From max at spevack.org Mon Jan 1 21:20:21 2007 From: max at spevack.org (Max Spevack) Date: Mon, 1 Jan 2007 16:20:21 -0500 (EST) Subject: fedora board meeting jan 2 Message-ID: Late notice with the end of the holiday, but we're scheduled for a Board meeting on January 2nd, at 10:00 AM Eastern time, 3:00 PM GMT. Assuming that we have a quorum of people back from vacation, we'll go ahead with the meeting. - get some update from Seth/Jeremy about how RPM is moving along after a few weeks since our announcement and creation of new repo/lists/etc - Max will talk about some of his priorities for Fedora in the first part of the year - Rex or others, get an update on how FESCO is right now, what the top concerns are regarding Fedora 7 and our progress on that. - whatever else comes up, or whatever folks think of between now and then. #fedora-board for the logs. Happy New Year, Max From matt at domsch.com Tue Jan 2 03:58:42 2007 From: matt at domsch.com (Matt Domsch) Date: Mon, 1 Jan 2007 21:58:42 -0600 Subject: fedora board meeting jan 2 In-Reply-To: References: Message-ID: <20070102035842.GA30961@domsch.com> On Mon, Jan 01, 2007 at 04:20:21PM -0500, Max Spevack wrote: > Late notice with the end of the holiday, but we're scheduled for a Board > meeting on January 2nd, at 10:00 AM Eastern time, 3:00 PM GMT. Assuming > that we have a quorum of people back from vacation, we'll go ahead with > the meeting. Logs and summary of the last meeting are at: http://fedoraproject.org/wiki/Board/Meetings/2006-12-19 Logs and summary of previous meetings are at http://fedoraproject.org/wiki/Board/Meetings All errors/omissions from the last 3 are mine. Thanks, Matt From blizzard at redhat.com Tue Jan 2 05:24:37 2007 From: blizzard at redhat.com (Christopher Blizzard) Date: Tue, 02 Jan 2007 00:24:37 -0500 Subject: fedora board meeting jan 2 In-Reply-To: References: Message-ID: <4599EC95.4080500@redhat.com> Have to send regrets - I will be on a plane at that time. --Chris From fedora at leemhuis.info Tue Jan 2 06:13:28 2007 From: fedora at leemhuis.info (Thorsten Leemhuis) Date: Tue, 02 Jan 2007 07:13:28 +0100 Subject: fedora board meeting jan 2 In-Reply-To: References: Message-ID: <4599F808.30103@leemhuis.info> On 01.01.2007 22:20, Max Spevack wrote: > Late notice with the end of the holiday, but we're scheduled for a Board > meeting on January 2nd, at 10:00 AM Eastern time, 3:00 PM GMT. Assuming > that we have a quorum of people back from vacation, we'll go ahead with > the meeting. > > - Rex or others, get an update on how FESCO is right now, what the top > concerns are regarding Fedora 7 and our progress on that. I think we have a deadlock here -- FESCo is mostly waiting for advice from the board how to proceed (?). Some people from the Board and FESCo probably should meet once or twice on IRC or somewhere else and just sort the details out. CU thl (?) -- think like FESCo size, mandatory level of community members in FESCo, how does FESCo interact with release engineering, red hat employees or the packaging committee, stuff like that From sundaram at fedoraproject.org Wed Jan 3 02:33:37 2007 From: sundaram at fedoraproject.org (Rahul Sundaram) Date: Wed, 03 Jan 2007 08:03:37 +0530 Subject: [fab] Re: Licensing audit for Fedora Extras In-Reply-To: <1165280311.20269.198.camel@localhost.localdomain> References: <45464A8A.6080503@fedoraproject.org> <1162234672.7933.206.camel@dhcp-32-122.ord.redhat.com> <4574C247.4090105@fedoraproject.org> <1165280311.20269.198.camel@localhost.localdomain> Message-ID: <459B1601.4020507@fedoraproject.org> Tom 'spot' Callaway wrote: > January? :) > > I hate to keep bumping this out, but the audit (even with help) will > take up ALL of my free time, and I really want to get Aurora 3 done (it > is taking up all of my free time, except for some Zelda while things are > building). > > The FSF will be there in 2007. Happy new year folks. January arrived not too long ago. Rahul From luis at tieguy.org Wed Jan 3 08:59:14 2007 From: luis at tieguy.org (Luis Villa) Date: Wed, 3 Jan 2007 03:59:14 -0500 Subject: why I'm using Ubuntu instead of Fedora ATM Message-ID: <2cb10c440701030059l49323ecaqeae7c7041f735d63@mail.gmail.com> So... after my list intro, where I mentioned that I'm using Ubuntu right now, Bill asked me to explain why I'm using Ubuntu instead of Fedora. Here they are. Note that while some of these reasons are probably broadly held, others are 'expert' issues and solving them probably wouldn't help Fedora's broader popularity. I should note that I'm not interested in getting into long discussions about whether or not my reasons were *correct*; they were what they were. If you think they were wrong, the discussion should be about how to communicate the correct message to the broad public, not about me or my reasons. reasons I went to Ubuntu (from suse, FWIW): * integrated/reasonably tested package availability- until recently, Ubuntu had a clear edge in the number of up-to-date, installable packages. That edge is no longer so clear-cut, but community perception seems to lag the reality. * support: whether or not it is reasonable/sustainable, Ubuntu's support policies for their non-LTS distros are more generous and more sane (i.e., all backports, no new features[1]) than Fedora's, which is a factor for someone like me who doesn't have much time to screw around with installations, re-installations, new releases that introduce new bugs, etc. * QA: Ubuntu aggressively pushes people to use their development branch and report problems, which leads to better, more stable final releases. At the time I chose to use Ubuntu, people were not just not encouraged, but actively discouraged from using rawhide. This is improving, but at the time, I felt that this difference meant that Ubuntu would be better QA'd (and it was no surprise, at least to me, that Ubuntu has not yet had bad release bugs like the FC5 GPL module kernel bug.[2]) * release cycle: the predictability of Ubuntu's release cycle is nice. Not a huge deal, but nice, and reassures me that I've got reasonably fresh software all the time. * release discipline: again, this is from an outsider's perspective, but it appears that Ubuntu adheres more closely to a series of reasonable freezes than Fedora does. This of course slows development and requires more divergence from upstream, but it also makes for a more reliable and better-tested end-product.[3] * reliable, timely access to GNOME development releases: this was probably the prime reason for my initial move to Ubuntu, since at the time I was trying to move away from building all of GNOME from CVS, but still wanted to use and test the development branch. I expect that this is a fairly minority issue. :) * liveCD, single-CD: these were useful to me personally, and I get the sense that people at least talk about using this to demo things for other people. Whether or not they actually use it that way is irrelevant; they *feel* that this makes Ubuntu more potentially useful to them. * 'independence': as some of you know, I advocate strong transparency in the relationship between the enterprise distro and the community distro. I think Ubuntu mostly has that; at least, no one can point to any potential undisclosed conflicts of interest between LTS and the other releases. Fedora does not; the fudging around Xen and FC6's release delays I think demonstrate that fairly clearly. This relationship may well be unavoidable, given where Fedora's resources come from, but in that case, I think Fedora should be more transparent and public about the relationship between the two; in general, I think such transparency would both aid Fedora's PR and make Fedora's decision making easier.[4] * 'trust': I don't personally particularly trust Mark; he has said repeatedly that Canonical is a business he will seek to profit from, so I trust him just as far as I can throw him and his contracts with me- the same treatment I give any profit-motivated CEO. However, I think that the conventional wisdom is that Mark is doing this primarily as a charity and hence is more trustworthy than Red Hat. I don't think that there is a whole lot Red Hat/Fedora can do here (short of substantially endowing Fedora as a Moz-style 501(c)(3)), but the more transparency in the relationship, the less it will come up. * community growth: this wasn't a factor for me, but Ubuntu has very aggressively and very publicly pursued non-engineering community involvement. Make people feel wanted and love, and they'll want to use your distro. I think that is about it. The main reasons I'm currently considering moving away from Ubuntu: * freedom- Ubuntu's lack of a clear patent policy and move towards more extensive use of non-free drivers is very troubling to me. * malone- god their bug tracker is a gigantic steaming pile of shit, even moreso than bugzilla. (And proprietary, but so far every distro except Debian has some proprietary tools, so I'm not going to jump too hard on them for that yet.) Anyway, hope all this is useful; I'd be happy to elaborate both on my own motivations and what I perceive to be the conventional wisdom if I've been unclear or underspecific. Luis Footnotes: [1] New upstream releases/features in a stable release give the lie to any claim that the distro is 'stable', or alternately, admits that the distro has no substantial QA role. Some might argue that the distro has no substantial QA role, but I think most people have that expectation. (The testing channel that has been mentioned would go a long way towards resolving my concerns here, but until that is up and has active users, and hard rules about regressions, the bulk of my objections still apply.) [2] http://lwn.net/Articles/175793/ ; in particular the comment at http://lwn.net/Articles/175800/ captures what I think is the conventional wisdom about Fedora's quality. [3] Note, for example, the upstream version freeze in https://wiki.ubuntu.com/FeistyReleaseSchedule , lacking in http://fedoraproject.org/wiki/Core/Schedule . Ubuntu's feature freeze is a week further from their final release (not a huge deal) but AFAICT more strictly adhered to (a big deal). [4] The prime example that comes to my mind is the BS about Fedora not being a beta for RHEL. That is crap; anyone who says otherwise is nobly deluding themselves. I'd probably argue that this isn't even *desirable*- if Fedora is a meaningful part of the QA process for RHEL, RH has greater incentive to invest in Fedora, and as long as Fedora has sane, transparent, and independent release freezes/QA (probably not the case right now, to wit Xen) Fedora being a beta for RHEL shouldn't damage Fedora's quality. From fedora at leemhuis.info Wed Jan 3 09:45:45 2007 From: fedora at leemhuis.info (Thorsten Leemhuis) Date: Wed, 03 Jan 2007 10:45:45 +0100 Subject: why I'm using Ubuntu instead of Fedora ATM In-Reply-To: <2cb10c440701030059l49323ecaqeae7c7041f735d63@mail.gmail.com> References: <2cb10c440701030059l49323ecaqeae7c7041f735d63@mail.gmail.com> Message-ID: <459B7B49.8040101@leemhuis.info> Hi! Thx Luis, I liked you mail. Some comments from my side: On 03.01.2007 09:59, Luis Villa wrote: > [...] > * support: whether or not it is reasonable/sustainable, Ubuntu's > support policies for their non-LTS distros are more generous and more > sane (i.e., all backports, no new features[1]) than Fedora's, which is > a factor for someone like me who doesn't have much time to screw > around with installations, re-installations, new releases that > introduce new bugs, etc. Agreed. We always point people to RHEL or CentOS and I think that becomes more and more a problem, especially now that Legacy is dead. A Fedora LTS (two years? maybe the server parts ever three?) now and then (every second or third release?) from a new Fedora Legacy (needs a different name) would IMHO a nice solution. > * QA: Ubuntu aggressively pushes people to use their development > branch and report problems, which leads to better, more stable final > releases. [...] Agreed. The biggest problem we have in this regard IMHO is that we always communicate "you can't get to stable release from rawhide or a test release". That scares people aways from the devel branch and the test releases. We should provide a clean solution so people at least can get from test3 to stable. > * release cycle: the predictability of Ubuntu's release cycle is nice. > Not a huge deal, but nice, and reassures me that I've got reasonably > fresh software all the time. Strongly agreed. > [...] > * reliable, timely access to GNOME development releases: this was > probably the prime reason for my initial move to Ubuntu, since at the > time I was trying to move away from building all of GNOME from CVS, > but still wanted to use and test the development branch. I expect that > this is a fairly minority issue. :) Strongly agreed. > * liveCD, single-CD: these were useful to me personally, and I get the > sense that people at least talk about using this to demo things for > other people. Whether or not they actually use it that way is > irrelevant; they *feel* that this makes Ubuntu more potentially useful > to them. Single-CD would be nice, but what I'd like to see even more is a single-sided one-layer DVD (e.g. 4,x GB) that contains a x86-Live-CD and is able to install x86 oder x64 to hard disk using anaconda. That would be ideal for fairs and computer magazines to distribute as it contains all on one DVD what most people need (yes, I know, I'm leaving regions aside where DVDs are still uncommon). > [...] > * community growth: this wasn't a factor for me, but Ubuntu has very > aggressively and very publicly pursued non-engineering community > involvement. Make people feel wanted and love, and they'll want to use > your distro. Strongly agreed. I more and more think we need a "Fedora Experimental Kitchen" project where stuff that's not yet covered by the Fedora Project can be developed while the users feel as being a part of the Fedora project -> this would help getting people involved and grow up. Kmods, alternate kernels, Firefox2 for FC6, Respins, Live-CDs, new Distributions Spins (Fedora Audio, Fedora BrandNewIdea anyone?) could be suitable to be done under the hood of the "Fedora Experimental Kitchen". >[...] > [1] New upstream releases/features in a stable release give the lie to > any claim that the distro is 'stable', or alternately, admits that the > distro has no substantial QA role. Some might argue that the distro > has no substantial QA role, but I think most people have that > expectation. (The testing channel that has been mentioned would go a > long way towards resolving my concerns here, but until that is up and > has active users, and hard rules about regressions, the bulk of my > objections still apply.) My 2 cent on this: There are a lot of people around that seem to like Fedora because you get a lot of new software even for a stable version. But yes, there are others that would like a more careful approach. I'm wondering if we could provide solutions for both users: One update channel that only gets security updates and important bugfixes while the other is a bit more bold -- we for example could have firefox2 in the bold channel for FC6 while shipping the latest firefox 1.5.x in the more conservative channel. >[...] CU thl From rgarth at gmail.com Wed Jan 3 10:10:40 2007 From: rgarth at gmail.com (Rob Garth) Date: Wed, 3 Jan 2007 21:10:40 +1100 Subject: why I'm using Ubuntu instead of Fedora ATM In-Reply-To: <459B7B49.8040101@leemhuis.info> References: <2cb10c440701030059l49323ecaqeae7c7041f735d63@mail.gmail.com> <459B7B49.8040101@leemhuis.info> Message-ID: <183d138b0701030210g7127c6a0s10669ce333231cd6@mail.gmail.com> I am not part of the Fedora Project and so I guees my opinion is fairly meaningless but on the LTS support stuff. FC6 is the upstream of RHEL5. This means that resources are already going into supporting basically the same code base, I know it is a fork, but there is more similar than different, so could the fedora releases that are upstream of RHEL be LTS releases? 2 years would be lovely. Agreed. We always point people to RHEL or CentOS and I think that > becomes more and more a problem, especially now that Legacy is dead. > > A Fedora LTS (two years? maybe the server parts ever three?) now and > then (every second or third release?) from a new Fedora Legacy (needs a > different name) would IMHO a nice solution. -------------- next part -------------- An HTML attachment was scrubbed... URL: From bugs.michael at gmx.net Wed Jan 3 12:42:25 2007 From: bugs.michael at gmx.net (Michael Schwendt) Date: Wed, 3 Jan 2007 13:42:25 +0100 Subject: why I'm using Ubuntu instead of Fedora ATM In-Reply-To: <2cb10c440701030059l49323ecaqeae7c7041f735d63@mail.gmail.com> References: <2cb10c440701030059l49323ecaqeae7c7041f735d63@mail.gmail.com> Message-ID: <20070103134225.de8a7906.bugs.michael@gmx.net> On Wed, 3 Jan 2007 03:59:14 -0500, Luis Villa wrote: > * QA: Ubuntu aggressively pushes people to use their development ^^^^^^ "people" == "all people"? i.e. not just users, but also packagers > branch and report problems, which leads to better, more stable final > releases. At the time I chose to use Ubuntu, people were not just not > encouraged, but actively discouraged from using rawhide. This is > improving, but at the time, I felt that this difference meant that > Ubuntu would be better QA'd (and it was no surprise, at least to me, > that Ubuntu has not yet had bad release bugs like the FC5 GPL module > kernel bug.[2]) The first thing that would need to happen is to push the developers and packagers to keep Rawhide *including* "Fedora Extras development" up-to-date. Plus bringing back the goal to have *all* packages rebuilt for Test1. It is not just the infamous "Package EVR problems in FC+FE" report, which has been posted to fedora-maintainers list for quite a while, that shows how the development distribution is neglected by the packagers. It lists problems where software _versions_ (not just package release numbers) are lower than what has been published for the older distributions. Only if it is possible to upgrade from FC-6 to Rawhide, users can be asked to try out the development packages. If, however, many packagers don't even care about submitting simple package rebuild jobs, there is a lot of undiscovered breakage. Compilation failures, API changes, run-time breakage after successful rebuilds, necessity to request help from upstream, which may take some time, plus other fun. And the late discovery of such problems when e.g. FESCO takes over and performs a semi-automated mass-rebuild close before the release of Fedora Core. From tiemann at redhat.com Wed Jan 3 14:22:43 2007 From: tiemann at redhat.com (Michael Tiemann) Date: Wed, 03 Jan 2007 09:22:43 -0500 Subject: why I'm using Ubuntu instead of Fedora ATM In-Reply-To: <183d138b0701030210g7127c6a0s10669ce333231cd6@mail.gmail.com> References: <2cb10c440701030059l49323ecaqeae7c7041f735d63@mail.gmail.com> <459B7B49.8040101@leemhuis.info> <183d138b0701030210g7127c6a0s10669ce333231cd6@mail.gmail.com> Message-ID: <1167834163.4869.2.camel@localhost.localdomain> On Wed, 2007-01-03 at 21:10 +1100, Rob Garth wrote: > I am not part of the Fedora Project and so I guees my opinion is > fairly meaningless but on the LTS support stuff. FC6 is the upstream > of RHEL5. ...and so much more. It is not *just* the upstream of Red Hat's commercial products. M From mmcgrath at fedoraproject.org Wed Jan 3 14:49:09 2007 From: mmcgrath at fedoraproject.org (Mike McGrath) Date: Wed, 3 Jan 2007 08:49:09 -0600 Subject: why I'm using Ubuntu instead of Fedora ATM In-Reply-To: <1167834163.4869.2.camel@localhost.localdomain> References: <2cb10c440701030059l49323ecaqeae7c7041f735d63@mail.gmail.com> <459B7B49.8040101@leemhuis.info> <183d138b0701030210g7127c6a0s10669ce333231cd6@mail.gmail.com> <1167834163.4869.2.camel@localhost.localdomain> Message-ID: <3237e4410701030649r51e1a4ccw30a89d6df6124d2d@mail.gmail.com> On 1/3/07, Michael Tiemann wrote: > On Wed, 2007-01-03 at 21:10 +1100, Rob Garth wrote: > > I am not part of the Fedora Project and so I guees my opinion is > > fairly meaningless but on the LTS support stuff. FC6 is the upstream > > of RHEL5. > > ...and so much more. It is not *just* the upstream of Red Hat's > commercial products. > > M > I hear this argument a lot. The fact is that Fedora is the 'upstream' for many projects. We're often the first distro to pick up a new technology and integrate it. Just because RHEL is one distro that gets it after us shouldn't be counted against Fedora. -Mike From sundaram at fedoraproject.org Wed Jan 3 15:00:35 2007 From: sundaram at fedoraproject.org (Rahul Sundaram) Date: Wed, 03 Jan 2007 20:30:35 +0530 Subject: why I'm using Ubuntu instead of Fedora ATM In-Reply-To: <183d138b0701030210g7127c6a0s10669ce333231cd6@mail.gmail.com> References: <2cb10c440701030059l49323ecaqeae7c7041f735d63@mail.gmail.com> <459B7B49.8040101@leemhuis.info> <183d138b0701030210g7127c6a0s10669ce333231cd6@mail.gmail.com> Message-ID: <459BC513.3020107@fedoraproject.org> Rob Garth wrote: > I am not part of the Fedora Project and so I guees my opinion is fairly > meaningless but on the LTS support stuff. FC6 is the upstream of RHEL5. > This means that resources are already going into supporting basically > the same code base, I know it is a fork, but there is more similar than > different, so could the fedora releases that are upstream of RHEL be LTS > releases? 2 years would be lovely. That fact that RHEL is forked and maintained for a longer period doesnt help Fedora retain a longer lifecycle. The general policy on updates is way different. This becomes more and more visible over after the initial release. Rahul From jkeating at redhat.com Wed Jan 3 15:29:23 2007 From: jkeating at redhat.com (Jesse Keating) Date: Wed, 3 Jan 2007 10:29:23 -0500 Subject: why I'm using Ubuntu instead of Fedora ATM In-Reply-To: <459B7B49.8040101@leemhuis.info> References: <2cb10c440701030059l49323ecaqeae7c7041f735d63@mail.gmail.com> <459B7B49.8040101@leemhuis.info> Message-ID: <200701031029.23569.jkeating@redhat.com> On Wednesday 03 January 2007 04:45, Thorsten Leemhuis wrote: > Agreed. We always point people to RHEL or CentOS and I think that > becomes more and more a problem, especially now that Legacy is dead. > > A Fedora LTS (two years? maybe the server parts ever three?) now and > then (every second or third release?) from a new Fedora Legacy (needs a > different name) would IMHO a nice solution. My question here is who will do the work? If you want to feed from RHEL updates, then you're basically only going to be doing backports, no new versions. You're also only going to be fixing bugs that are deemed worthy through the RHEL process, which usually needs a paid customer reporting the problem, etc, etc... They are different beasts and the work still has to be done to build packages, push them through a QA / update process, take bugs in, do QA, etc, etc, etc... -- Jesse Keating Release Engineer: Fedora -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 189 bytes Desc: not available URL: From katzj at redhat.com Wed Jan 3 15:32:57 2007 From: katzj at redhat.com (Jeremy Katz) Date: Wed, 03 Jan 2007 10:32:57 -0500 Subject: why I'm using Ubuntu instead of Fedora ATM In-Reply-To: <200701031029.23569.jkeating@redhat.com> References: <2cb10c440701030059l49323ecaqeae7c7041f735d63@mail.gmail.com> <459B7B49.8040101@leemhuis.info> <200701031029.23569.jkeating@redhat.com> Message-ID: <1167838377.14822.6.camel@aglarond.local> On Wed, 2007-01-03 at 10:29 -0500, Jesse Keating wrote: > On Wednesday 03 January 2007 04:45, Thorsten Leemhuis wrote: > > Agreed. We always point people to RHEL or CentOS and I think that > > becomes more and more a problem, especially now that Legacy is dead. > > > > A Fedora LTS (two years? maybe the server parts ever three?) now and > > then (every second or third release?) from a new Fedora Legacy (needs a > > different name) would IMHO a nice solution. > > My question here is who will do the work? If you want to feed from RHEL > updates, then you're basically only going to be doing backports, no new > versions. You're also only going to be fixing bugs that are deemed worthy > through the RHEL process, which usually needs a paid customer reporting the > problem, etc, etc... They are different beasts and the work still has to be > done to build packages, push them through a QA / update process, take bugs > in, do QA, etc, etc, etc... There are also (non-trivial) differences in the package set. The Fedora universe is large... :) Jeremy From notting at redhat.com Wed Jan 3 16:39:03 2007 From: notting at redhat.com (Bill Nottingham) Date: Wed, 3 Jan 2007 11:39:03 -0500 Subject: why I'm using Ubuntu instead of Fedora ATM In-Reply-To: <459B7B49.8040101@leemhuis.info> References: <2cb10c440701030059l49323ecaqeae7c7041f735d63@mail.gmail.com> <459B7B49.8040101@leemhuis.info> Message-ID: <20070103163903.GA22376@nostromo.devel.redhat.com> Thorsten Leemhuis (fedora at leemhuis.info) said: > >* support: whether or not it is reasonable/sustainable, Ubuntu's > >support policies for their non-LTS distros are more generous and more > >sane (i.e., all backports, no new features[1]) than Fedora's, which is > >a factor for someone like me who doesn't have much time to screw > >around with installations, re-installations, new releases that > >introduce new bugs, etc. > > Agreed. We always point people to RHEL or CentOS and I think that > becomes more and more a problem, especially now that Legacy is dead. > > A Fedora LTS (two years? maybe the server parts ever three?) now and > then (every second or third release?) from a new Fedora Legacy (needs a > different name) would IMHO a nice solution. I think you missed the point of what he said. He was talking about how the *normal* release was handled - nothing about long-term support. As for any sort of long-term Fedora support, what we need to see is some sort of market for it - we had the inital Legacy, and, realistically, NO ONE WANTED IT ENOUGH to actually work on it. Considering that we are dependent in many ways on community resources for this, I don't think it's worth throwing resources at a problem that may not exist. It's all a tradeoff - is a better use of resource to make the next release better, or maintaining the old releases? Moreover, considering he's talking about 'all backports - no new features'; Extras is *significantly* worse than the current Core in this regard. > >* community growth: this wasn't a factor for me, but Ubuntu has very > >aggressively and very publicly pursued non-engineering community > >involvement. Make people feel wanted and love, and they'll want to use > >your distro. > > Strongly agreed. I more and more think we need a "Fedora Experimental > Kitchen" project where stuff that's not yet covered by the Fedora > Project can be developed while the users feel as being a part of the > Fedora project -> this would help getting people involved and grow up. > > Kmods, alternate kernels, Firefox2 for FC6, Respins, Live-CDs, new > Distributions Spins (Fedora Audio, Fedora BrandNewIdea anyone?) could be > suitable to be done under the hood of the "Fedora Experimental Kitchen". OK, you've lost me. How is adding an experimental repo for highly technical things (kmods, alternate kernels, etc) about embracing the non-engineering community? > I'm wondering if we could provide solutions for both users: One update > channel that only gets security updates and important bugfixes while the > other is a bit more bold -- we for example could have firefox2 in the > bold channel for FC6 while shipping the latest firefox 1.5.x in the more > conservative channel. So, an idea like this: - starts to exponentially expand the QA problem - breaks dependencies across repositories (we have no xulrunner) - fractures the community into different splinters Bill From luis at tieguy.org Wed Jan 3 16:58:05 2007 From: luis at tieguy.org (Luis Villa) Date: Wed, 3 Jan 2007 11:58:05 -0500 Subject: why I'm using Ubuntu instead of Fedora ATM In-Reply-To: <459B7B49.8040101@leemhuis.info> References: <2cb10c440701030059l49323ecaqeae7c7041f735d63@mail.gmail.com> <459B7B49.8040101@leemhuis.info> Message-ID: <2cb10c440701030858g4a0c84cdi3cc1f67186b3258f@mail.gmail.com> On 1/3/07, Thorsten Leemhuis wrote: > Hi! > > Thx Luis, I liked you mail. Some comments from my side: > > On 03.01.2007 09:59, Luis Villa wrote: > > [...] > > > * support: whether or not it is reasonable/sustainable, Ubuntu's > > support policies for their non-LTS distros are more generous and more > > sane (i.e., all backports, no new features[1]) than Fedora's, which is > > a factor for someone like me who doesn't have much time to screw > > around with installations, re-installations, new releases that > > introduce new bugs, etc. > > Agreed. We always point people to RHEL or CentOS and I think that > becomes more and more a problem, especially now that Legacy is dead. > > A Fedora LTS (two years? maybe the server parts ever three?) now and > then (every second or third release?) from a new Fedora Legacy (needs a > different name) would IMHO a nice solution. I don't think the need (at least my need) is for a Fedora LTS; I merely ask for one *stable* release at a time that: (1) doesn't get new features/new upstream releases (2) upgrades cleanly to the next release with a simple command If I have this, I can (as a hobbyist) reasonably rely on Fedora for most of my needs with minimal pain between releases. Right now I can't do that. I understand that (2) is basically there; (1) is (AFAICT) definitely not. (FWIW, I think it is unreasaonble to expect a true-community distro to do real LTS-y stuff- most volunteers don't have the patience to do the necessary backporting for the necessary length of time. (See Fedora Legacy.) And it is obviously not reasonable to expect RH to pay for anyone to do LTS-y stuff with Fedora.) > > * QA: Ubuntu aggressively pushes people to use their development > > branch and report problems, which leads to better, more stable final > > releases. [...] > > Agreed. The biggest problem we have in this regard IMHO is that we > always communicate "you can't get to stable release from rawhide or a > test release". That scares people aways from the devel branch and the > test releases. We should provide a clean solution so people at least can > get from test3 to stable. No, you should provide a clean path from every package ver X to every package ver X+1. Period. That is what Debian has done since time basically immemorial, and what Ubuntu does. Clean upgrades from only test 3 means that you only get good testing after test 3- again, insufficient for real quality. You need testing of the latest code *every day* to get the best quality. Let me be clear- I feel that quality is one of the biggest possible advantages free software can have over proprietary software, specifically because you can have hundreds or thousands of people testing the latest code every day. If you're not taking advantage of that, you're throwing away one of the biggest advantages we have over proprietary software. And I think Fedora currently does that, in the main. Luis From gdk at redhat.com Wed Jan 3 16:56:26 2007 From: gdk at redhat.com (Greg Dekoenigsberg) Date: Wed, 3 Jan 2007 11:56:26 -0500 (EST) Subject: why I'm using Ubuntu instead of Fedora ATM In-Reply-To: <20070103163903.GA22376@nostromo.devel.redhat.com> References: <2cb10c440701030059l49323ecaqeae7c7041f735d63@mail.gmail.com> <459B7B49.8040101@leemhuis.info> <20070103163903.GA22376@nostromo.devel.redhat.com> Message-ID: On Wed, 3 Jan 2007, Bill Nottingham wrote: >> Strongly agreed. I more and more think we need a "Fedora Experimental >> Kitchen" project where stuff that's not yet covered by the Fedora >> Project can be developed while the users feel as being a part of the >> Fedora project -> this would help getting people involved and grow up. >> >> Kmods, alternate kernels, Firefox2 for FC6, Respins, Live-CDs, new >> Distributions Spins (Fedora Audio, Fedora BrandNewIdea anyone?) could be >> suitable to be done under the hood of the "Fedora Experimental Kitchen". > > OK, you've lost me. How is adding an experimental repo for highly > technical things (kmods, alternate kernels, etc) about embracing > the non-engineering community? It's not. It's about embracing the influential communities of developers who are trying to Get Shit Done -- which is, from my perspective, even *more* important. If Fedora becomes the platform from which innovators choose to share their innovations, Fedora becomes stronger. It's all about the applications. >> I'm wondering if we could provide solutions for both users: One update >> channel that only gets security updates and important bugfixes while the >> other is a bit more bold -- we for example could have firefox2 in the >> bold channel for FC6 while shipping the latest firefox 1.5.x in the more >> conservative channel. > > So, an idea like this: > > - starts to exponentially expand the QA problem Bill, I love you, but I'm going to have to call bullshit on this one. The answer is not "avoid making QA harder", the answer is to SOLVE THE QA PROBLEM. Most users don't report bugs worth a crap, and when they do report bugs, those bug reports, as a rule, suck ass. Why? Because they have no idea how to do it -- and we haven't done *even a little bit of work* to make it any easier for them. If there's two things I'd like to see Will doing, right now, they would be: 1. Come up with a better tool for reporting bugs -- maybe in bugzilla, maybe elsewhere, maybe a layer in-between. Some kind of sysdump tool that can do stuff like, oh, look in /etc/yum.repos.d/ and, if experimental repos are found, mark the new bug accordingly. 2. Make the Fedora Triage team happen. > - breaks dependencies across repositories (we have no xulrunner) Let's work on tools and policy to address this issue, then. Because for a lot of users, this is already a problem. I still use the CCRMA repo, because CCRMA still hasn't integrated into Fedora, because CCRMA uses a RT kernel. > - fractures the community into different splinters Smells like FUD to me. --g ------------------------------------------------------------- Greg DeKoenigsberg || Fedora Project || fedoraproject.org Be an Ambassador || http://fedoraproject.org/wiki/Ambassadors ------------------------------------------------------------- From luis at tieguy.org Wed Jan 3 17:00:56 2007 From: luis at tieguy.org (Luis Villa) Date: Wed, 3 Jan 2007 12:00:56 -0500 Subject: why I'm using Ubuntu instead of Fedora ATM In-Reply-To: <1167834163.4869.2.camel@localhost.localdomain> References: <2cb10c440701030059l49323ecaqeae7c7041f735d63@mail.gmail.com> <459B7B49.8040101@leemhuis.info> <183d138b0701030210g7127c6a0s10669ce333231cd6@mail.gmail.com> <1167834163.4869.2.camel@localhost.localdomain> Message-ID: <2cb10c440701030900g379d3fcdw5b14bfd21925b2a1@mail.gmail.com> On 1/3/07, Michael Tiemann wrote: > On Wed, 2007-01-03 at 21:10 +1100, Rob Garth wrote: > > I am not part of the Fedora Project and so I guees my opinion is > > fairly meaningless but on the LTS support stuff. FC6 is the upstream > > of RHEL5. > > ...and so much more. It is not *just* the upstream of Red Hat's > commercial products. Sure, but that is still the primary motivation for most of the hours and money that goes into Fedora. To pretend that that is just one goal among many makes Fedora look dishonest. To be sure, it is admirable (and I think admired) to say that the long-term goal is for RHEL to be just one consumer among many. But that isn't the reality on the ground right now, and to deny that (as many Fedora-affiliated people have tried to do from time to time) is silly. Luis From gdk at redhat.com Wed Jan 3 16:59:24 2007 From: gdk at redhat.com (Greg Dekoenigsberg) Date: Wed, 3 Jan 2007 11:59:24 -0500 (EST) Subject: why I'm using Ubuntu instead of Fedora ATM In-Reply-To: <2cb10c440701030900g379d3fcdw5b14bfd21925b2a1@mail.gmail.com> References: <2cb10c440701030059l49323ecaqeae7c7041f735d63@mail.gmail.com> <459B7B49.8040101@leemhuis.info> <183d138b0701030210g7127c6a0s10669ce333231cd6@mail.gmail.com> <1167834163.4869.2.camel@localhost.localdomain> <2cb10c440701030900g379d3fcdw5b14bfd21925b2a1@mail.gmail.com> Message-ID: On Wed, 3 Jan 2007, Luis Villa wrote: > On 1/3/07, Michael Tiemann wrote: >> On Wed, 2007-01-03 at 21:10 +1100, Rob Garth wrote: >> > I am not part of the Fedora Project and so I guees my opinion is >> > fairly meaningless but on the LTS support stuff. FC6 is the upstream >> > of RHEL5. >> >> ...and so much more. It is not *just* the upstream of Red Hat's >> commercial products. > > Sure, but that is still the primary motivation for most of the hours > and money that goes into Fedora. To pretend that that is just one goal > among many makes Fedora look dishonest. To be sure, it is admirable > (and I think admired) to say that the long-term goal is for RHEL to be > just one consumer among many. But that isn't the reality on the ground > right now, and to deny that (as many Fedora-affiliated people have > tried to do from time to time) is silly. +1. It's worthwhile stating, as a goal, that Fedora seeks to be the upstream for many projects -- but RHEL will *always* be the, ah, "first among equals" in that regard. --g ------------------------------------------------------------- Greg DeKoenigsberg || Fedora Project || fedoraproject.org Be an Ambassador || http://fedoraproject.org/wiki/Ambassadors ------------------------------------------------------------- From sundaram at fedoraproject.org Wed Jan 3 16:59:35 2007 From: sundaram at fedoraproject.org (Rahul Sundaram) Date: Wed, 03 Jan 2007 22:29:35 +0530 Subject: why I'm using Ubuntu instead of Fedora ATM In-Reply-To: <2cb10c440701030900g379d3fcdw5b14bfd21925b2a1@mail.gmail.com> References: <2cb10c440701030059l49323ecaqeae7c7041f735d63@mail.gmail.com> <459B7B49.8040101@leemhuis.info> <183d138b0701030210g7127c6a0s10669ce333231cd6@mail.gmail.com> <1167834163.4869.2.camel@localhost.localdomain> <2cb10c440701030900g379d3fcdw5b14bfd21925b2a1@mail.gmail.com> Message-ID: <459BE0F7.7050401@fedoraproject.org> Luis Villa wrote: > On 1/3/07, Michael Tiemann wrote: >> On Wed, 2007-01-03 at 21:10 +1100, Rob Garth wrote: >> > I am not part of the Fedora Project and so I guees my opinion is >> > fairly meaningless but on the LTS support stuff. FC6 is the upstream >> > of RHEL5. >> >> ...and so much more. It is not *just* the upstream of Red Hat's >> commercial products. > > Sure, but that is still the primary motivation for most of the hours > and money that goes into Fedora. To pretend that that is just one goal > among many makes Fedora look dishonest. To be sure, it is admirable > (and I think admired) to say that the long-term goal is for RHEL to be > just one consumer among many. But that isn't the reality on the ground > right now, and to deny that (as many Fedora-affiliated people have > tried to do from time to time) is silly. It's not. Think OLPC for example. There is also stuff out there like Sony play stations, Cell processors etc. It already is being used as a base for variety of stuff out there. Rahul From notting at redhat.com Wed Jan 3 17:08:05 2007 From: notting at redhat.com (Bill Nottingham) Date: Wed, 3 Jan 2007 12:08:05 -0500 Subject: why I'm using Ubuntu instead of Fedora ATM In-Reply-To: References: <2cb10c440701030059l49323ecaqeae7c7041f735d63@mail.gmail.com> <459B7B49.8040101@leemhuis.info> <20070103163903.GA22376@nostromo.devel.redhat.com> Message-ID: <20070103170805.GA23011@nostromo.devel.redhat.com> Greg Dekoenigsberg (gdk at redhat.com) said: > >>Strongly agreed. I more and more think we need a "Fedora Experimental > >>Kitchen" project where stuff that's not yet covered by the Fedora > >>Project can be developed while the users feel as being a part of the > >>Fedora project -> this would help getting people involved and grow up. > >> > >>Kmods, alternate kernels, Firefox2 for FC6, Respins, Live-CDs, new > >>Distributions Spins (Fedora Audio, Fedora BrandNewIdea anyone?) could be > >>suitable to be done under the hood of the "Fedora Experimental Kitchen". > > > >OK, you've lost me. How is adding an experimental repo for highly > >technical things (kmods, alternate kernels, etc) about embracing > >the non-engineering community? > > It's not. It's about embracing the influential communities of developers > who are trying to Get Shit Done -- which is, from my perspective, even > *more* important. If Fedora becomes the platform from which innovators > choose to share their innovations, Fedora becomes stronger. It's all > about the applications. OK. That's different from what Luis said, though. :) > >>I'm wondering if we could provide solutions for both users: One update > >>channel that only gets security updates and important bugfixes while the > >>other is a bit more bold -- we for example could have firefox2 in the > >>bold channel for FC6 while shipping the latest firefox 1.5.x in the more > >>conservative channel. > > > >So, an idea like this: > > > >- starts to exponentially expand the QA problem > > Bill, I love you, but I'm going to have to call bullshit on this one. > The answer is not "avoid making QA harder", the answer is to SOLVE THE QA > PROBLEM. Sure, but which of these plans make more sense: 1) - Solve the QA problem for our repo configurations as they exist - Expand the QA solution to new, multiple, disparate and conflicting repositories 2) - Expand into new, multiple, disparate and conflicting repositories - Then try to solve the QA problem Honestly, before we can do multiple experimental repositories of this, that, and the other, we need to get our *OWN* house in order. Bill From fedora at leemhuis.info Wed Jan 3 17:12:13 2007 From: fedora at leemhuis.info (Thorsten Leemhuis) Date: Wed, 03 Jan 2007 18:12:13 +0100 Subject: why I'm using Ubuntu instead of Fedora ATM In-Reply-To: <20070103163903.GA22376@nostromo.devel.redhat.com> References: <2cb10c440701030059l49323ecaqeae7c7041f735d63@mail.gmail.com> <459B7B49.8040101@leemhuis.info> <20070103163903.GA22376@nostromo.devel.redhat.com> Message-ID: <459BE3ED.6060207@leemhuis.info> Bill Nottingham schrieb: > Thorsten Leemhuis (fedora at leemhuis.info) said: >>> * support: whether or not it is reasonable/sustainable, Ubuntu's >>> support policies for their non-LTS distros are more generous and more >>> sane (i.e., all backports, no new features[1]) than Fedora's, which is >>> a factor for someone like me who doesn't have much time to screw >>> around with installations, re-installations, new releases that >>> introduce new bugs, etc. >> Agreed. We always point people to RHEL or CentOS and I think that >> becomes more and more a problem, especially now that Legacy is dead. >> A Fedora LTS (two years? maybe the server parts ever three?) now and >> then (every second or third release?) from a new Fedora Legacy (needs a >> different name) would IMHO a nice solution. > I think you missed the point of what he said. He was talking about > how the *normal* release was handled - nothing about long-term support. Seems so. /me slashes thl /me makes mental note: don#t write mails in a hurry, looks bad > As for any sort of long-term Fedora support, what we need to see is > some sort of market for it - we had the inital Legacy, and, realistically, > NO ONE WANTED IT ENOUGH to actually work on it. Maybe it died because it/we tried to much? I think we should be able to get enough people together to support only one distro for a longer time at a certain period, e.g.: FC6 -> supported until FC8 get's out + one month = 13 months basic support. Support FC6 after that by a new Fedora Legacy for for another 18 months = 31 Month or round about two and a half years in total. FC11 would be out by then and we could start maintaining FC9 for another 18 months (it would be otherwise EOL by then)... > [...] > Moreover, considering he's talking about 'all backports - no new features'; > Extras is *significantly* worse than the current Core in this regard. Fully agreed, but Extras started as a rolling release scheme and never stopped and nobody yelled loudly to stop that. That will change with F7. >>> * community growth: this wasn't a factor for me, but Ubuntu has very >>> aggressively and very publicly pursued non-engineering community >>> involvement. Make people feel wanted and love, and they'll want to use >>> your distro. >> Strongly agreed. I more and more think we need a "Fedora Experimental >> Kitchen" project where stuff that's not yet covered by the Fedora >> Project can be developed while the users feel as being a part of the >> Fedora project -> this would help getting people involved and grow up. >> >> Kmods, alternate kernels, Firefox2 for FC6, Respins, Live-CDs, new >> Distributions Spins (Fedora Audio, Fedora BrandNewIdea anyone?) could be >> suitable to be done under the hood of the "Fedora Experimental Kitchen". > OK, you've lost me. How is adding an experimental repo for highly > technical things (kmods, alternate kernels, etc) about embracing > the non-engineering community? It was an idea that came to my mind when the recent kmod kernel thing came up and I thought I could boot my feed out a bit into the cold water with outlining them it a bit and to see what comes out of it. >> I'm wondering if we could provide solutions for both users: One update >> channel that only gets security updates and important bugfixes while the >> other is a bit more bold -- we for example could have firefox2 in the >> bold channel for FC6 while shipping the latest firefox 1.5.x in the more >> conservative channel. > So, an idea like this: > - starts to exponentially expand the QA problem The stable update channel could remain the official repo with a slightly more conservative approach then now. > - breaks dependencies across repositories (we have no xulrunner) The maintainers of the more bold channel would need to fix this, too. > - fractures the community into different splinters People want different things: group1: some want always the newest and greatest -> rawhide group2: some want a slightly more stable approach, but also quite new stuff (Exrtas rolling release) -> bold repo group3: some want only security fixes and important bugfixes (traditional approach) I'd like to have group2 on board, even if there is a small split. *Maybe* we would not need it if the devel tree was a bit more stable. CU thl From luis at tieguy.org Wed Jan 3 17:13:07 2007 From: luis at tieguy.org (Luis Villa) Date: Wed, 3 Jan 2007 12:13:07 -0500 Subject: why I'm using Ubuntu instead of Fedora ATM In-Reply-To: <459BE0F7.7050401@fedoraproject.org> References: <2cb10c440701030059l49323ecaqeae7c7041f735d63@mail.gmail.com> <459B7B49.8040101@leemhuis.info> <183d138b0701030210g7127c6a0s10669ce333231cd6@mail.gmail.com> <1167834163.4869.2.camel@localhost.localdomain> <2cb10c440701030900g379d3fcdw5b14bfd21925b2a1@mail.gmail.com> <459BE0F7.7050401@fedoraproject.org> Message-ID: <2cb10c440701030913j635d455ajef2c065b2a791295@mail.gmail.com> On 1/3/07, Rahul Sundaram wrote: > Luis Villa wrote: > > On 1/3/07, Michael Tiemann wrote: > >> On Wed, 2007-01-03 at 21:10 +1100, Rob Garth wrote: > >> > I am not part of the Fedora Project and so I guees my opinion is > >> > fairly meaningless but on the LTS support stuff. FC6 is the upstream > >> > of RHEL5. > >> > >> ...and so much more. It is not *just* the upstream of Red Hat's > >> commercial products. > > > > Sure, but that is still the primary motivation for most of the hours > > and money that goes into Fedora. To pretend that that is just one goal > > among many makes Fedora look dishonest. To be sure, it is admirable > > (and I think admired) to say that the long-term goal is for RHEL to be > > just one consumer among many. But that isn't the reality on the ground > > right now, and to deny that (as many Fedora-affiliated people have > > tried to do from time to time) is silly. > > It's not. Think OLPC for example. There is also stuff out there like > Sony play stations, Cell processors etc. It already is being used as a > base for variety of stuff out there. This is exactly the kind of claim that I'm talking about- it makes Fedora look delusionally optimistic at best and deceptive at worst. Every time Fedora people say things like this, the rest of the world *thinks less of Fedora.* Compare and contrast: RH is probably the kernel's biggest contributor (maybe IBM), but if either RH or IBM went away tomorrow, there would still be a ton of paid kernel developers and a fair number of very deeply skilled volunteers. If RH went away tomorrow, how many paid Fedora developers would there be? How many volunteers would be able to spin a new release? That is the difference between where you are, where you sometimes claim to be, and where you want to be some day. Do OLPC and various Cell-based things consume and even contribute to Fedora? Absolutely. If RHEL went away tomorrow, could those projects support Fedora? No way in hell, as far as I can see as an outsider. Pretending otherwise is silly. Admit that RHEL is the driving force behind Fedora right now, state that it is a goal that RHEL be just one of many, and work in that direction. But don't pretend it is otherwise right now- it makes you look bad. Luis From mmcgrath at fedoraproject.org Wed Jan 3 17:18:28 2007 From: mmcgrath at fedoraproject.org (Mike McGrath) Date: Wed, 3 Jan 2007 11:18:28 -0600 Subject: why I'm using Ubuntu instead of Fedora ATM In-Reply-To: <2cb10c440701030913j635d455ajef2c065b2a791295@mail.gmail.com> References: <2cb10c440701030059l49323ecaqeae7c7041f735d63@mail.gmail.com> <459B7B49.8040101@leemhuis.info> <183d138b0701030210g7127c6a0s10669ce333231cd6@mail.gmail.com> <1167834163.4869.2.camel@localhost.localdomain> <2cb10c440701030900g379d3fcdw5b14bfd21925b2a1@mail.gmail.com> <459BE0F7.7050401@fedoraproject.org> <2cb10c440701030913j635d455ajef2c065b2a791295@mail.gmail.com> Message-ID: <3237e4410701030918p7ef9d3afr6b939b3c86042b39@mail.gmail.com> On 1/3/07, Luis Villa wrote: > On 1/3/07, Rahul Sundaram wrote: > > > > This is exactly the kind of claim that I'm talking about- it makes > Fedora look delusionally optimistic at best and deceptive at worst. > Every time Fedora people say things like this, the rest of the world > *thinks less of Fedora.* > I don't. -Mike From fedora at leemhuis.info Wed Jan 3 17:19:02 2007 From: fedora at leemhuis.info (Thorsten Leemhuis) Date: Wed, 03 Jan 2007 18:19:02 +0100 Subject: why I'm using Ubuntu instead of Fedora ATM In-Reply-To: <2cb10c440701030858g4a0c84cdi3cc1f67186b3258f@mail.gmail.com> References: <2cb10c440701030059l49323ecaqeae7c7041f735d63@mail.gmail.com> <459B7B49.8040101@leemhuis.info> <2cb10c440701030858g4a0c84cdi3cc1f67186b3258f@mail.gmail.com> Message-ID: <459BE586.9080509@leemhuis.info> Luis Villa schrieb: > On 1/3/07, Thorsten Leemhuis wrote: >> Thx Luis, I liked you mail. Some comments from my side: >> On 03.01.2007 09:59, Luis Villa wrote: >>> * QA: Ubuntu aggressively pushes people to use their development >>> branch and report problems, which leads to better, more stable final >>> releases. [...] >> Agreed. The biggest problem we have in this regard IMHO is that we >> always communicate "you can't get to stable release from rawhide or a >> test release". That scares people aways from the devel branch and the >> test releases. We should provide a clean solution so people at least can >> get from test3 to stable. > No, you should provide a clean path from every package ver X to every > package ver X+1. Period. [...] Actually I'd like to see this, too, but Fedora people seem to be so unwilling to support something like that so I tried the "better ask for something that's not that hard to realize and lies between the current solution and the best/ideal solution; maybe then the chances to get what we ask for are higher and it make one big part of the problem go away, even if parts of the problem remain"-approach ;-) CU thl From davej at redhat.com Wed Jan 3 17:26:12 2007 From: davej at redhat.com (Dave Jones) Date: Wed, 3 Jan 2007 12:26:12 -0500 Subject: why I'm using Ubuntu instead of Fedora ATM In-Reply-To: <2cb10c440701030858g4a0c84cdi3cc1f67186b3258f@mail.gmail.com> References: <2cb10c440701030059l49323ecaqeae7c7041f735d63@mail.gmail.com> <459B7B49.8040101@leemhuis.info> <2cb10c440701030858g4a0c84cdi3cc1f67186b3258f@mail.gmail.com> Message-ID: <20070103172612.GC26533@redhat.com> On Wed, Jan 03, 2007 at 11:58:05AM -0500, Luis Villa wrote: > I don't think the need (at least my need) is for a Fedora LTS; I > merely ask for one *stable* release at a time that: > > (1) doesn't get new features/new upstream releases Right now, rebases to newer kernel releases is the single provider of the majority of bugfixes we get reported. If we stopped doing that, we'd pretty much be giving up all hope of fixing kernel bugs. It's already completely out of control (right now ~1000 open bugs), and with the limited resources we have to attack this problem, staying close to upstream so that we can get upstream involvement in fixing bugs is our only hope. Yes, occasionally there are regressions, but tbh these days this happens to a much lesser extent. As an avid bugzilla fan, you might be interested to see the kernel bug count over time - http://people.redhat.com/davej/bugzilla-stats.txt For the last six months, we've made pretty much zero overall progress in reducing the overall count, despite hundreds of bugs being closed. By taking away the ability to move to a new upstream, the number of unfixed bugs will skyrocket. We already get a bad rap from some users who claim the process is = file bug = bug sits there = reaches end of life = closed->nextrelease. This will happen more and more if we stagnate on single versions. Backporting fixes is a *ton* of work. We have a huge team of people who do this exclusively for RHEL, and they don't even get to cover everything, just the bugs deemed important. Fedora gets a *lot* more bugs filed, and has a lot less manpower. It's just a losing proposition from every angle from where I'm sitting. Dave -- http://www.codemonkey.org.uk From luis at tieguy.org Wed Jan 3 17:21:29 2007 From: luis at tieguy.org (Luis Villa) Date: Wed, 3 Jan 2007 12:21:29 -0500 Subject: why I'm using Ubuntu instead of Fedora ATM In-Reply-To: <459BE586.9080509@leemhuis.info> References: <2cb10c440701030059l49323ecaqeae7c7041f735d63@mail.gmail.com> <459B7B49.8040101@leemhuis.info> <2cb10c440701030858g4a0c84cdi3cc1f67186b3258f@mail.gmail.com> <459BE586.9080509@leemhuis.info> Message-ID: <2cb10c440701030921t3310c6f0kf768eaaddf39dc5d@mail.gmail.com> On 1/3/07, Thorsten Leemhuis wrote: > Luis Villa schrieb: > > On 1/3/07, Thorsten Leemhuis wrote: > >> Thx Luis, I liked you mail. Some comments from my side: > >> On 03.01.2007 09:59, Luis Villa wrote: > >>> * QA: Ubuntu aggressively pushes people to use their development > >>> branch and report problems, which leads to better, more stable final > >>> releases. [...] > >> Agreed. The biggest problem we have in this regard IMHO is that we > >> always communicate "you can't get to stable release from rawhide or a > >> test release". That scares people aways from the devel branch and the > >> test releases. We should provide a clean solution so people at least can > >> get from test3 to stable. > > No, you should provide a clean path from every package ver X to every > > package ver X+1. Period. [...] > > Actually I'd like to see this, too, but Fedora people seem to be so > unwilling to support something like that so I tried the "better ask for > something that's not that hard to realize and lies between the current > solution and the best/ideal solution; maybe then the chances to get what > we ask for are higher and it make one big part of the problem go away, > even if parts of the problem remain"-approach ;-) I'm not (at least in this thread) trying to explain what Fedora *can* do; I'm trying to state what others *are* doing and where Fedora is falling short. Fedora may feel they can't do it for whatever reason, but others are doing it (or at least a reasonable fascimile of it), and seeing benefits from it, and Fedora should be aware of that. Luis From jwboyer at jdub.homelinux.org Wed Jan 3 17:30:08 2007 From: jwboyer at jdub.homelinux.org (Josh Boyer) Date: Wed, 03 Jan 2007 11:30:08 -0600 Subject: why I'm using Ubuntu instead of Fedora ATM In-Reply-To: <3237e4410701030918p7ef9d3afr6b939b3c86042b39@mail.gmail.com> References: <2cb10c440701030059l49323ecaqeae7c7041f735d63@mail.gmail.com> <459B7B49.8040101@leemhuis.info> <183d138b0701030210g7127c6a0s10669ce333231cd6@mail.gmail.com> <1167834163.4869.2.camel@localhost.localdomain> <2cb10c440701030900g379d3fcdw5b14bfd21925b2a1@mail.gmail.com> <459BE0F7.7050401@fedoraproject.org> <2cb10c440701030913j635d455ajef2c065b2a791295@mail.gmail.com> <3237e4410701030918p7ef9d3afr6b939b3c86042b39@mail.gmail.com> Message-ID: <1167845408.3836.28.camel@zod.rchland.ibm.com> On Wed, 2007-01-03 at 11:18 -0600, Mike McGrath wrote: > On 1/3/07, Luis Villa wrote: > > On 1/3/07, Rahul Sundaram wrote: > > > > > > > > This is exactly the kind of claim that I'm talking about- it makes > > Fedora look delusionally optimistic at best and deceptive at worst. > > Every time Fedora people say things like this, the rest of the world > > *thinks less of Fedora.* > > > > I don't. You work for Red Hat ;). To be fair, Luis has a point. OLPC semi-counts. It's another Red Hat driven (and funded) project, however with a much more philanthropic goal. If RHEL (and Red Hat) were to disappear today, I believe Fedora would live on. But I believe it would live on as a shadow of it's current state today for quite some time. Which leads me to say, as a Fedora user,: Thanks Red Hat! josh From luis at tieguy.org Wed Jan 3 17:31:23 2007 From: luis at tieguy.org (Luis Villa) Date: Wed, 3 Jan 2007 12:31:23 -0500 Subject: why I'm using Ubuntu instead of Fedora ATM In-Reply-To: <20070103172612.GC26533@redhat.com> References: <2cb10c440701030059l49323ecaqeae7c7041f735d63@mail.gmail.com> <459B7B49.8040101@leemhuis.info> <2cb10c440701030858g4a0c84cdi3cc1f67186b3258f@mail.gmail.com> <20070103172612.GC26533@redhat.com> Message-ID: <2cb10c440701030931u4952585ev8909363d47e4ff31@mail.gmail.com> On 1/3/07, Dave Jones wrote: > On Wed, Jan 03, 2007 at 11:58:05AM -0500, Luis Villa wrote: > > > I don't think the need (at least my need) is for a Fedora LTS; I > > merely ask for one *stable* release at a time that: > > > > (1) doesn't get new features/new upstream releases > > Right now, rebases to newer kernel releases is the single provider > of the majority of bugfixes we get reported. If we stopped doing > that, we'd pretty much be giving up all hope of fixing kernel bugs. > > It's already completely out of control (right now ~1000 open bugs), > and with the limited resources we have to attack this problem, > staying close to upstream so that we can get upstream involvement > in fixing bugs is our only hope. > > Yes, occasionally there are regressions, but tbh these days this > happens to a much lesser extent. > As an avid bugzilla fan, you might be interested to see > the kernel bug count over time - http://people.redhat.com/davej/bugzilla-stats.txt > For the last six months, we've made pretty much zero overall progress > in reducing the overall count, despite hundreds of bugs being closed. > > By taking away the ability to move to a new upstream, the number > of unfixed bugs will skyrocket. We already get a bad rap from some > users who claim the process is > > = file bug > = bug sits there > = reaches end of life > = closed->nextrelease. > > This will happen more and more if we stagnate on single versions. > > Backporting fixes is a *ton* of work. We have a huge team of > people who do this exclusively for RHEL, and they don't even > get to cover everything, just the bugs deemed important. > Fedora gets a *lot* more bugs filed, and has a lot less manpower. > It's just a losing proposition from every angle from where I'm sitting. Kernel, and for slightly different reasons GNOME, are likely exceptions to the rule. (In fact, Ubuntu explicitly exempts both from their no-new-upstream freeze.) Both have extremely active upstream development which includes pretty good QA processes. (Kernel more upstream development and less QA, GNOME less upstream development and better organized QA.) I'm sure there are some other exceptions, and the bar for making new exceptions can be lowered if Fedora has its own pre-release QA mechanism like the testing channel I mentioned. Luis From jkeating at redhat.com Wed Jan 3 17:31:09 2007 From: jkeating at redhat.com (Jesse Keating) Date: Wed, 3 Jan 2007 12:31:09 -0500 Subject: why I'm using Ubuntu instead of Fedora ATM In-Reply-To: <459BE586.9080509@leemhuis.info> References: <2cb10c440701030059l49323ecaqeae7c7041f735d63@mail.gmail.com> <2cb10c440701030858g4a0c84cdi3cc1f67186b3258f@mail.gmail.com> <459BE586.9080509@leemhuis.info> Message-ID: <200701031231.09906.jkeating@redhat.com> On Wednesday 03 January 2007 12:19, Thorsten Leemhuis wrote: > Actually I'd like to see this, too, but Fedora people seem to be so > unwilling to support something like that so I tried the "better ask for > something that's not that hard to realize and lies between the current > solution and the best/ideal solution; maybe then the chances to get what > we ask for are higher and it make one big part of the problem go away, > even if parts of the problem remain"-approach Well, we could do this, but it would involve an explosion of epochs :/ Our ability to roll back package versions that were 'mistakes' is very very useful. The only way to keep this is to use epochs to "force" a lower version package to win. Once epoch, always epoch. -- Jesse Keating Release Engineer: Fedora -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 189 bytes Desc: not available URL: From sundaram at fedoraproject.org Wed Jan 3 17:44:59 2007 From: sundaram at fedoraproject.org (Rahul Sundaram) Date: Wed, 03 Jan 2007 23:14:59 +0530 Subject: why I'm using Ubuntu instead of Fedora ATM In-Reply-To: <2cb10c440701030913j635d455ajef2c065b2a791295@mail.gmail.com> References: <2cb10c440701030059l49323ecaqeae7c7041f735d63@mail.gmail.com> <459B7B49.8040101@leemhuis.info> <183d138b0701030210g7127c6a0s10669ce333231cd6@mail.gmail.com> <1167834163.4869.2.camel@localhost.localdomain> <2cb10c440701030900g379d3fcdw5b14bfd21925b2a1@mail.gmail.com> <459BE0F7.7050401@fedoraproject.org> <2cb10c440701030913j635d455ajef2c065b2a791295@mail.gmail.com> Message-ID: <459BEB9B.8030100@fedoraproject.org> Luis Villa wrote: > Do OLPC and various Cell-based things consume and even contribute to > Fedora? Absolutely. If RHEL went away tomorrow, could those projects > support Fedora? No way in hell, as far as I can see as an outsider. > Pretending otherwise is silly. Admit that RHEL is the driving force > behind Fedora right now, state that it is a goal that RHEL be just one > of many, and work in that direction. But don't pretend it is > otherwise right now- it makes you look bad. We are talking about two different things here as far as I can see * What is the primary driving force of Fedora development today? RHEL * Who are the consumers? Many. People frequently see RHEL as the only consumer and then argue that it is ok for Fedora to be driven based on only that requirements which is the wrong approach IMO> Rahul From sundaram at fedoraproject.org Wed Jan 3 17:48:05 2007 From: sundaram at fedoraproject.org (Rahul Sundaram) Date: Wed, 03 Jan 2007 23:18:05 +0530 Subject: why I'm using Ubuntu instead of Fedora ATM In-Reply-To: <2cb10c440701030931u4952585ev8909363d47e4ff31@mail.gmail.com> References: <2cb10c440701030059l49323ecaqeae7c7041f735d63@mail.gmail.com> <459B7B49.8040101@leemhuis.info> <2cb10c440701030858g4a0c84cdi3cc1f67186b3258f@mail.gmail.com> <20070103172612.GC26533@redhat.com> <2cb10c440701030931u4952585ev8909363d47e4ff31@mail.gmail.com> Message-ID: <459BEC55.8050702@fedoraproject.org> Luis Villa wrote: I'm sure there are some other exceptions, and > the bar for making new exceptions can be lowered if Fedora has its own > pre-release QA mechanism like the testing channel I mentioned. > What do you mean by "testing channel"? Fedora already has updates-testing repository. Are we talking about that? Rahul From max at spevack.org Wed Jan 3 17:52:20 2007 From: max at spevack.org (Max Spevack) Date: Wed, 3 Jan 2007 12:52:20 -0500 (EST) Subject: why I'm using Ubuntu instead of Fedora ATM In-Reply-To: <2cb10c440701030913j635d455ajef2c065b2a791295@mail.gmail.com> References: <2cb10c440701030059l49323ecaqeae7c7041f735d63@mail.gmail.com> <459B7B49.8040101@leemhuis.info> <183d138b0701030210g7127c6a0s10669ce333231cd6@mail.gmail.com> <1167834163.4869.2.camel@localhost.localdomain> <2cb10c440701030900g379d3fcdw5b14bfd21925b2a1@mail.gmail.com> <459BE0F7.7050401@fedoraproject.org> <2cb10c440701030913j635d455ajef2c065b2a791295@mail.gmail.com> Message-ID: On Wed, 3 Jan 2007, Luis Villa wrote: > Admit that RHEL is the driving force behind Fedora right now, state that > it is a goal that RHEL be just one of many, and work in that direction. > But don't pretend it is otherwise right now- it makes you look bad. I don't think we've been shy in saying that RHEL is the primary consumer of the upstream that Fedora provides. But at the same time, the entire point of Fedora 7 is to detach the RHEL and Fedora in a way that makes community (non-RH) ownership and contribution to Fedora easier and greater than it ever has been. The challenge is then on RHEL to adapt itself to take full advantage of what is going on in Fedora. I think that the Fedora messaging that I see coming from the Board-level types, as well as myself, is very honest about the realities of the Fedora-RHEL relationship. I try very hard to make sure that when we talk about Fedora, we are the least BS or corporate-speak people around. But I prefer to acknowledge those things and then talk about how we want it to change. A year ago at this time the "how we want it to change" part was all vaporware. Now it's a partial reality, more each day. Look at pungi, the livecd, the build system/fedora 7 goals. So I think we're on the right track. --Max From davej at redhat.com Wed Jan 3 17:53:52 2007 From: davej at redhat.com (Dave Jones) Date: Wed, 3 Jan 2007 12:53:52 -0500 Subject: why I'm using Ubuntu instead of Fedora ATM In-Reply-To: <459BEB9B.8030100@fedoraproject.org> References: <2cb10c440701030059l49323ecaqeae7c7041f735d63@mail.gmail.com> <459B7B49.8040101@leemhuis.info> <183d138b0701030210g7127c6a0s10669ce333231cd6@mail.gmail.com> <1167834163.4869.2.camel@localhost.localdomain> <2cb10c440701030900g379d3fcdw5b14bfd21925b2a1@mail.gmail.com> <459BE0F7.7050401@fedoraproject.org> <2cb10c440701030913j635d455ajef2c065b2a791295@mail.gmail.com> <459BEB9B.8030100@fedoraproject.org> Message-ID: <20070103175352.GD26533@redhat.com> On Wed, Jan 03, 2007 at 11:14:59PM +0530, Rahul Sundaram wrote: > * What is the primary driving force of Fedora development today? > RHEL In some ways I think this is over exaggerated. A lot of the really cool stuff tha that happens in Fedora has had zero input from RHEL-focused folks in Red Hat. To use one example: Suspend support was something me and Jeremy thought would be cool to hack on, so we just did it. This sort of thing happens a lot more often than Fedora gets credit for. Yes, sometimes mandated things like Xen come down the pipe, but its really the minority case. For the most part, it's less of a "we want xyz in rhel x+1" and more "what's happened upstream/in fedora for the next rhel?" For RHEL5, Xen is pretty much the only RHEL driven feature that ended up in Fedora. Dave -- http://www.codemonkey.org.uk From notting at redhat.com Wed Jan 3 17:54:53 2007 From: notting at redhat.com (Bill Nottingham) Date: Wed, 3 Jan 2007 12:54:53 -0500 Subject: why I'm using Ubuntu instead of Fedora ATM In-Reply-To: <20070103175352.GD26533@redhat.com> References: <2cb10c440701030059l49323ecaqeae7c7041f735d63@mail.gmail.com> <459B7B49.8040101@leemhuis.info> <183d138b0701030210g7127c6a0s10669ce333231cd6@mail.gmail.com> <1167834163.4869.2.camel@localhost.localdomain> <2cb10c440701030900g379d3fcdw5b14bfd21925b2a1@mail.gmail.com> <459BE0F7.7050401@fedoraproject.org> <2cb10c440701030913j635d455ajef2c065b2a791295@mail.gmail.com> <459BEB9B.8030100@fedoraproject.org> <20070103175352.GD26533@redhat.com> Message-ID: <20070103175453.GC23011@nostromo.devel.redhat.com> Dave Jones (davej at redhat.com) said: > > * What is the primary driving force of Fedora development today? > > RHEL > > In some ways I think this is over exaggerated. > A lot of the really cool stuff tha that happens in Fedora > has had zero input from RHEL-focused folks in Red Hat. True, but if you're looking at a budgetary force behind the engineers working.... Bill From davej at redhat.com Wed Jan 3 17:57:18 2007 From: davej at redhat.com (Dave Jones) Date: Wed, 3 Jan 2007 12:57:18 -0500 Subject: why I'm using Ubuntu instead of Fedora ATM In-Reply-To: <2cb10c440701030931u4952585ev8909363d47e4ff31@mail.gmail.com> References: <2cb10c440701030059l49323ecaqeae7c7041f735d63@mail.gmail.com> <459B7B49.8040101@leemhuis.info> <2cb10c440701030858g4a0c84cdi3cc1f67186b3258f@mail.gmail.com> <20070103172612.GC26533@redhat.com> <2cb10c440701030931u4952585ev8909363d47e4ff31@mail.gmail.com> Message-ID: <20070103175718.GA30432@redhat.com> On Wed, Jan 03, 2007 at 12:31:23PM -0500, Luis Villa wrote: > Kernel, and for slightly different reasons GNOME, are likely > exceptions to the rule. (In fact, Ubuntu explicitly exempts both from > their no-new-upstream freeze.) Both have extremely active upstream > development which includes pretty good QA processes. (Kernel more > upstream development and less QA, GNOME less upstream development and > better organized QA.) I'm sure there are some other exceptions, and > the bar for making new exceptions can be lowered if Fedora has its own > pre-release QA mechanism like the testing channel I mentioned. The question then becomes where do you draw the line? 'Gnome' covers a pretty large package set, and I'll wager that 'gnome + kernel' probably covers the majority of the bugs that get filed. Dave -- http://www.codemonkey.org.uk From max at spevack.org Wed Jan 3 17:59:03 2007 From: max at spevack.org (Max Spevack) Date: Wed, 3 Jan 2007 12:59:03 -0500 (EST) Subject: why I'm using Ubuntu instead of Fedora ATM In-Reply-To: <1167845408.3836.28.camel@zod.rchland.ibm.com> References: <2cb10c440701030059l49323ecaqeae7c7041f735d63@mail.gmail.com> <459B7B49.8040101@leemhuis.info> <183d138b0701030210g7127c6a0s10669ce333231cd6@mail.gmail.com> <1167834163.4869.2.camel@localhost.localdomain> <2cb10c440701030900g379d3fcdw5b14bfd21925b2a1@mail.gmail.com> <459BE0F7.7050401@fedoraproject.org> <2cb10c440701030913j635d455ajef2c065b2a791295@mail.gmail.com> <3237e4410701030918p7ef9d3afr6b939b3c86042b39@mail.gmail.com> <1167845408.3836.28.camel@zod.rchland.ibm.com> Message-ID: On Wed, 3 Jan 2007, Josh Boyer wrote: > To be fair, Luis has a point. OLPC semi-counts. It's another Red Hat > driven (and funded) project, however with a much more philanthropic > goal. If RHEL (and Red Hat) were to disappear today, I believe Fedora > would live on. But I believe it would live on as a shadow of it's > current state today for quite some time. As Elliot Lee once said, "my goal at work is to make myself unnecessary". If Red Hat disappeared tomorrow, would Fedora be in better shape than if Red Hat had disappeared a year ago? Yes. As long as that statement is still true a year from now, we're doing the right things. Part of my job (and you guys are the ones who tell me when I'm doing it well and when I'm dropping the ball) is to make sure that the resources Red Hat *does* give to Fedora are in the best interests of Fedora as a community project. And all the credit in the world here to Jeremy Katz for his tireless work in selling the Fedora 7 vision (which is all about using Red Hat resources to make Fedora both better and more indepenent) to lots of folks within Red Hat. -- Max Spevack + http://spevack.org + gpg key -- http://spevack.org/max.asc + fingerprint -- CD52 5E72 369B B00D 9E9A 773E 2FDB CB46 5A17 CF21 From davej at redhat.com Wed Jan 3 18:02:00 2007 From: davej at redhat.com (Dave Jones) Date: Wed, 3 Jan 2007 13:02:00 -0500 Subject: why I'm using Ubuntu instead of Fedora ATM In-Reply-To: <20070103175453.GC23011@nostromo.devel.redhat.com> References: <2cb10c440701030059l49323ecaqeae7c7041f735d63@mail.gmail.com> <459B7B49.8040101@leemhuis.info> <183d138b0701030210g7127c6a0s10669ce333231cd6@mail.gmail.com> <1167834163.4869.2.camel@localhost.localdomain> <2cb10c440701030900g379d3fcdw5b14bfd21925b2a1@mail.gmail.com> <459BE0F7.7050401@fedoraproject.org> <2cb10c440701030913j635d455ajef2c065b2a791295@mail.gmail.com> <459BEB9B.8030100@fedoraproject.org> <20070103175352.GD26533@redhat.com> <20070103175453.GC23011@nostromo.devel.redhat.com> Message-ID: <20070103180200.GA31104@redhat.com> On Wed, Jan 03, 2007 at 12:54:53PM -0500, Bill Nottingham wrote: > Dave Jones (davej at redhat.com) said: > > > * What is the primary driving force of Fedora development today? > > > RHEL > > > > In some ways I think this is over exaggerated. > > A lot of the really cool stuff tha that happens in Fedora > > has had zero input from RHEL-focused folks in Red Hat. > > True, but if you're looking at a budgetary force behind the > engineers working.... Where the money comes from != "driving force of development". Fedora is where innovation is happening, just like it says on the tin. To say that RHEL is driving Fedora is for the most part, just plain wrong. Dave -- http://www.codemonkey.org.uk From fedora at leemhuis.info Wed Jan 3 18:02:44 2007 From: fedora at leemhuis.info (Thorsten Leemhuis) Date: Wed, 03 Jan 2007 19:02:44 +0100 Subject: why I'm using Ubuntu instead of Fedora ATM In-Reply-To: <200701031231.09906.jkeating@redhat.com> References: <2cb10c440701030059l49323ecaqeae7c7041f735d63@mail.gmail.com> <2cb10c440701030858g4a0c84cdi3cc1f67186b3258f@mail.gmail.com> <459BE586.9080509@leemhuis.info> <200701031231.09906.jkeating@redhat.com> Message-ID: <459BEFC4.4020909@leemhuis.info> Jesse Keating schrieb: > On Wednesday 03 January 2007 12:19, Thorsten Leemhuis wrote: >> Actually I'd like to see this, too, but Fedora people seem to be so >> unwilling to support something like that so I tried the "better ask for >> something that's not that hard to realize and lies between the current >> solution and the best/ideal solution; maybe then the chances to get what >> we ask for are higher and it make one big part of the problem go away, >> even if parts of the problem remain"-approach > Well, we could do this, but it would involve an explosion of epochs :/ Our > ability to roll back package versions that were 'mistakes' is very very > useful. [...] Sorry, but we IMHO did something seriously wrong beforehand already if we have to roll back packages *after* test3. Cu thl From luis at tieguy.org Wed Jan 3 18:03:39 2007 From: luis at tieguy.org (Luis Villa) Date: Wed, 3 Jan 2007 13:03:39 -0500 Subject: why I'm using Ubuntu instead of Fedora ATM In-Reply-To: <20070103175718.GA30432@redhat.com> References: <2cb10c440701030059l49323ecaqeae7c7041f735d63@mail.gmail.com> <459B7B49.8040101@leemhuis.info> <2cb10c440701030858g4a0c84cdi3cc1f67186b3258f@mail.gmail.com> <20070103172612.GC26533@redhat.com> <2cb10c440701030931u4952585ev8909363d47e4ff31@mail.gmail.com> <20070103175718.GA30432@redhat.com> Message-ID: <2cb10c440701031003q68d272ffs70470e4bc238d88e@mail.gmail.com> On 1/3/07, Dave Jones wrote: > On Wed, Jan 03, 2007 at 12:31:23PM -0500, Luis Villa wrote: > > > Kernel, and for slightly different reasons GNOME, are likely > > exceptions to the rule. (In fact, Ubuntu explicitly exempts both from > > their no-new-upstream freeze.) Both have extremely active upstream > > development which includes pretty good QA processes. (Kernel more > > upstream development and less QA, GNOME less upstream development and > > better organized QA.) I'm sure there are some other exceptions, and > > the bar for making new exceptions can be lowered if Fedora has its own > > pre-release QA mechanism like the testing channel I mentioned. > > The question then becomes where do you draw the line? > 'Gnome' covers a pretty large package set, and I'll wager that > 'gnome + kernel' probably covers the majority of the bugs that get filed. FWIW, Ubuntu's line around GNOME is 'core GNOME release set', which is the only thing GNOME makes a serious attempt to QA. But no, I don't have a great answer to your broader line-drawing question- I can only say that I'm pretty sure that it is more narrow than Fedora is currently drawing it, and Ubuntu's line appears to have worked pretty well for them. Luis From dwmw2 at infradead.org Wed Jan 3 18:05:05 2007 From: dwmw2 at infradead.org (David Woodhouse) Date: Wed, 03 Jan 2007 18:05:05 +0000 Subject: why I'm using Ubuntu instead of Fedora ATM In-Reply-To: <20070103175352.GD26533@redhat.com> References: <2cb10c440701030059l49323ecaqeae7c7041f735d63@mail.gmail.com> <459B7B49.8040101@leemhuis.info> <183d138b0701030210g7127c6a0s10669ce333231cd6@mail.gmail.com> <1167834163.4869.2.camel@localhost.localdomain> <2cb10c440701030900g379d3fcdw5b14bfd21925b2a1@mail.gmail.com> <459BE0F7.7050401@fedoraproject.org> <2cb10c440701030913j635d455ajef2c065b2a791295@mail.gmail.com> <459BEB9B.8030100@fedoraproject.org> <20070103175352.GD26533@redhat.com> Message-ID: <1167847505.22068.482.camel@pmac.infradead.org> On Wed, 2007-01-03 at 12:53 -0500, Dave Jones wrote: > For RHEL5, Xen is pretty much the only RHEL driven feature > that ended up in Fedora. Well, that and the bloody stupid 64KiB pages on ppc64, which were supposed to have been _only_ in RHEL and not Fedora. -- dwmw2 From gdk at redhat.com Wed Jan 3 18:02:39 2007 From: gdk at redhat.com (Greg Dekoenigsberg) Date: Wed, 3 Jan 2007 13:02:39 -0500 (EST) Subject: why I'm using Ubuntu instead of Fedora ATM In-Reply-To: <20070103170805.GA23011@nostromo.devel.redhat.com> References: <2cb10c440701030059l49323ecaqeae7c7041f735d63@mail.gmail.com> <459B7B49.8040101@leemhuis.info> <20070103163903.GA22376@nostromo.devel.redhat.com> <20070103170805.GA23011@nostromo.devel.redhat.com> Message-ID: On Wed, 3 Jan 2007, Bill Nottingham wrote: >>>> I'm wondering if we could provide solutions for both users: One update >>>> channel that only gets security updates and important bugfixes while the >>>> other is a bit more bold -- we for example could have firefox2 in the >>>> bold channel for FC6 while shipping the latest firefox 1.5.x in the more >>>> conservative channel. >>> >>> So, an idea like this: >>> >>> - starts to exponentially expand the QA problem >> >> Bill, I love you, but I'm going to have to call bullshit on this one. >> The answer is not "avoid making QA harder", the answer is to SOLVE THE QA >> PROBLEM. > > Sure, but which of these plans make more sense: > > 1) > > - Solve the QA problem for our repo configurations as they exist > - Expand the QA solution to new, multiple, disparate and conflicting > repositories > > 2) > > - Expand into new, multiple, disparate and conflicting repositories > - Then try to solve the QA problem > > Honestly, before we can do multiple experimental repositories of this, > that, and the other, we need to get our *OWN* house in order. I agree with this completely -- if, and only if, we actually make a concerted effort at fixing the QA problem. Which means articulating the QA problem, actually. --g ------------------------------------------------------------- Greg DeKoenigsberg || Fedora Project || fedoraproject.org Be an Ambassador || http://fedoraproject.org/wiki/Ambassadors ------------------------------------------------------------- From jwboyer at jdub.homelinux.org Wed Jan 3 18:06:15 2007 From: jwboyer at jdub.homelinux.org (Josh Boyer) Date: Wed, 03 Jan 2007 12:06:15 -0600 Subject: why I'm using Ubuntu instead of Fedora ATM In-Reply-To: References: <2cb10c440701030059l49323ecaqeae7c7041f735d63@mail.gmail.com> <459B7B49.8040101@leemhuis.info> <183d138b0701030210g7127c6a0s10669ce333231cd6@mail.gmail.com> <1167834163.4869.2.camel@localhost.localdomain> <2cb10c440701030900g379d3fcdw5b14bfd21925b2a1@mail.gmail.com> <459BE0F7.7050401@fedoraproject.org> <2cb10c440701030913j635d455ajef2c065b2a791295@mail.gmail.com> <3237e4410701030918p7ef9d3afr6b939b3c86042b39@mail.gmail.com> <1167845408.3836.28.camel@zod.rchland.ibm.com> Message-ID: <1167847575.3836.34.camel@zod.rchland.ibm.com> On Wed, 2007-01-03 at 12:59 -0500, Max Spevack wrote: > On Wed, 3 Jan 2007, Josh Boyer wrote: > > > To be fair, Luis has a point. OLPC semi-counts. It's another Red Hat > > driven (and funded) project, however with a much more philanthropic > > goal. If RHEL (and Red Hat) were to disappear today, I believe Fedora > > would live on. But I believe it would live on as a shadow of it's > > current state today for quite some time. > > As Elliot Lee once said, "my goal at work is to make myself unnecessary". > > If Red Hat disappeared tomorrow, would Fedora be in better shape than if > Red Hat had disappeared a year ago? Yes. > > As long as that statement is still true a year from now, we're doing the > right things. Sure, and I'm not saying Fedora isn't heading in the right direction. I think it truly is. All I was saying is that without dedicated, paid people work on Fedora, it would move at a much slower pace for some time. But it would still evolve. And that is most definitely because of work that's been done over the life of the project. > Part of my job (and you guys are the ones who tell me when I'm doing it > well and when I'm dropping the ball) is to make sure that the resources > Red Hat *does* give to Fedora are in the best interests of Fedora as a > community project. And all the credit in the world here to Jeremy Katz > for his tireless work in selling the Fedora 7 vision (which is all about > using Red Hat resources to make Fedora both better and more indepenent) to > lots of folks within Red Hat. Of course. Fedora 7 is by far the largest step I've seen so far and I can't wait for it to be realized. josh From luis at tieguy.org Wed Jan 3 18:06:57 2007 From: luis at tieguy.org (Luis Villa) Date: Wed, 3 Jan 2007 13:06:57 -0500 Subject: why I'm using Ubuntu instead of Fedora ATM In-Reply-To: References: <2cb10c440701030059l49323ecaqeae7c7041f735d63@mail.gmail.com> <459B7B49.8040101@leemhuis.info> <183d138b0701030210g7127c6a0s10669ce333231cd6@mail.gmail.com> <1167834163.4869.2.camel@localhost.localdomain> <2cb10c440701030900g379d3fcdw5b14bfd21925b2a1@mail.gmail.com> <459BE0F7.7050401@fedoraproject.org> <2cb10c440701030913j635d455ajef2c065b2a791295@mail.gmail.com> Message-ID: <2cb10c440701031006g16ec694ak1531d53237f99e1f@mail.gmail.com> On 1/3/07, Max Spevack wrote: > I think that the Fedora messaging that I see coming from the Board-level > types, as well as myself, is very honest about the realities of the > Fedora-RHEL relationship. I try very hard to make sure that when we talk > about Fedora, we are the least BS or corporate-speak people around. > > But I prefer to acknowledge those things and then talk about how we want > it to change. > > A year ago at this time the "how we want it to change" part was all > vaporware. Now it's a partial reality, more each day. Look at pungi, the > livecd, the build system/fedora 7 goals. So I think we're on the right > track. The direction Fedora is heading in is about as good as can possibly be imagined right now, as far as I can see. I can't speak more highly about what you guys have done and appear to be doing there. That is the only reason I'm even bothering to be on f-a-b. I'm not so positive about the messaging; not everyone is as careful as you are, Max. And even when it is careful, it seems like it isn't getting picked up by the conventional wisdom- I'm not sure I have good advice on that last part, I'm afraid. Luis From luis at tieguy.org Wed Jan 3 18:22:17 2007 From: luis at tieguy.org (Luis Villa) Date: Wed, 3 Jan 2007 13:22:17 -0500 Subject: why I'm using Ubuntu instead of Fedora ATM In-Reply-To: <459BEC55.8050702@fedoraproject.org> References: <2cb10c440701030059l49323ecaqeae7c7041f735d63@mail.gmail.com> <459B7B49.8040101@leemhuis.info> <2cb10c440701030858g4a0c84cdi3cc1f67186b3258f@mail.gmail.com> <20070103172612.GC26533@redhat.com> <2cb10c440701030931u4952585ev8909363d47e4ff31@mail.gmail.com> <459BEC55.8050702@fedoraproject.org> Message-ID: <2cb10c440701031022i1ecfb95ela17c9cf083bfbf64@mail.gmail.com> On 1/3/07, Rahul Sundaram wrote: > Luis Villa wrote: > I'm sure there are some other exceptions, and > > the bar for making new exceptions can be lowered if Fedora has its own > > pre-release QA mechanism like the testing channel I mentioned. > > > > What do you mean by "testing channel"? Fedora already has > updates-testing repository. Are we talking about that? Yeah, that's what I mean. It is underpublicized and underused, AFAICS- that I didn't know about it until a couple months ago, when I obsessively follow Linux QA, is not a good sign. Luis From notting at redhat.com Wed Jan 3 18:22:40 2007 From: notting at redhat.com (Bill Nottingham) Date: Wed, 3 Jan 2007 13:22:40 -0500 Subject: why I'm using Ubuntu instead of Fedora ATM In-Reply-To: <20070103180200.GA31104@redhat.com> References: <459B7B49.8040101@leemhuis.info> <183d138b0701030210g7127c6a0s10669ce333231cd6@mail.gmail.com> <1167834163.4869.2.camel@localhost.localdomain> <2cb10c440701030900g379d3fcdw5b14bfd21925b2a1@mail.gmail.com> <459BE0F7.7050401@fedoraproject.org> <2cb10c440701030913j635d455ajef2c065b2a791295@mail.gmail.com> <459BEB9B.8030100@fedoraproject.org> <20070103175352.GD26533@redhat.com> <20070103175453.GC23011@nostromo.devel.redhat.com> <20070103180200.GA31104@redhat.com> Message-ID: <20070103182240.GA24042@nostromo.devel.redhat.com> Dave Jones (davej at redhat.com) said: > > True, but if you're looking at a budgetary force behind the > > engineers working.... > > Where the money comes from != "driving force of development". > Fedora is where innovation is happening, just like it says > on the tin. To say that RHEL is driving Fedora is for > the most part, just plain wrong. RHEL? No. Red Hat? Yes. Fedora is not now self-sustaining without Red Hat, and I doubt it will be in the near future, even with the strides we're making. (Which isn't to say it can't be a goal) Bill From notting at redhat.com Wed Jan 3 18:23:59 2007 From: notting at redhat.com (Bill Nottingham) Date: Wed, 3 Jan 2007 13:23:59 -0500 Subject: why I'm using Ubuntu instead of Fedora ATM In-Reply-To: <20070103175718.GA30432@redhat.com> References: <2cb10c440701030059l49323ecaqeae7c7041f735d63@mail.gmail.com> <459B7B49.8040101@leemhuis.info> <2cb10c440701030858g4a0c84cdi3cc1f67186b3258f@mail.gmail.com> <20070103172612.GC26533@redhat.com> <2cb10c440701030931u4952585ev8909363d47e4ff31@mail.gmail.com> <20070103175718.GA30432@redhat.com> Message-ID: <20070103182359.GD23011@nostromo.devel.redhat.com> Dave Jones (davej at redhat.com) said: > The question then becomes where do you draw the line? > 'Gnome' covers a pretty large package set, and I'll wager that > 'gnome + kernel' probably covers the majority of the bugs that get filed. Total FC6 bugs filed: 2826. Kernel: 427 GNOME: 578 (includes: apps, OO.o, firefox and kin, as they're really 'desktop') So, 35%, give or take. Bill From jkeating at redhat.com Wed Jan 3 18:30:59 2007 From: jkeating at redhat.com (Jesse Keating) Date: Wed, 3 Jan 2007 13:30:59 -0500 Subject: why I'm using Ubuntu instead of Fedora ATM In-Reply-To: <459BEFC4.4020909@leemhuis.info> References: <2cb10c440701030059l49323ecaqeae7c7041f735d63@mail.gmail.com> <200701031231.09906.jkeating@redhat.com> <459BEFC4.4020909@leemhuis.info> Message-ID: <200701031331.02846.jkeating@redhat.com> On Wednesday 03 January 2007 13:02, Thorsten Leemhuis wrote: > Sorry, but we IMHO did something seriously wrong beforehand already if > we have to roll back packages *after* test3. Test3 isn't what Luis wants. He wants to be able to go from F7, to F8Test1 to F8Test2/3 to F8, cleanly and without problems. -- Jesse Keating Release Engineer: Fedora -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 189 bytes Desc: not available URL: From luis at tieguy.org Wed Jan 3 18:40:16 2007 From: luis at tieguy.org (Luis Villa) Date: Wed, 3 Jan 2007 13:40:16 -0500 Subject: why I'm using Ubuntu instead of Fedora ATM In-Reply-To: <20070103182359.GD23011@nostromo.devel.redhat.com> References: <2cb10c440701030059l49323ecaqeae7c7041f735d63@mail.gmail.com> <459B7B49.8040101@leemhuis.info> <2cb10c440701030858g4a0c84cdi3cc1f67186b3258f@mail.gmail.com> <20070103172612.GC26533@redhat.com> <2cb10c440701030931u4952585ev8909363d47e4ff31@mail.gmail.com> <20070103175718.GA30432@redhat.com> <20070103182359.GD23011@nostromo.devel.redhat.com> Message-ID: <2cb10c440701031040u6dffb838o1490301d0a7b472e@mail.gmail.com> On 1/3/07, Bill Nottingham wrote: > Dave Jones (davej at redhat.com) said: > > The question then becomes where do you draw the line? > > 'Gnome' covers a pretty large package set, and I'll wager that > > 'gnome + kernel' probably covers the majority of the bugs that get filed. > > Total FC6 bugs filed: 2826. > > Kernel: 427 > GNOME: 578 (includes: apps, OO.o, firefox and kin, as they're really > 'desktop') > > So, 35%, give or take. For the purposes of this line drawing you have to look at development organizations, not functional groupings, so GNOME != desktop. I'd be embarassed if GNOME's bug fixing and prioritization were as bad as OOo's, for example, or if GNOME's release cycle were as broken as firefoxes. Luis From luis at tieguy.org Wed Jan 3 18:43:26 2007 From: luis at tieguy.org (Luis Villa) Date: Wed, 3 Jan 2007 13:43:26 -0500 Subject: why I'm using Ubuntu instead of Fedora ATM In-Reply-To: <2cb10c440701031040u6dffb838o1490301d0a7b472e@mail.gmail.com> References: <2cb10c440701030059l49323ecaqeae7c7041f735d63@mail.gmail.com> <459B7B49.8040101@leemhuis.info> <2cb10c440701030858g4a0c84cdi3cc1f67186b3258f@mail.gmail.com> <20070103172612.GC26533@redhat.com> <2cb10c440701030931u4952585ev8909363d47e4ff31@mail.gmail.com> <20070103175718.GA30432@redhat.com> <20070103182359.GD23011@nostromo.devel.redhat.com> <2cb10c440701031040u6dffb838o1490301d0a7b472e@mail.gmail.com> Message-ID: <2cb10c440701031043m2830b7a5ra0a33702dcd195ad@mail.gmail.com> On 1/3/07, Luis Villa wrote: > On 1/3/07, Bill Nottingham wrote: > > Dave Jones (davej at redhat.com) said: > > > The question then becomes where do you draw the line? > > > 'Gnome' covers a pretty large package set, and I'll wager that > > > 'gnome + kernel' probably covers the majority of the bugs that get filed. > > > > Total FC6 bugs filed: 2826. > > > > Kernel: 427 > > GNOME: 578 (includes: apps, OO.o, firefox and kin, as they're really > > 'desktop') > > > > So, 35%, give or take. > > For the purposes of this line drawing you have to look at development > organizations, not functional groupings, so GNOME != desktop. I'd be > embarassed if GNOME's bug fixing and prioritization were as bad as > OOo's, for example, or if GNOME's release cycle were as broken as > firefoxes. (Not to say that GNOME is perfect, and in fact I believe it has been getting steadily worse over the past few years as Novell and Sun have withdrawn from using HEAD. But it does some things much better than OOo or ffox do.) Luis From dwmw2 at infradead.org Wed Jan 3 18:43:40 2007 From: dwmw2 at infradead.org (David Woodhouse) Date: Wed, 03 Jan 2007 18:43:40 +0000 Subject: why I'm using Ubuntu instead of Fedora ATM In-Reply-To: <20070103172612.GC26533@redhat.com> References: <2cb10c440701030059l49323ecaqeae7c7041f735d63@mail.gmail.com> <459B7B49.8040101@leemhuis.info> <2cb10c440701030858g4a0c84cdi3cc1f67186b3258f@mail.gmail.com> <20070103172612.GC26533@redhat.com> Message-ID: <1167849820.19205.10.camel@pmac.infradead.org> On Wed, 2007-01-03 at 12:26 -0500, Dave Jones wrote: > Right now, rebases to newer kernel releases is the single provider > of the majority of bugfixes we get reported. If we stopped doing > that, we'd pretty much be giving up all hope of fixing kernel bugs. Er, I think you mean "bug closures" not necessarily "fixes"? We close a whole load of kernel bugs with a message along the lines of "we rebased the kernel and have no clue if we addressed this bug, but we're closing it anyway -- please reopen if it's not fixed". Do we have a clue how many of those bugs remain closed because the bug is actually fixed, and how many of them remain closed because the reporter never looked back after filing it? (Not that I'm arguing against this way of closing bugs -- with the kwality of bug reporting and follow-up we get, I think it's the right thing for us to do. But let's be careful about what the numbers mean.) -- dwmw2 From notting at redhat.com Wed Jan 3 18:45:39 2007 From: notting at redhat.com (Bill Nottingham) Date: Wed, 3 Jan 2007 13:45:39 -0500 Subject: why I'm using Ubuntu instead of Fedora ATM In-Reply-To: <2cb10c440701031040u6dffb838o1490301d0a7b472e@mail.gmail.com> References: <2cb10c440701030059l49323ecaqeae7c7041f735d63@mail.gmail.com> <459B7B49.8040101@leemhuis.info> <2cb10c440701030858g4a0c84cdi3cc1f67186b3258f@mail.gmail.com> <20070103172612.GC26533@redhat.com> <2cb10c440701030931u4952585ev8909363d47e4ff31@mail.gmail.com> <20070103175718.GA30432@redhat.com> <20070103182359.GD23011@nostromo.devel.redhat.com> <2cb10c440701031040u6dffb838o1490301d0a7b472e@mail.gmail.com> Message-ID: <20070103184539.GA24288@nostromo.devel.redhat.com> Luis Villa (luis at tieguy.org) said: > >> 'Gnome' covers a pretty large package set, and I'll wager that > >> 'gnome + kernel' probably covers the majority of the bugs that get filed. > > > >Total FC6 bugs filed: 2826. > > > >Kernel: 427 > >GNOME: 578 (includes: apps, OO.o, firefox and kin, as they're really > >'desktop') > > > >So, 35%, give or take. > > For the purposes of this line drawing you have to look at development > organizations, not functional groupings, so GNOME != desktop. I'd be > embarassed if GNOME's bug fixing and prioritization were as bad as > OOo's, for example, or if GNOME's release cycle were as broken as > firefoxes. True. I was looking at it from a user perspective, not a devel group perspective. Taking out ff/tbird/oo.o removes 80 bugs, although that's still leaving apps at the evolution/rhythmbox/totem level. Bill From davej at redhat.com Wed Jan 3 18:53:47 2007 From: davej at redhat.com (Dave Jones) Date: Wed, 3 Jan 2007 13:53:47 -0500 Subject: why I'm using Ubuntu instead of Fedora ATM In-Reply-To: <1167849820.19205.10.camel@pmac.infradead.org> References: <2cb10c440701030059l49323ecaqeae7c7041f735d63@mail.gmail.com> <459B7B49.8040101@leemhuis.info> <2cb10c440701030858g4a0c84cdi3cc1f67186b3258f@mail.gmail.com> <20070103172612.GC26533@redhat.com> <1167849820.19205.10.camel@pmac.infradead.org> Message-ID: <20070103185347.GA7251@redhat.com> On Wed, Jan 03, 2007 at 06:43:40PM +0000, David Woodhouse wrote: > On Wed, 2007-01-03 at 12:26 -0500, Dave Jones wrote: > > Right now, rebases to newer kernel releases is the single provider > > of the majority of bugfixes we get reported. If we stopped doing > > that, we'd pretty much be giving up all hope of fixing kernel bugs. > > Er, I think you mean "bug closures" not necessarily "fixes"? > > We close a whole load of kernel bugs with a message along the lines of > "we rebased the kernel and have no clue if we addressed this bug, but > we're closing it anyway -- please reopen if it's not fixed". Err, no. On a rebase I put them into NEEDINFO_REPORTER and ask people to try again. If they report it got fixed, /then/ they get closed. Sadly a lot of them remain in NEEDINFO, and then get closed out a month or so later due to inactivity. Typically, we're talking >100 or so end up like this, and about a dozen get reopened or followed up with "yes it did get fixed, sorry I didnt respond". Dave -- http://www.codemonkey.org.uk From fedora at leemhuis.info Wed Jan 3 19:04:45 2007 From: fedora at leemhuis.info (Thorsten Leemhuis) Date: Wed, 03 Jan 2007 20:04:45 +0100 Subject: why I'm using Ubuntu instead of Fedora ATM In-Reply-To: <200701031331.02846.jkeating@redhat.com> References: <2cb10c440701030059l49323ecaqeae7c7041f735d63@mail.gmail.com> <200701031231.09906.jkeating@redhat.com> <459BEFC4.4020909@leemhuis.info> <200701031331.02846.jkeating@redhat.com> Message-ID: <459BFE4D.2050901@leemhuis.info> Jesse Keating schrieb: > On Wednesday 03 January 2007 13:02, Thorsten Leemhuis wrote: >> Sorry, but we IMHO did something seriously wrong beforehand already if >> we have to roll back packages *after* test3. > Test3 isn't what Luis wants. I know, but sorry, it was not clear to me that you meant what Luis said in your last mail, and not my compromise-approach I posted/you quoted. > He wants to be able to go from F7, to F8Test1 to > F8Test2/3 to F8, cleanly and without problems. And it would be nice (as I said already). But I'd already be glad if we could have the F7TEST3 to F7 way. CU thl P.S:Ahh, this subthread once again shows that discussions on mailing lists sometimes don't work very well... From jkeating at redhat.com Wed Jan 3 19:20:39 2007 From: jkeating at redhat.com (Jesse Keating) Date: Wed, 3 Jan 2007 14:20:39 -0500 Subject: why I'm using Ubuntu instead of Fedora ATM In-Reply-To: <459BFE4D.2050901@leemhuis.info> References: <2cb10c440701030059l49323ecaqeae7c7041f735d63@mail.gmail.com> <200701031331.02846.jkeating@redhat.com> <459BFE4D.2050901@leemhuis.info> Message-ID: <200701031420.39525.jkeating@redhat.com> On Wednesday 03 January 2007 14:04, Thorsten Leemhuis wrote: > And it would be nice (as I said already). But I'd already be glad if we > could have the F7TEST3 to F7 way. Yes, F7T3 to F7 would be nice, but as you said, sometimes we screw up. And if we screw up, would you rather we introduce epoch, or roll back the version? If we come up with a hard and fast rule that T3 must be upgradable to Final, that has some pretty serious QA implications, and the QA team (I'm looking at you Will) needs to have a say in that. -- Jesse Keating Release Engineer: Fedora -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 189 bytes Desc: not available URL: From wwoods at redhat.com Wed Jan 3 19:41:36 2007 From: wwoods at redhat.com (Will Woods) Date: Wed, 03 Jan 2007 14:41:36 -0500 Subject: The QA Problem In-Reply-To: References: <2cb10c440701030059l49323ecaqeae7c7041f735d63@mail.gmail.com> <459B7B49.8040101@leemhuis.info> <20070103163903.GA22376@nostromo.devel.redhat.com> <20070103170805.GA23011@nostromo.devel.redhat.com> Message-ID: <1167853296.7767.87.camel@metroid.rdu.redhat.com> On Wed, 2007-01-03 at 13:02 -0500, Greg Dekoenigsberg wrote: > On Wed, 3 Jan 2007, Bill Nottingham wrote: > > >>>> I'm wondering if we could provide solutions for both users: One update > >>>> channel that only gets security updates and important bugfixes while the > >>>> other is a bit more bold -- we for example could have firefox2 in the > >>>> bold channel for FC6 while shipping the latest firefox 1.5.x in the more > >>>> conservative channel. > >>> > >>> So, an idea like this: > >>> > >>> - starts to exponentially expand the QA problem > >> > >> Bill, I love you, but I'm going to have to call bullshit on this one. > >> The answer is not "avoid making QA harder", the answer is to SOLVE THE QA > >> PROBLEM. > > > > Sure, but which of these plans make more sense: > > > > 1) > > > > - Solve the QA problem for our repo configurations as they exist > > - Expand the QA solution to new, multiple, disparate and conflicting > > repositories > > > > 2) > > > > - Expand into new, multiple, disparate and conflicting repositories > > - Then try to solve the QA problem > > > > Honestly, before we can do multiple experimental repositories of this, > > that, and the other, we need to get our *OWN* house in order. > > I agree with this completely -- if, and only if, we actually make a > concerted effort at fixing the QA problem. > > Which means articulating the QA problem, actually. Okay. "The QA Problem" is a big, fat, multi-headed monster. I mentally group QA into four tasks, ordered by priority: Task #1: Testing updates to stable releases Task #2: Testing new releases before they go out (rawhide, TestX, etc.) Task #3: Bug triage (this is the stuff we miss in #1 and #2) Task #4: Writing tools and docs to make the previous tasks easier Examining the first task - we don't currently have the *dedicated* manpower to *guarantee* that every package will be tested and approved by the QA group before it leaves updates-testing and goes into updates. I don't think we actually lack for *available* manpower. AFAIK there's plenty of people willing and able to install packages from updates-testing. The problems here are, I believe, a good summary of "The QA Problem". Problem #1: Testing currently requires a lot of skill, which reduces usable manpower. We lack how-to-test documents, so each tester must know how to set up and test any given package/feature on his own. New features don't necessarily come with much documentation (e.g. iSCSI). - Possible solution(s): More docs would lower the barrier to entry. Setting up an official Fedora QA group will help keep track of team strength and help everyone work together. Problem #2: Testing requires a lot of time. The simple tasks - e.g. "start httpd and make sure it can actually serve files" - are not automated. Nobody knows if anyone else has already tested a package, so there's duplicated work. - Possible solutions: Automation (beaker) and infrastructure work (fedora-updates-system) will help cut down on wasted time. Problem #3: Motivation to report bugs / testresults is low. Even though it's easy to install packages from updates-testing, reporting problems with them is far harder than it needs to be. The same old bugs get reported over and over while some new problems don't get reported because the tester didn't want to spend 15 minutes looking for duplicate bugs, figuring out the appropriate component/version/etc. - Possible solutions: The updates system should have an easy way to report common problems with packages in updates-testing. A modified bug-buddy for Fedora would be very helpful here. These tools should also show the user commonly-reported bugs, and allow them to easily add a "me too" comment. A "Bugzilla RPG" or other ranking system (like GNOME's point system or Launchpad's Karma ranking) makes bug reporting and triage more interesting. So. Does that define The QA Problem? Or are there other issues I'm forgetting? -w -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 189 bytes Desc: This is a digitally signed message part URL: From mmcgrath at fedoraproject.org Wed Jan 3 19:41:49 2007 From: mmcgrath at fedoraproject.org (Mike McGrath) Date: Wed, 3 Jan 2007 13:41:49 -0600 Subject: why I'm using Ubuntu instead of Fedora ATM In-Reply-To: <200701031420.39525.jkeating@redhat.com> References: <2cb10c440701030059l49323ecaqeae7c7041f735d63@mail.gmail.com> <200701031331.02846.jkeating@redhat.com> <459BFE4D.2050901@leemhuis.info> <200701031420.39525.jkeating@redhat.com> Message-ID: <3237e4410701031141n214ac434t53cacee42b416b5b@mail.gmail.com> On 1/3/07, Jesse Keating wrote: > On Wednesday 03 January 2007 14:04, Thorsten Leemhuis wrote: > > And it would be nice (as I said already). But I'd already be glad if we > > could have the F7TEST3 to F7 way. > > Yes, F7T3 to F7 would be nice, but as you said, sometimes we screw up. And if > we screw up, would you rather we introduce epoch, or roll back the version? > If we come up with a hard and fast rule that T3 must be upgradable to Final, > that has some pretty serious QA implications, and the QA team (I'm looking at > you Will) needs to have a say in that. > T[1-3] and rawhide ain't grandma's Linux, those expecting it not to break shouldn't be using it. It would be nice to have more of a choice though when upgrading from FC5 though. Our options are FC6 (to be EOL'd in 6 months) or FC7t2. I'm sure this has been discussed before but I'm bringing it up again. -Mike From davidz at redhat.com Wed Jan 3 19:52:45 2007 From: davidz at redhat.com (David Zeuthen) Date: Wed, 03 Jan 2007 14:52:45 -0500 Subject: why I'm using Ubuntu instead of Fedora ATM In-Reply-To: <200701031420.39525.jkeating@redhat.com> References: <2cb10c440701030059l49323ecaqeae7c7041f735d63@mail.gmail.com> <200701031331.02846.jkeating@redhat.com> <459BFE4D.2050901@leemhuis.info> <200701031420.39525.jkeating@redhat.com> Message-ID: <1167853965.5694.7.camel@zelda.fubar.dk> On Wed, 2007-01-03 at 14:20 -0500, Jesse Keating wrote: > On Wednesday 03 January 2007 14:04, Thorsten Leemhuis wrote: > > And it would be nice (as I said already). But I'd already be glad if we > > could have the F7TEST3 to F7 way. > > Yes, F7T3 to F7 would be nice, but as you said, sometimes we screw up. And if > we screw up, would you rather we introduce epoch, or roll back the version? Maybe I'm missing something or maybe I just don't get it, but how can rolling back version numbers *instead* of bumping another number (Epoch) ever be considered a solution? Have we done this before, and if so, what was the justification? Thanks. David From wwoods at redhat.com Wed Jan 3 19:58:11 2007 From: wwoods at redhat.com (Will Woods) Date: Wed, 03 Jan 2007 14:58:11 -0500 Subject: why I'm using Ubuntu instead of Fedora ATM In-Reply-To: <200701031420.39525.jkeating@redhat.com> References: <2cb10c440701030059l49323ecaqeae7c7041f735d63@mail.gmail.com> <200701031331.02846.jkeating@redhat.com> <459BFE4D.2050901@leemhuis.info> <200701031420.39525.jkeating@redhat.com> Message-ID: <1167854291.7767.104.camel@metroid.rdu.redhat.com> On Wed, 2007-01-03 at 14:20 -0500, Jesse Keating wrote: > On Wednesday 03 January 2007 14:04, Thorsten Leemhuis wrote: > > And it would be nice (as I said already). But I'd already be glad if we > > could have the F7TEST3 to F7 way. > > Yes, F7T3 to F7 would be nice, but as you said, sometimes we screw up. And if > we screw up, would you rather we introduce epoch, or roll back the version? > If we come up with a hard and fast rule that T3 must be upgradable to Final, > that has some pretty serious QA implications, and the QA team (I'm looking at > you Will) needs to have a say in that. Correct me if I'm wrong, but it's always been *possible* to upgrade to Final from Test3 already, right? It may require a little massaging but it's not impossible, or even all that hard. We just don't *guarantee* that it will magically Just Work. Rather than putting extra effort into making that guarantee, I'd prefer to make major release upgrades magically Just Work. If you can't handle installing TestX without that guarantee (i.e. if the idea of having to manually remove/downgrade a few packages first is too much for you), you could always try the LiveCD of TestX instead. We could have a compromise - we will at least *try* to help anyone using Test3 upgrade to Final when it finally comes. We can keep a wiki page with info/workarounds, but I don't think it's wise or desirable to force the devs to cram that stuff into anaconda and such. -w -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 189 bytes Desc: This is a digitally signed message part URL: From davidz at redhat.com Wed Jan 3 20:00:04 2007 From: davidz at redhat.com (David Zeuthen) Date: Wed, 03 Jan 2007 15:00:04 -0500 Subject: why I'm using Ubuntu instead of Fedora ATM In-Reply-To: <3237e4410701031141n214ac434t53cacee42b416b5b@mail.gmail.com> References: <2cb10c440701030059l49323ecaqeae7c7041f735d63@mail.gmail.com> <200701031331.02846.jkeating@redhat.com> <459BFE4D.2050901@leemhuis.info> <200701031420.39525.jkeating@redhat.com> <3237e4410701031141n214ac434t53cacee42b416b5b@mail.gmail.com> Message-ID: <1167854404.5694.14.camel@zelda.fubar.dk> On Wed, 2007-01-03 at 13:41 -0600, Mike McGrath wrote: > On 1/3/07, Jesse Keating wrote: > > On Wednesday 03 January 2007 14:04, Thorsten Leemhuis wrote: > > > And it would be nice (as I said already). But I'd already be glad if we > > > could have the F7TEST3 to F7 way. > > > > Yes, F7T3 to F7 would be nice, but as you said, sometimes we screw up. And if > > we screw up, would you rather we introduce epoch, or roll back the version? > > If we come up with a hard and fast rule that T3 must be upgradable to Final, > > that has some pretty serious QA implications, and the QA team (I'm looking at > > you Will) needs to have a say in that. > > > > T[1-3] and rawhide ain't grandma's Linux, those expecting it not to > break shouldn't be using it. I refuse to buy arguments like this. We want our users to at least have an upgrade path and if the only thing preventing this is certain people being afraid of bumping a superficial number (the Epoch) and doing things like rolling back version numbers then, perhaps, our package maintenance procedures needs to be reconsidered. (Then again, as I noted in my other mail, perhaps I don't fully understand why one would ever roll back a version number (rather than bumping the Epoch) knowing it will break upgrade paths.) David From jkeating at redhat.com Wed Jan 3 20:02:35 2007 From: jkeating at redhat.com (Jesse Keating) Date: Wed, 3 Jan 2007 15:02:35 -0500 Subject: why I'm using Ubuntu instead of Fedora ATM In-Reply-To: <3237e4410701031141n214ac434t53cacee42b416b5b@mail.gmail.com> References: <2cb10c440701030059l49323ecaqeae7c7041f735d63@mail.gmail.com> <200701031420.39525.jkeating@redhat.com> <3237e4410701031141n214ac434t53cacee42b416b5b@mail.gmail.com> Message-ID: <200701031502.35599.jkeating@redhat.com> On Wednesday 03 January 2007 14:41, Mike McGrath wrote: > T[1-3] and rawhide ain't grandma's Linux, those expecting it not to > break shouldn't be using it. ?It would be nice to have more of a > choice though when upgrading from FC5 though. ?Our options are FC6 (to > be EOL'd in 6 months) or FC7t2. ?I'm sure this has been discussed > before but I'm bringing it up again. The new lifespan proposal would make FC6 viable until one month after F8 is released. -- Jesse Keating Release Engineer: Fedora -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 189 bytes Desc: not available URL: From jkeating at redhat.com Wed Jan 3 20:04:18 2007 From: jkeating at redhat.com (Jesse Keating) Date: Wed, 3 Jan 2007 15:04:18 -0500 Subject: why I'm using Ubuntu instead of Fedora ATM In-Reply-To: <1167853965.5694.7.camel@zelda.fubar.dk> References: <2cb10c440701030059l49323ecaqeae7c7041f735d63@mail.gmail.com> <200701031420.39525.jkeating@redhat.com> <1167853965.5694.7.camel@zelda.fubar.dk> Message-ID: <200701031504.18618.jkeating@redhat.com> On Wednesday 03 January 2007 14:52, David Zeuthen wrote: > Maybe I'm missing something or maybe I just don't get it, but how can > rolling back version numbers *instead* of bumping another number (Epoch) > ever be considered a solution? Have we done this before, and if so, what > was the justification? Thanks. Yes, we've rolled back before. I don't have specific examples, but it happens. We've also flat out removed packages introduced during test times. Introducing epochs is ugly and will hang over the package forever, and every effort should be made to avoid it. Because we have the 'right' currently to not worry much about going from T1 to T2 or T2 to T3 or T3 to final, we have a way of preventing epoch. -- Jesse Keating Release Engineer: Fedora -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 189 bytes Desc: not available URL: From notting at redhat.com Wed Jan 3 20:04:40 2007 From: notting at redhat.com (Bill Nottingham) Date: Wed, 3 Jan 2007 15:04:40 -0500 Subject: The QA Problem In-Reply-To: <1167853296.7767.87.camel@metroid.rdu.redhat.com> References: <2cb10c440701030059l49323ecaqeae7c7041f735d63@mail.gmail.com> <459B7B49.8040101@leemhuis.info> <20070103163903.GA22376@nostromo.devel.redhat.com> <20070103170805.GA23011@nostromo.devel.redhat.com> <1167853296.7767.87.camel@metroid.rdu.redhat.com> Message-ID: <20070103200440.GA25367@nostromo.devel.redhat.com> Will Woods (wwoods at redhat.com) said: > So. Does that define The QA Problem? Or are there other issues I'm > forgetting? If we're going to support 'always being able to upgrade in the devel stream to stable, and vice versa', we need to eliminate things that break upgrades. This involves: 1) automated post-build testing. Find broken packages before they go anywhere. Which means we need to find all the things that could break the repository, and write tests for them. 2) non-automated post-build testing before packages are pushed. Includes: - smoke testing - %post/%preun/%etc auditing Presumably more stuff that I'm missing here. Bill From davidz at redhat.com Wed Jan 3 20:14:36 2007 From: davidz at redhat.com (David Zeuthen) Date: Wed, 03 Jan 2007 15:14:36 -0500 Subject: bump epoch, don't roll back versions (Was Re: why I'm using Ubuntu instead of Fedora ATM) In-Reply-To: <200701031504.18618.jkeating@redhat.com> References: <2cb10c440701030059l49323ecaqeae7c7041f735d63@mail.gmail.com> <200701031420.39525.jkeating@redhat.com> <1167853965.5694.7.camel@zelda.fubar.dk> <200701031504.18618.jkeating@redhat.com> Message-ID: <1167855276.5694.22.camel@zelda.fubar.dk> On Wed, 2007-01-03 at 15:04 -0500, Jesse Keating wrote: > On Wednesday 03 January 2007 14:52, David Zeuthen wrote: > > Maybe I'm missing something or maybe I just don't get it, but how can > > rolling back version numbers *instead* of bumping another number (Epoch) > > ever be considered a solution? Have we done this before, and if so, what > > was the justification? Thanks. > > Yes, we've rolled back before. I don't have specific examples, but it > happens. We've also flat out removed packages introduced during test times. > Introducing epochs is ugly and will hang over the package forever, and every > effort should be made to avoid it. Because we have the 'right' currently to > not worry much about going from T1 to T2 or T2 to T3 or T3 to final, we have > a way of preventing epoch. Sorry to sound non-constructive, but can we please stop breaking upgrade paths just because someone happens to think that "epochs is ugly" (left over packages is much less of a problem). I'd like to go even further and ask for our build system to enforce this rule. The justification is that it's only a number, and this practice is bordering introducing bugs by refactoring source code just because it's "less ugly" that way. Please tell me where I can file a bug (I'm not exactly sure, I used the vauge term "Build System" but is this in Bugzilla yet?) so this issue won't be forgotten. Thanks. David From sundaram at fedoraproject.org Wed Jan 3 20:12:20 2007 From: sundaram at fedoraproject.org (Rahul Sundaram) Date: Thu, 04 Jan 2007 01:42:20 +0530 Subject: why I'm using Ubuntu instead of Fedora ATM In-Reply-To: <1167854291.7767.104.camel@metroid.rdu.redhat.com> References: <2cb10c440701030059l49323ecaqeae7c7041f735d63@mail.gmail.com> <200701031331.02846.jkeating@redhat.com> <459BFE4D.2050901@leemhuis.info> <200701031420.39525.jkeating@redhat.com> <1167854291.7767.104.camel@metroid.rdu.redhat.com> Message-ID: <459C0E24.1020106@fedoraproject.org> Will Woods wrote: > We could have a compromise - we will at least *try* to help anyone using > Test3 upgrade to Final when it finally comes. We can keep a wiki page > with info/workarounds, but I don't think it's wise or desirable to force > the devs to cram that stuff into anaconda and such. The updates system can prevent problems by checking and refusing to release a update with packaging issues that would be hindrance for updates. We could have a separate upgrade module which does that release specific massaging/hacks/workarounds/obsoleting dropped packages etc and that can be called from a live upgrade tool as well as Anaconda. Pup should notify users when the release they are using reaches end of life as well as availability of new releases and the live upgrade tool should be hooked into pup. For the next release, we might maintain it as a experimental feature and stabilize this as we have move forward. There are some cases where a live upgrade wont work efficiently but for majority of use cases the only reason live upgrades dont work properly are real packaging problems which we should be fixing anyway. Rahul From mmcgrath at fedoraproject.org Wed Jan 3 20:19:28 2007 From: mmcgrath at fedoraproject.org (Mike McGrath) Date: Wed, 3 Jan 2007 14:19:28 -0600 Subject: bump epoch, don't roll back versions (Was Re: why I'm using Ubuntu instead of Fedora ATM) In-Reply-To: <1167855276.5694.22.camel@zelda.fubar.dk> References: <2cb10c440701030059l49323ecaqeae7c7041f735d63@mail.gmail.com> <200701031420.39525.jkeating@redhat.com> <1167853965.5694.7.camel@zelda.fubar.dk> <200701031504.18618.jkeating@redhat.com> <1167855276.5694.22.camel@zelda.fubar.dk> Message-ID: <3237e4410701031219j46ea4009y196418cff4e0c0b@mail.gmail.com> On 1/3/07, David Zeuthen wrote: > On Wed, 2007-01-03 at 15:04 -0500, Jesse Keating wrote: > > On Wednesday 03 January 2007 14:52, David Zeuthen wrote: > > > Maybe I'm missing something or maybe I just don't get it, but how can > > > rolling back version numbers *instead* of bumping another number (Epoch) > > > ever be considered a solution? Have we done this before, and if so, what > > > was the justification? Thanks. > > > > Yes, we've rolled back before. I don't have specific examples, but it > > happens. We've also flat out removed packages introduced during test times. > > Introducing epochs is ugly and will hang over the package forever, and every > > effort should be made to avoid it. Because we have the 'right' currently to > > not worry much about going from T1 to T2 or T2 to T3 or T3 to final, we have > > a way of preventing epoch. > > Sorry to sound non-constructive, but can we please stop breaking upgrade > paths just because someone happens to think that "epochs is ugly" (left > over packages is much less of a problem). I'd like to go even further > and ask for our build system to enforce this rule. The justification is > that it's only a number, and this practice is bordering introducing bugs > by refactoring source code just because it's "less ugly" that way. > > Please tell me where I can file a bug (I'm not exactly sure, I used the > vauge term "Build System" but is this in Bugzilla yet?) so this issue > won't be forgotten. Thanks. > > David > > This probably belongs on fedora-devel -Mike From jkeating at redhat.com Wed Jan 3 20:21:53 2007 From: jkeating at redhat.com (Jesse Keating) Date: Wed, 3 Jan 2007 15:21:53 -0500 Subject: bump epoch, don't roll back versions (Was Re: why I'm using Ubuntu instead of Fedora ATM) In-Reply-To: <1167855276.5694.22.camel@zelda.fubar.dk> References: <2cb10c440701030059l49323ecaqeae7c7041f735d63@mail.gmail.com> <200701031504.18618.jkeating@redhat.com> <1167855276.5694.22.camel@zelda.fubar.dk> Message-ID: <200701031521.54012.jkeating@redhat.com> On Wednesday 03 January 2007 15:14, David Zeuthen wrote: > Sorry to sound non-constructive, but can we please stop breaking upgrade > paths just because someone happens to think that "epochs is ugly" (left > over packages is much less of a problem). I'd like to go even further > and ask for our build system to enforce this rule. The justification is > that it's only a number, and this practice is bordering introducing bugs > by refactoring source code just because it's "less ugly" that way. I'll let conversation happen on the merrits of bumping or not bumping epoch. However for buildsystem to enforce this, that's a pretty tough nut to solve, since any build could get tagged for any variety of collections, regardless of nvr. In fact, the buildsystem (by design) only enforces unique n-v-r, so you couldn't do 1:n-v-r and 2:n-v-r, the buildsystem would freak. You'd have to do 2:n-v-r+1 or some other unique 'n-v-r'. -- Jesse Keating Release Engineer: Fedora -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 189 bytes Desc: not available URL: From jkeating at redhat.com Wed Jan 3 20:27:50 2007 From: jkeating at redhat.com (Jesse Keating) Date: Wed, 3 Jan 2007 15:27:50 -0500 Subject: bump epoch, don't roll back versions (Was Re: why I'm using Ubuntu instead of Fedora ATM) In-Reply-To: <200701031521.54012.jkeating@redhat.com> References: <2cb10c440701030059l49323ecaqeae7c7041f735d63@mail.gmail.com> <1167855276.5694.22.camel@zelda.fubar.dk> <200701031521.54012.jkeating@redhat.com> Message-ID: <200701031527.50804.jkeating@redhat.com> On Wednesday 03 January 2007 15:21, Jesse Keating wrote: > I'll let conversation happen on the merrits of bumping or not bumping > epoch. Hrm, I meant let conversation happen by folks smarter than me. -- Jesse Keating Release Engineer: Fedora -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 189 bytes Desc: not available URL: From davej at redhat.com Wed Jan 3 20:41:05 2007 From: davej at redhat.com (Dave Jones) Date: Wed, 3 Jan 2007 15:41:05 -0500 Subject: The QA Problem In-Reply-To: <1167853296.7767.87.camel@metroid.rdu.redhat.com> References: <2cb10c440701030059l49323ecaqeae7c7041f735d63@mail.gmail.com> <459B7B49.8040101@leemhuis.info> <20070103163903.GA22376@nostromo.devel.redhat.com> <20070103170805.GA23011@nostromo.devel.redhat.com> <1167853296.7767.87.camel@metroid.rdu.redhat.com> Message-ID: <20070103204105.GA24013@redhat.com> On Wed, Jan 03, 2007 at 02:41:36PM -0500, Will Woods wrote: > Task #1: Testing updates to stable releases > Task #2: Testing new releases before they go out (rawhide, TestX, etc.) > Task #3: Bug triage (this is the stuff we miss in #1 and #2) > Task #4: Writing tools and docs to make the previous tasks easier > > Examining the first task - we don't currently have the *dedicated* > manpower to *guarantee* that every package will be tested and approved > by the QA group before it leaves updates-testing and goes into updates. > > I don't think we actually lack for *available* manpower. AFAIK there's > plenty of people willing and able to install packages from > updates-testing. We could definitly use more. I've been bitten a few times by silly bugs that should have been caught by users running -testing, which didn't become obvious until it got moved to -final. One potential way to increase usage of -testing is to offer a question in firstboot whether or not to enable it. As Luis mentioned, he didn't even know about this until recently. Makes me wonder just how many other users we have who aren't aware of it. Quite often I see in bugzilla.. 'This should be fixed in the build in updates-testing, can you give it a try?' and then the bug goes silent. Eventually after it gets pushed to final the reporter comes back "Yes, that fixed it". We're definitly missing out on some potential testers. Dave -- http://www.codemonkey.org.uk From katzj at redhat.com Wed Jan 3 21:30:16 2007 From: katzj at redhat.com (Jeremy Katz) Date: Wed, 03 Jan 2007 16:30:16 -0500 Subject: The QA Problem In-Reply-To: <20070103204105.GA24013@redhat.com> References: <2cb10c440701030059l49323ecaqeae7c7041f735d63@mail.gmail.com> <459B7B49.8040101@leemhuis.info> <20070103163903.GA22376@nostromo.devel.redhat.com> <20070103170805.GA23011@nostromo.devel.redhat.com> <1167853296.7767.87.camel@metroid.rdu.redhat.com> <20070103204105.GA24013@redhat.com> Message-ID: <1167859816.14822.30.camel@aglarond.local> On Wed, 2007-01-03 at 15:41 -0500, Dave Jones wrote: > One potential way to increase usage of -testing is to offer > a question in firstboot whether or not to enable it. > As Luis mentioned, he didn't even know about this until recently. > Makes me wonder just how many other users we have who aren't > aware of it. Another thing that will help here is actually having some reasonable UI for configuring available repositories[1] which I'm planning to make sure gets done for F7. Jeremy [1] This includes at the very least (and in order of importance) enabling/disabling existing repos "available" on disk, adding repos and removing repos. From katzj at redhat.com Wed Jan 3 21:31:49 2007 From: katzj at redhat.com (Jeremy Katz) Date: Wed, 03 Jan 2007 16:31:49 -0500 Subject: why I'm using Ubuntu instead of Fedora ATM In-Reply-To: <459C0E24.1020106@fedoraproject.org> References: <2cb10c440701030059l49323ecaqeae7c7041f735d63@mail.gmail.com> <200701031331.02846.jkeating@redhat.com> <459BFE4D.2050901@leemhuis.info> <200701031420.39525.jkeating@redhat.com> <1167854291.7767.104.camel@metroid.rdu.redhat.com> <459C0E24.1020106@fedoraproject.org> Message-ID: <1167859909.14822.33.camel@aglarond.local> On Thu, 2007-01-04 at 01:42 +0530, Rahul Sundaram wrote: > Pup should notify users when the release they are using reaches end of > life as well as availability of new releases and the live upgrade tool > should be hooked into pup. For the next release, we might maintain it as > a experimental feature and stabilize this as we have move forward. There > are some cases where a live upgrade wont work efficiently but for > majority of use cases the only reason live upgrades dont work properly > are real packaging problems which we should be fixing anyway. Sadly, this isn't actually the case. The number of "weird" cases that are just "remove this package" is far smaller than the other bizarre bits that have caused live upgrades to not work reliably. Jeremy From blizzard at redhat.com Wed Jan 3 21:34:15 2007 From: blizzard at redhat.com (Christopher Blizzard) Date: Wed, 03 Jan 2007 16:34:15 -0500 Subject: why I'm using Ubuntu instead of Fedora ATM In-Reply-To: <200701031231.09906.jkeating@redhat.com> References: <2cb10c440701030059l49323ecaqeae7c7041f735d63@mail.gmail.com> <2cb10c440701030858g4a0c84cdi3cc1f67186b3258f@mail.gmail.com> <459BE586.9080509@leemhuis.info> <200701031231.09906.jkeating@redhat.com> Message-ID: <459C2157.7070205@redhat.com> Jesse Keating wrote: > On Wednesday 03 January 2007 12:19, Thorsten Leemhuis wrote: >> Actually I'd like to see this, too, but Fedora people seem to be so >> unwilling to support something like that so I tried the "better ask for >> something that's not that hard to realize and lies between the current >> solution and the best/ideal solution; maybe then the chances to get what >> we ask for are higher and it make one big part of the problem go away, >> even if parts of the problem remain"-approach > > Well, we could do this, but it would involve an explosion of epochs :/ Our > ability to roll back package versions that were 'mistakes' is very very > useful. The only way to keep this is to use epochs to "force" a lower > version package to win. Once epoch, always epoch. I would like to say that this is the only way that we can fix this _today_. This is one of the areas where I would love to see some changes in RPM. It's a constant source of pain for our developers and by effect our users. (Add an Oops: tag that has a list of versions that this package is _actually_ older than?) Just that thinking only of what we can do today is not always the best way to think about the problem. Luis has a point - let's apply some thinking about not only how we can solve the problem today (or not) and figure out what changes we can make to help it down the road. --Chris From dwmw2 at infradead.org Wed Jan 3 21:42:51 2007 From: dwmw2 at infradead.org (David Woodhouse) Date: Wed, 03 Jan 2007 21:42:51 +0000 Subject: why I'm using Ubuntu instead of Fedora ATM In-Reply-To: <1167859909.14822.33.camel@aglarond.local> References: <2cb10c440701030059l49323ecaqeae7c7041f735d63@mail.gmail.com> <200701031331.02846.jkeating@redhat.com> <459BFE4D.2050901@leemhuis.info> <200701031420.39525.jkeating@redhat.com> <1167854291.7767.104.camel@metroid.rdu.redhat.com> <459C0E24.1020106@fedoraproject.org> <1167859909.14822.33.camel@aglarond.local> Message-ID: <1167860571.19205.31.camel@pmac.infradead.org> On Wed, 2007-01-03 at 16:31 -0500, Jeremy Katz wrote: > On Thu, 2007-01-04 at 01:42 +0530, Rahul Sundaram wrote: > > Pup should notify users when the release they are using reaches end of > > life as well as availability of new releases and the live upgrade tool > > should be hooked into pup. For the next release, we might maintain it as > > a experimental feature and stabilize this as we have move forward. There > > are some cases where a live upgrade wont work efficiently but for > > majority of use cases the only reason live upgrades dont work properly > > are real packaging problems which we should be fixing anyway. > > Sadly, this isn't actually the case. The number of "weird" cases that > are just "remove this package" is far smaller than the other bizarre > bits that have caused live upgrades to not work reliably. That's not my experience. Live upgrades have worked fine for me ever since about RHL5. Sometimes there's a glitch which you need to deal with manually (like having to reformat swap space for 64KiB pages), but stuff like that happens even with "installer" upgrades too. The most painful thing about it is fixing Apache after the upgrade -- or maybe it's rebooting the remote machines onto a much newer kernel. But neither of those are special to the "live" upgrade path. -- dwmw2 From wwoods at redhat.com Wed Jan 3 21:43:25 2007 From: wwoods at redhat.com (Will Woods) Date: Wed, 03 Jan 2007 16:43:25 -0500 Subject: why I'm using Ubuntu instead of Fedora ATM In-Reply-To: <1167859909.14822.33.camel@aglarond.local> References: <2cb10c440701030059l49323ecaqeae7c7041f735d63@mail.gmail.com> <200701031331.02846.jkeating@redhat.com> <459BFE4D.2050901@leemhuis.info> <200701031420.39525.jkeating@redhat.com> <1167854291.7767.104.camel@metroid.rdu.redhat.com> <459C0E24.1020106@fedoraproject.org> <1167859909.14822.33.camel@aglarond.local> Message-ID: <1167860605.7767.120.camel@metroid.rdu.redhat.com> On Wed, 2007-01-03 at 16:31 -0500, Jeremy Katz wrote: > On Thu, 2007-01-04 at 01:42 +0530, Rahul Sundaram wrote: > > Pup should notify users when the release they are using reaches end of > > life as well as availability of new releases and the live upgrade tool > > should be hooked into pup. For the next release, we might maintain it as > > a experimental feature and stabilize this as we have move forward. There > > are some cases where a live upgrade wont work efficiently but for > > majority of use cases the only reason live upgrades dont work properly > > are real packaging problems which we should be fixing anyway. > > Sadly, this isn't actually the case. The number of "weird" cases that > are just "remove this package" is far smaller than the other bizarre > bits that have caused live upgrades to not work reliably. I agree - there's just too many pitfalls for a live upgrade to be viable. However! I think we can have a *mostly*-live update - package selection, dep resolution, and package downloading can all happen without rebooting, and those are the parts that take the longest. Once all that's done, we can reboot into the installer kernel/initrd and upgrade using the packages that are already sitting on the local disk. We even know which device it's on. That part could be basically automatic. I think. -w -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 189 bytes Desc: This is a digitally signed message part URL: From rgarth at gmail.com Wed Jan 3 21:43:48 2007 From: rgarth at gmail.com (Rob Garth) Date: Thu, 4 Jan 2007 08:43:48 +1100 Subject: why I'm using Ubuntu instead of Fedora ATM In-Reply-To: <3237e4410701030649r51e1a4ccw30a89d6df6124d2d@mail.gmail.com> References: <2cb10c440701030059l49323ecaqeae7c7041f735d63@mail.gmail.com> <459B7B49.8040101@leemhuis.info> <183d138b0701030210g7127c6a0s10669ce333231cd6@mail.gmail.com> <1167834163.4869.2.camel@localhost.localdomain> <3237e4410701030649r51e1a4ccw30a89d6df6124d2d@mail.gmail.com> Message-ID: <183d138b0701031343m268e4933r119c3a94df501b6@mail.gmail.com> > > I hear this argument a lot. The fact is that Fedora is the 'upstream' > for many projects. We're often the first distro to pick up a new > technology and integrate it. Just because RHEL is one distro that > gets it after us shouldn't be counted against Fedora. Fedora is the upstream of many projects. Some of which are considered productions systems, RHEL and OLPC, and probably others. I am guessing that these projects all maintian there own patches, and may indeed have a longer support life the Fedora. Doesn't it make sense for some of the maintainence work to be done in the upstream. I will shut up about this now though, sorry to stir the pot. -------------- next part -------------- An HTML attachment was scrubbed... URL: From wwoods at redhat.com Wed Jan 3 21:48:37 2007 From: wwoods at redhat.com (Will Woods) Date: Wed, 03 Jan 2007 16:48:37 -0500 Subject: The QA Problem In-Reply-To: <1167859816.14822.30.camel@aglarond.local> References: <2cb10c440701030059l49323ecaqeae7c7041f735d63@mail.gmail.com> <459B7B49.8040101@leemhuis.info> <20070103163903.GA22376@nostromo.devel.redhat.com> <20070103170805.GA23011@nostromo.devel.redhat.com> <1167853296.7767.87.camel@metroid.rdu.redhat.com> <20070103204105.GA24013@redhat.com> <1167859816.14822.30.camel@aglarond.local> Message-ID: <1167860917.7767.123.camel@metroid.rdu.redhat.com> On Wed, 2007-01-03 at 16:30 -0500, Jeremy Katz wrote: > On Wed, 2007-01-03 at 15:41 -0500, Dave Jones wrote: > > One potential way to increase usage of -testing is to offer > > a question in firstboot whether or not to enable it. > > As Luis mentioned, he didn't even know about this until recently. > > Makes me wonder just how many other users we have who aren't > > aware of it. > > Another thing that will help here is actually having some reasonable UI > for configuring available repositories[1] which I'm planning to make > sure gets done for F7. > > Jeremy > > [1] This includes at the very least (and in order of importance) > enabling/disabling existing repos "available" on disk, adding repos and > removing repos. Oooh. I hope this will also include some nice backend bits so that (for example) a theoretical fedora-bug-buddy could offer to turn on updates-testing and look for an updated package. -w -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 189 bytes Desc: This is a digitally signed message part URL: From katzj at redhat.com Wed Jan 3 21:54:20 2007 From: katzj at redhat.com (Jeremy Katz) Date: Wed, 03 Jan 2007 16:54:20 -0500 Subject: why I'm using Ubuntu instead of Fedora ATM In-Reply-To: <1167860571.19205.31.camel@pmac.infradead.org> References: <2cb10c440701030059l49323ecaqeae7c7041f735d63@mail.gmail.com> <200701031331.02846.jkeating@redhat.com> <459BFE4D.2050901@leemhuis.info> <200701031420.39525.jkeating@redhat.com> <1167854291.7767.104.camel@metroid.rdu.redhat.com> <459C0E24.1020106@fedoraproject.org> <1167859909.14822.33.camel@aglarond.local> <1167860571.19205.31.camel@pmac.infradead.org> Message-ID: <1167861260.14822.46.camel@aglarond.local> On Wed, 2007-01-03 at 21:42 +0000, David Woodhouse wrote: > On Wed, 2007-01-03 at 16:31 -0500, Jeremy Katz wrote: > > On Thu, 2007-01-04 at 01:42 +0530, Rahul Sundaram wrote: > > > Pup should notify users when the release they are using reaches end of > > > life as well as availability of new releases and the live upgrade tool > > > should be hooked into pup. For the next release, we might maintain it as > > > a experimental feature and stabilize this as we have move forward. There > > > are some cases where a live upgrade wont work efficiently but for > > > majority of use cases the only reason live upgrades dont work properly > > > are real packaging problems which we should be fixing anyway. > > > > Sadly, this isn't actually the case. The number of "weird" cases that > > are just "remove this package" is far smaller than the other bizarre > > bits that have caused live upgrades to not work reliably. > > That's not my experience. Live upgrades have worked fine for me ever > since about RHL5. I can generally get them to work as well. But there are weird cases that have been dealt with over time; sure they're work-aroundable if you're pretty saavy. It's not, though, the experience that most users want. Live upgrades can also lead to "really interesting" states if you, say, unplug your machine in the middle[1]. For a quick list of "off the top of my head weird problem" examples: * Going from static /dev -> udev involved some interesting shenanigans to avoid all of your devices disappearing mid-upgrade. * Newer SELinux policy depending on a kernel supporting that policy version. * glibc being built with a dependency on newer kernels for some new feature * New rpm functionality used by packages requiring a new version of rpm to install them The number of releases where there is one of these sorts of things is higher than the number where we don't have them. > Sometimes there's a glitch which you need to deal with manually (like > having to reformat swap space for 64KiB pages), but stuff like that > happens even with "installer" upgrades too. The key being that for these "manual" things, anaconda can and *DOES* take care of them[2]. At least, for the cases where people have tested upgrades over test releases and noticed a problem Jeremy [1] Yes, this isn't "good" if it happens when they're upgrading via anaconda also, but they are at least more likely to be able to boot back to anaconda and be able to, at the very least, reinstall, if not restart the upgrade and have things come out mostly okay. [2] Support for this specific case is in anaconda now From jkeating at redhat.com Wed Jan 3 22:01:10 2007 From: jkeating at redhat.com (Jesse Keating) Date: Wed, 3 Jan 2007 17:01:10 -0500 Subject: why I'm using Ubuntu instead of Fedora ATM In-Reply-To: <1167860605.7767.120.camel@metroid.rdu.redhat.com> References: <2cb10c440701030059l49323ecaqeae7c7041f735d63@mail.gmail.com> <1167859909.14822.33.camel@aglarond.local> <1167860605.7767.120.camel@metroid.rdu.redhat.com> Message-ID: <200701031701.11093.jkeating@redhat.com> On Wednesday 03 January 2007 16:43, Will Woods wrote: > Once all that's done, we can reboot into the installer kernel/initrd and > upgrade using the packages that are already sitting on the local disk. > We even know which device it's on. That part could be basically > automatic. /me puts his crazy hat on... and at that point, after the install, you could just switchroot to the new root right? No second reboot? -- Jesse Keating Release Engineer: Fedora -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 189 bytes Desc: not available URL: From sundaram at fedoraproject.org Wed Jan 3 22:03:08 2007 From: sundaram at fedoraproject.org (Rahul Sundaram) Date: Thu, 04 Jan 2007 03:33:08 +0530 Subject: why I'm using Ubuntu instead of Fedora ATM In-Reply-To: <183d138b0701031343m268e4933r119c3a94df501b6@mail.gmail.com> References: <2cb10c440701030059l49323ecaqeae7c7041f735d63@mail.gmail.com> <459B7B49.8040101@leemhuis.info> <183d138b0701030210g7127c6a0s10669ce333231cd6@mail.gmail.com> <1167834163.4869.2.camel@localhost.localdomain> <3237e4410701030649r51e1a4ccw30a89d6df6124d2d@mail.gmail.com> <183d138b0701031343m268e4933r119c3a94df501b6@mail.gmail.com> Message-ID: <459C281C.1040208@fedoraproject.org> Rob Garth wrote: > Fedora is the upstream of many projects. Some of which are considered > productions systems, RHEL and OLPC, and probably others. I am guessing > that these projects all maintian there own patches, and may indeed have > a longer support life the Fedora. Doesn't it make sense for some of the > maintainence work to be done in the upstream. I will shut up about this > now though, sorry to stir the pot. Pretty much all of these derivative systems contribute in one way or the other. Much of RHEL development is shared by Fedora but once they fork for release they diverge quickly with Fedora updates sticking closer to the upstream versions and RHEL usually getting only backported security and bug fixes. OLPC software development work on things like the sugar user interface which doesnt affect Fedora directly yet but they have helped in various fixing performance issues throughout the distribution, dependency issues etc but these derivates dont help much on extending the lifecycle of Fedora itself. Rahul From dwmw2 at infradead.org Wed Jan 3 22:09:51 2007 From: dwmw2 at infradead.org (David Woodhouse) Date: Wed, 03 Jan 2007 22:09:51 +0000 Subject: why I'm using Ubuntu instead of Fedora ATM In-Reply-To: <1167861260.14822.46.camel@aglarond.local> References: <2cb10c440701030059l49323ecaqeae7c7041f735d63@mail.gmail.com> <200701031331.02846.jkeating@redhat.com> <459BFE4D.2050901@leemhuis.info> <200701031420.39525.jkeating@redhat.com> <1167854291.7767.104.camel@metroid.rdu.redhat.com> <459C0E24.1020106@fedoraproject.org> <1167859909.14822.33.camel@aglarond.local> <1167860571.19205.31.camel@pmac.infradead.org> <1167861260.14822.46.camel@aglarond.local> Message-ID: <1167862191.19205.45.camel@pmac.infradead.org> On Wed, 2007-01-03 at 16:54 -0500, Jeremy Katz wrote: > I can generally get them to work as well. But there are weird cases > that have been dealt with over time; sure they're work-aroundable if > you're pretty saavy. It's not, though, the experience that most users > want. Live upgrades can also lead to "really interesting" states if > you, say, unplug your machine in the middle[1]. Yeah, but a normal 'yum update' does that too, when you're just fetching errata. It tends to leave two versions of various packages installed, etc. > For a quick list of "off the top of my head weird problem" examples: > * Going from static /dev -> udev involved some interesting shenanigans > to avoid all of your devices disappearing mid-upgrade. > * Newer SELinux policy depending on a kernel supporting that policy > version. > * glibc being built with a dependency on newer kernels for some new > feature > * New rpm functionality used by packages requiring a new version of rpm > to install them Yeah, some of those can be fun -- you're reminding me of things I'd subconsciously blanked from memory :) But they're not that common -- I think that if we _wanted_ a live upgrade to work, it wouldn't be particularly hard to _make_ it work. And to a large extent it _does_ work already. One of the reasons I'm happy using Fedora on servers is because I _am_ entirely confident that I can do a live upgrade on an unattended remote system. I'm no more concerned about it than I am about any kernel update on those machines. > [2] Support for this specific case is in anaconda now Is it clever enough to switch back for F7? I already fixed the kernel :) -- dwmw2 From dwmw2 at infradead.org Wed Jan 3 22:12:56 2007 From: dwmw2 at infradead.org (David Woodhouse) Date: Wed, 03 Jan 2007 22:12:56 +0000 Subject: why I'm using Ubuntu instead of Fedora ATM In-Reply-To: <459BE0F7.7050401@fedoraproject.org> References: <2cb10c440701030059l49323ecaqeae7c7041f735d63@mail.gmail.com> <459B7B49.8040101@leemhuis.info> <183d138b0701030210g7127c6a0s10669ce333231cd6@mail.gmail.com> <1167834163.4869.2.camel@localhost.localdomain> <2cb10c440701030900g379d3fcdw5b14bfd21925b2a1@mail.gmail.com> <459BE0F7.7050401@fedoraproject.org> Message-ID: <1167862376.19205.48.camel@pmac.infradead.org> On Wed, 2007-01-03 at 22:29 +0530, Rahul Sundaram wrote: > It's not. Think OLPC for example. There is also stuff out there like > Sony play stations, Cell processors etc. It already is being used as a > base for variety of stuff out there. Yellow Dog Linux being a commercial example with real non-Red-Hat folks being paid to work on it. -- dwmw2 From wwoods at redhat.com Wed Jan 3 22:15:40 2007 From: wwoods at redhat.com (Will Woods) Date: Wed, 03 Jan 2007 17:15:40 -0500 Subject: why I'm using Ubuntu instead of Fedora ATM In-Reply-To: <200701031701.11093.jkeating@redhat.com> References: <2cb10c440701030059l49323ecaqeae7c7041f735d63@mail.gmail.com> <1167859909.14822.33.camel@aglarond.local> <1167860605.7767.120.camel@metroid.rdu.redhat.com> <200701031701.11093.jkeating@redhat.com> Message-ID: <1167862540.7767.131.camel@metroid.rdu.redhat.com> On Wed, 2007-01-03 at 17:01 -0500, Jesse Keating wrote: > On Wednesday 03 January 2007 16:43, Will Woods wrote: > > Once all that's done, we can reboot into the installer kernel/initrd and > > upgrade using the packages that are already sitting on the local disk. > > We even know which device it's on. That part could be basically > > automatic. > > /me puts his crazy hat on... > > and at that point, after the install, you could just switchroot to the new > root right? No second reboot? Why stop there? Let's use kexec[1] to boot into the installer! No reboots at all! I'm SURE that's safe and well-supported.. right? Maybe? -w [1] http://fedoraproject.org/wiki/FC6KdumpKexecHowTo -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 189 bytes Desc: This is a digitally signed message part URL: From notting at redhat.com Wed Jan 3 22:16:02 2007 From: notting at redhat.com (Bill Nottingham) Date: Wed, 3 Jan 2007 17:16:02 -0500 Subject: why I'm using Ubuntu instead of Fedora ATM In-Reply-To: <1167862376.19205.48.camel@pmac.infradead.org> References: <2cb10c440701030059l49323ecaqeae7c7041f735d63@mail.gmail.com> <459B7B49.8040101@leemhuis.info> <183d138b0701030210g7127c6a0s10669ce333231cd6@mail.gmail.com> <1167834163.4869.2.camel@localhost.localdomain> <2cb10c440701030900g379d3fcdw5b14bfd21925b2a1@mail.gmail.com> <459BE0F7.7050401@fedoraproject.org> <1167862376.19205.48.camel@pmac.infradead.org> Message-ID: <20070103221602.GA26923@nostromo.devel.redhat.com> David Woodhouse (dwmw2 at infradead.org) said: > On Wed, 2007-01-03 at 22:29 +0530, Rahul Sundaram wrote: > > It's not. Think OLPC for example. There is also stuff out there like > > Sony play stations, Cell processors etc. It already is being used as a > > base for variety of stuff out there. > > Yellow Dog Linux being a commercial example with real non-Red-Hat folks > being paid to work on it. And yet, oddly, they ship Enlightement as the default desktop, but I don't see it in Extras. Bill From max at spevack.org Wed Jan 3 22:53:54 2007 From: max at spevack.org (Max Spevack) Date: Wed, 3 Jan 2007 17:53:54 -0500 (EST) Subject: why I'm using Ubuntu instead of Fedora ATM In-Reply-To: <20070103221602.GA26923@nostromo.devel.redhat.com> References: <2cb10c440701030059l49323ecaqeae7c7041f735d63@mail.gmail.com> <459B7B49.8040101@leemhuis.info> <183d138b0701030210g7127c6a0s10669ce333231cd6@mail.gmail.com> <1167834163.4869.2.camel@localhost.localdomain> <2cb10c440701030900g379d3fcdw5b14bfd21925b2a1@mail.gmail.com> <459BE0F7.7050401@fedoraproject.org> <1167862376.19205.48.camel@pmac.infradead.org> <20070103221602.GA26923@nostromo.devel.redhat.com> Message-ID: On Wed, 3 Jan 2007, Bill Nottingham wrote: > And yet, oddly, they ship Enlightement as the default desktop, but I > don't see it in Extras. My first window manager. As a Fedora user, I'd use E in a second if it was in Extras. A bit off topic, I know. But permit me the moment of nostalgia. :-) From smooge at gmail.com Wed Jan 3 23:22:10 2007 From: smooge at gmail.com (Stephen John Smoogen) Date: Wed, 3 Jan 2007 16:22:10 -0700 Subject: The QA Problem In-Reply-To: <1167853296.7767.87.camel@metroid.rdu.redhat.com> References: <2cb10c440701030059l49323ecaqeae7c7041f735d63@mail.gmail.com> <459B7B49.8040101@leemhuis.info> <20070103163903.GA22376@nostromo.devel.redhat.com> <20070103170805.GA23011@nostromo.devel.redhat.com> <1167853296.7767.87.camel@metroid.rdu.redhat.com> Message-ID: <80d7e4090701031522k21af22c0wd21512f0b4f42b40@mail.gmail.com> On 1/3/07, Will Woods wrote: > On Wed, 2007-01-03 at 13:02 -0500, Greg Dekoenigsberg wrote: > > On Wed, 3 Jan 2007, Bill Nottingham wrote: > > > > >>>> I'm wondering if we could provide solutions for both users: One update > > >>>> channel that only gets security updates and important bugfixes while the > > >>>> other is a bit more bold -- we for example could have firefox2 in the > > >>>> bold channel for FC6 while shipping the latest firefox 1.5.x in the more > > >>>> conservative channel. > > >>> > > >>> So, an idea like this: > > >>> > > >>> - starts to exponentially expand the QA problem > > >> > > >> Bill, I love you, but I'm going to have to call bullshit on this one. > > >> The answer is not "avoid making QA harder", the answer is to SOLVE THE QA > > >> PROBLEM. > > > > > > Sure, but which of these plans make more sense: > > > > > > 1) > > > > > > - Solve the QA problem for our repo configurations as they exist > > > - Expand the QA solution to new, multiple, disparate and conflicting > > > repositories > > > > > > 2) > > > > > > - Expand into new, multiple, disparate and conflicting repositories > > > - Then try to solve the QA problem > > > > > > Honestly, before we can do multiple experimental repositories of this, > > > that, and the other, we need to get our *OWN* house in order. > > > > I agree with this completely -- if, and only if, we actually make a > > concerted effort at fixing the QA problem. > > > > Which means articulating the QA problem, actually. > > Okay. "The QA Problem" is a big, fat, multi-headed monster. I mentally > group QA into four tasks, ordered by priority: > > Task #1: Testing updates to stable releases > Task #2: Testing new releases before they go out (rawhide, TestX, etc.) > Task #3: Bug triage (this is the stuff we miss in #1 and #2) > Task #4: Writing tools and docs to make the previous tasks easier > > Examining the first task - we don't currently have the *dedicated* > manpower to *guarantee* that every package will be tested and approved > by the QA group before it leaves updates-testing and goes into updates. > > I don't think we actually lack for *available* manpower. AFAIK there's > plenty of people willing and able to install packages from > updates-testing. The problems here are, I believe, a good summary of > "The QA Problem". > > Problem #1: Testing currently requires a lot of skill, which reduces > usable manpower. > > We lack how-to-test documents, so each tester must know how to set up > and test any given package/feature on his own. New features don't > necessarily come with much documentation (e.g. iSCSI). > > - Possible solution(s): More docs would lower the barrier to entry. > Setting up an official Fedora QA group will help keep track of team > strength and help everyone w uork together. > > This would be great. Knowing when you have tested enough, and covered the 'standard' way a package is meant to be used would help cut down on the trying to hit every corner case. > > Problem #3: Motivation to report bugs / testresults is low. > > Even though it's easy to install packages from updates-testing, > reporting problems with them is far harder than it needs to be. The same > old bugs get reported over and over while some new problems don't get > reported because the tester didn't want to spend 15 minutes looking for > duplicate bugs, figuring out the appropriate component/version/etc. > Or when you do take the 15 minutes you don't find the duplicate because it was described differently than you would have done so (little solution to that). I sent an intern last month to report a bug he found.. frankly bugzilla scared him when trying to pin something down.. there are 14 states that an existing bug could be in (but only really 3 maybe are used.) Putting stuff for RHEL and Fedora probably causes confusion on the RH side also. The biggest grief for me is when I find a bug and then go to search and find that there are 30+ open NEW bugs on the package all the way back from 1999 or so. This is the immediate "It is obvious that the Developer refuses to use bugzilla or does not care about this package so why post another bug that wont be looked at." I would second the point system.. and I would like to add a wall of shame. If a developer doesnt move a bug from NEW to ASSIGNED (LOOKED AT?) within a week his mug is posted to Mugshot. At the end of the quarter, their fate is voted on by members of the community (will it be the dish rack? the cushions? the COMFY CHAIR? 2 hours of a board meeting?) > - Possible solutions: The updates system should have an easy way to > report common problems with packages in updates-testing. A modified > bug-buddy for Fedora would be very helpful here. These tools should also > show the user commonly-reported bugs, and allow them to easily add a "me > too" comment. A "Bugzilla RPG" or other ranking system (like GNOME's > point system or Launchpad's Karma ranking) makes bug reporting and > triage more interesting. > > > So. Does that define The QA Problem? Or are there other issues I'm > forgetting? > > -w > > > _______________________________________________ > fedora-advisory-board mailing list > fedora-advisory-board at redhat.com > http://www.redhat.com/mailman/listinfo/fedora-advisory-board > > > > -- Stephen J Smoogen. -- CSIRT/Linux System Administrator How far that little candle throws his beams! So shines a good deed in a naughty world. = Shakespeare. "The Merchant of Venice" From skvidal at linux.duke.edu Thu Jan 4 00:20:18 2007 From: skvidal at linux.duke.edu (seth vidal) Date: Wed, 03 Jan 2007 19:20:18 -0500 Subject: bump epoch, don't roll back versions (Was Re: why I'm using Ubuntu instead of Fedora ATM) In-Reply-To: <200701031521.54012.jkeating@redhat.com> References: <2cb10c440701030059l49323ecaqeae7c7041f735d63@mail.gmail.com> <200701031504.18618.jkeating@redhat.com> <1167855276.5694.22.camel@zelda.fubar.dk> <200701031521.54012.jkeating@redhat.com> Message-ID: <1167870018.11355.2.camel@cutter> On Wed, 2007-01-03 at 15:21 -0500, Jesse Keating wrote: > On Wednesday 03 January 2007 15:14, David Zeuthen wrote: > > Sorry to sound non-constructive, but can we please stop breaking upgrade > > paths just because someone happens to think that "epochs is ugly" (left > > over packages is much less of a problem). I'd like to go even further > > and ask for our build system to enforce this rule. The justification is > > that it's only a number, and this practice is bordering introducing bugs > > by refactoring source code just because it's "less ugly" that way. > > I'll let conversation happen on the merrits of bumping or not bumping epoch. > > However for buildsystem to enforce this, that's a pretty tough nut to solve, > since any build could get tagged for any variety of collections, regardless > of nvr. In fact, the buildsystem (by design) only enforces unique n-v-r, so > you couldn't do 1:n-v-r and 2:n-v-r, the buildsystem would freak. You'd have > to do 2:n-v-r+1 or some other unique 'n-v-r'. You could, if you wanted to, starting from today have EVERY package have an epoch. The epoch is incremented by the buildsystem/cvs. So if I checkin a change and type 'make plague' it bumps the epoch so that I know that EVERY new build I have would DEFINITELY > via EVR comparisons than the one I built before. We'd have to do some tricky stuff to make sure FC6EPOCH is ALWAYS less than FC7EPOCH but that's do-able. -sv From smooge at gmail.com Thu Jan 4 00:26:41 2007 From: smooge at gmail.com (Stephen John Smoogen) Date: Wed, 3 Jan 2007 17:26:41 -0700 Subject: why I'm using Ubuntu instead of Fedora ATM In-Reply-To: <459BE3ED.6060207@leemhuis.info> References: <2cb10c440701030059l49323ecaqeae7c7041f735d63@mail.gmail.com> <459B7B49.8040101@leemhuis.info> <20070103163903.GA22376@nostromo.devel.redhat.com> <459BE3ED.6060207@leemhuis.info> Message-ID: <80d7e4090701031626l68bff836l2bd64758ff59a59@mail.gmail.com> On 1/3/07, Thorsten Leemhuis wrote: > Bill Nottingham schrieb: > > As for any sort of long-term Fedora support, what we need to see is > > some sort of market for it - we had the inital Legacy, and, realistically, > > NO ONE WANTED IT ENOUGH to actually work on it. > > Maybe it died because it/we tried to much? I think we should be able to > get enough people together to support only one distro for a longer time > at a certain period, e.g.: > > FC6 -> supported until FC8 get's out + one month = 13 months basic > support. Support FC6 after that by a new Fedora Legacy for for another > 18 months = 31 Month or round about two and a half years in total. FC11 > would be out by then and we could start maintaining FC9 for another 18 > months (it would be otherwise EOL by then)... The biggest problem was that the most Get Work Done (GWD[TM]) people were interested in a single release or two, and once that release was past.. they went elsewhere. The vast number of people/companies who asked questions were basically free-loaders. [Technically I consider myself a free-loader because I dont think I did enough to help things] They wanted to get updates for free and they either wanted the latest stuff (PHP-5 on Fedora Core 1) or they wanted to make sure it was stable (only bug fixes, NO API changes). There is nothing but inertia from stopping you from doing this.. but there is no impetus to start it because A) people who want super stable are served better by other products (Centos, Ubuntu, etc) or B) those who want the newest will stay with Fedora X. -- Stephen J Smoogen. -- CSIRT/Linux System Administrator How far that little candle throws his beams! So shines a good deed in a naughty world. = Shakespeare. "The Merchant of Venice" From mmcgrath at fedoraproject.org Thu Jan 4 00:37:37 2007 From: mmcgrath at fedoraproject.org (Mike McGrath) Date: Wed, 3 Jan 2007 18:37:37 -0600 Subject: The QA Problem In-Reply-To: <80d7e4090701031522k21af22c0wd21512f0b4f42b40@mail.gmail.com> References: <2cb10c440701030059l49323ecaqeae7c7041f735d63@mail.gmail.com> <459B7B49.8040101@leemhuis.info> <20070103163903.GA22376@nostromo.devel.redhat.com> <20070103170805.GA23011@nostromo.devel.redhat.com> <1167853296.7767.87.camel@metroid.rdu.redhat.com> <80d7e4090701031522k21af22c0wd21512f0b4f42b40@mail.gmail.com> Message-ID: <3237e4410701031637t457f11bfx46eb5dcf8fa5bf8b@mail.gmail.com> On 1/3/07, Stephen John Smoogen wrote: > On 1/3/07, Will Woods wrote: > > So. Does that define The QA Problem? Or are there other issues I'm > > forgetting? I think we should also strongly look at ways to help users solve their own issues and submit patches for them. Part of this is motivation. Sure there are many users out there who are just not competent enough to fix their problems, not everyone is a developer. We should, however, encourage the ones are able and make it as easy as possible for them to fix stuff. Possibly by having a point system or something. I mean, if I was Joe Schmo coming out of college and I was looking for a job I might mention that I have resolved 1,200 Fedora bugs, here's a website that shows my ranking, etc. -Mike From katzj at redhat.com Thu Jan 4 02:56:30 2007 From: katzj at redhat.com (Jeremy Katz) Date: Wed, 03 Jan 2007 21:56:30 -0500 Subject: why I'm using Ubuntu instead of Fedora ATM In-Reply-To: <1167862191.19205.45.camel@pmac.infradead.org> References: <2cb10c440701030059l49323ecaqeae7c7041f735d63@mail.gmail.com> <200701031331.02846.jkeating@redhat.com> <459BFE4D.2050901@leemhuis.info> <200701031420.39525.jkeating@redhat.com> <1167854291.7767.104.camel@metroid.rdu.redhat.com> <459C0E24.1020106@fedoraproject.org> <1167859909.14822.33.camel@aglarond.local> <1167860571.19205.31.camel@pmac.infradead.org> <1167861260.14822.46.camel@aglarond.local> <1167862191.19205.45.camel@pmac.infradead.org> Message-ID: <1167879390.26072.5.camel@aglarond.local> On Wed, 2007-01-03 at 22:09 +0000, David Woodhouse wrote: > On Wed, 2007-01-03 at 16:54 -0500, Jeremy Katz wrote: > > For a quick list of "off the top of my head weird problem" examples: [snip treehouse of horrors] > Yeah, some of those can be fun -- you're reminding me of things I'd > subconsciously blanked from memory :) I worry what I've subconsciously blanked ;-) > But they're not that common -- I think that if we _wanted_ a live > upgrade to work, it wouldn't be particularly hard to _make_ it work. And > to a large extent it _does_ work already. Realistically, there's on the order of one or two of these per release. And they don't tend to be the same so "making it work" isn't something that's practically worth the, in some cases quite large, effort. Hell, handling some of them even in the context of a not live upgrade can be painful enough. > > [2] Support for this specific case is in anaconda now > > Is it clever enough to switch back for F7? I already fixed the kernel :) Right now, it just unconditionally formats swap on ppc on upgrade. So perhaps it's not that it's clever enough; more that it's too dumb not to do so ;-) Jeremy From Christian.Iseli at licr.org Thu Jan 4 08:42:18 2007 From: Christian.Iseli at licr.org (Christian Iseli) Date: Thu, 4 Jan 2007 09:42:18 +0100 Subject: why I'm using Ubuntu instead of Fedora ATM In-Reply-To: References: <2cb10c440701030059l49323ecaqeae7c7041f735d63@mail.gmail.com> <459B7B49.8040101@leemhuis.info> <183d138b0701030210g7127c6a0s10669ce333231cd6@mail.gmail.com> <1167834163.4869.2.camel@localhost.localdomain> <2cb10c440701030900g379d3fcdw5b14bfd21925b2a1@mail.gmail.com> <459BE0F7.7050401@fedoraproject.org> <1167862376.19205.48.camel@pmac.infradead.org> <20070103221602.GA26923@nostromo.devel.redhat.com> Message-ID: <20070104094218.57278ce2@ludwig-alpha.unil.ch> On Wed, 3 Jan 2007 17:53:54 -0500 (EST), Max Spevack wrote: > On Wed, 3 Jan 2007, Bill Nottingham wrote: > > > And yet, oddly, they ship Enlightement as the default desktop, but I > > don't see it in Extras. > > My first window manager. As a Fedora user, I'd use E in a second if it > was in Extras. > > A bit off topic, I know. But permit me the moment of nostalgia. :-) There was an attempt to get it in Extras: https://www.redhat.com/archives/fedora-extras-list/2005-May/msg01041.html Dunno what happened... Cheers, C From fedora at leemhuis.info Thu Jan 4 08:52:53 2007 From: fedora at leemhuis.info (Thorsten Leemhuis) Date: Thu, 04 Jan 2007 09:52:53 +0100 Subject: e1[67]Re: why I'm using Ubuntu instead of Fedora ATM In-Reply-To: <20070104094218.57278ce2@ludwig-alpha.unil.ch> References: <2cb10c440701030059l49323ecaqeae7c7041f735d63@mail.gmail.com> <459B7B49.8040101@leemhuis.info> <183d138b0701030210g7127c6a0s10669ce333231cd6@mail.gmail.com> <1167834163.4869.2.camel@localhost.localdomain> <2cb10c440701030900g379d3fcdw5b14bfd21925b2a1@mail.gmail.com> <459BE0F7.7050401@fedoraproject.org> <1167862376.19205.48.camel@pmac.infradead.org> <20070103221602.GA26923@nostromo.devel.redhat.com> <20070104094218.57278ce2@ludwig-alpha.unil.ch> Message-ID: <459CC065.2010102@leemhuis.info> On 04.01.2007 09:42, Christian Iseli wrote: > On Wed, 3 Jan 2007 17:53:54 -0500 (EST), Max Spevack wrote: >> On Wed, 3 Jan 2007, Bill Nottingham wrote: >>> And yet, oddly, they ship Enlightement as the default desktop, but I >>> don't see it in Extras. >> My first window manager. As a Fedora user, I'd use E in a second if it >> was in Extras. >> A bit off topic, I know. But permit me the moment of nostalgia. :-) > There was an attempt to get it in Extras: > https://www.redhat.com/archives/fedora-extras-list/2005-May/msg01041.html > Dunno what happened... The usual problem iirc; the packages depended directly on some non-free stuff. One of the culprits was xine-lib iirc. That in Extras these days, so maybe it's possible to ship e1[67] now. BTW, I once looked at the e17 SRPM from ydl5 and ran away horrified after a few seconds... CU thl From stickster at gmail.com Thu Jan 4 13:05:47 2007 From: stickster at gmail.com (Paul W. Frields) Date: Thu, 04 Jan 2007 08:05:47 -0500 Subject: The QA Problem In-Reply-To: <1167853296.7767.87.camel@metroid.rdu.redhat.com> References: <2cb10c440701030059l49323ecaqeae7c7041f735d63@mail.gmail.com> <459B7B49.8040101@leemhuis.info> <20070103163903.GA22376@nostromo.devel.redhat.com> <20070103170805.GA23011@nostromo.devel.redhat.com> <1167853296.7767.87.camel@metroid.rdu.redhat.com> Message-ID: <1167915947.4299.7.camel@localhost.localdomain> On Wed, 2007-01-03 at 14:41 -0500, Will Woods wrote: > Problem #1: Testing currently requires a lot of skill, which reduces > usable manpower. > > We lack how-to-test documents, so each tester must know how to set up > and test any given package/feature on his own. New features don't > necessarily come with much documentation (e.g. iSCSI). > > - Possible solution(s): More docs would lower the barrier to entry. > Setting up an official Fedora QA group will help keep track of team > strength and help everyone work together. FWIW, the Docs team is willing to help with this part to the extent we can. Since we don't know your methodology, it's significantly harder to draft this ourselves. :-) Getting the stuff into maintainable, easily digestible form and publishing/advertising it, though, we can and should help do. -- Paul W. Frields, RHCE http://paul.frields.org/ gpg fingerprint: 3DA6 A0AC 6D58 FEC4 0233 5906 ACDB C937 BD11 3717 Fedora Project Board: http://fedoraproject.org/wiki/Board Fedora Docs Project: http://fedoraproject.org/wiki/DocsProject -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 189 bytes Desc: This is a digitally signed message part URL: From wwoods at redhat.com Thu Jan 4 20:57:00 2007 From: wwoods at redhat.com (Will Woods) Date: Thu, 04 Jan 2007 15:57:00 -0500 Subject: mailing-list reorganisation, round 2 In-Reply-To: <45896DDE.3080102@leemhuis.info> References: <45896DDE.3080102@leemhuis.info> Message-ID: <1167944220.7767.136.camel@metroid.rdu.redhat.com> On Wed, 2006-12-20 at 18:07 +0100, Thorsten Leemhuis wrote: > Hi All, > > second round in the process to solve the "we have to many mailinglists" > problem (first round starts at > https://www.redhat.com/archives/fedora-advisory-board/2006-December/msg00104.html > ; diff against first proposal: > http://www.fedoraproject.org/wiki/ThorstenLeemhuis/MailingListReorganization?action=diff&rev2=2&rev1=1 > ). Comments? If I some "yeah, I tend to like it" I'll post it to the > relevant mailing lists asking people to comment before we realize it. > == Rename == > > === fedora-qa-list == > > Rename fedora-triage-list to fedora-qa-list and use it for wwoods > efforts and his recruits. This is fine. I've also got #fedora-qa on freenode now, so we can merge all the -triage/testing stuff to -QA. Should make life a little easier. When can fedora-qa-list be made active? -w -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 189 bytes Desc: This is a digitally signed message part URL: From wtogami at redhat.com Fri Jan 5 01:05:59 2007 From: wtogami at redhat.com (Warren Togami) Date: Thu, 04 Jan 2007 20:05:59 -0500 Subject: why I'm using Ubuntu instead of Fedora ATM In-Reply-To: <1167834163.4869.2.camel@localhost.localdomain> References: <2cb10c440701030059l49323ecaqeae7c7041f735d63@mail.gmail.com> <459B7B49.8040101@leemhuis.info> <183d138b0701030210g7127c6a0s10669ce333231cd6@mail.gmail.com> <1167834163.4869.2.camel@localhost.localdomain> Message-ID: <459DA477.6080809@redhat.com> Michael Tiemann wrote: > On Wed, 2007-01-03 at 21:10 +1100, Rob Garth wrote: >> I am not part of the Fedora Project and so I guees my opinion is >> fairly meaningless but on the LTS support stuff. FC6 is the upstream >> of RHEL5. > > ...and so much more. It is not *just* the upstream of Red Hat's > commercial products. > Notable is the small and ever shrinking % of Fedora that becomes RHEL. There is something to this.... the scope of Fedora is much larger than RHEL. Warren Togami wtogami at redhat.com From wtogami at redhat.com Fri Jan 5 01:15:17 2007 From: wtogami at redhat.com (Warren Togami) Date: Thu, 04 Jan 2007 20:15:17 -0500 Subject: why I'm using Ubuntu instead of Fedora ATM In-Reply-To: <459BE586.9080509@leemhuis.info> References: <2cb10c440701030059l49323ecaqeae7c7041f735d63@mail.gmail.com> <459B7B49.8040101@leemhuis.info> <2cb10c440701030858g4a0c84cdi3cc1f67186b3258f@mail.gmail.com> <459BE586.9080509@leemhuis.info> Message-ID: <459DA6A5.7070203@redhat.com> Thorsten Leemhuis wrote: > Luis Villa schrieb: >> On 1/3/07, Thorsten Leemhuis wrote: >>> Thx Luis, I liked you mail. Some comments from my side: >>> On 03.01.2007 09:59, Luis Villa wrote: >>>> * QA: Ubuntu aggressively pushes people to use their development >>>> branch and report problems, which leads to better, more stable final >>>> releases. [...] >>> Agreed. The biggest problem we have in this regard IMHO is that we >>> always communicate "you can't get to stable release from rawhide or a >>> test release". That scares people aways from the devel branch and the >>> test releases. We should provide a clean solution so people at least can >>> get from test3 to stable. >> No, you should provide a clean path from every package ver X to every >> package ver X+1. Period. [...] > > Actually I'd like to see this, too, but Fedora people seem to be so > unwilling to support something like that so I tried the "better ask for > something that's not that hard to realize and lies between the current > solution and the best/ideal solution; maybe then the chances to get what > we ask for are higher and it make one big part of the problem go away, > even if parts of the problem remain"-approach ;-) > I've been annoyed with Red Hat's prior stance where mistakes (like "pre" within version or bad epochs) that happened during rawhide are rolled back silently. This makes it impossible for folks using rawhide to continue upgrading in a smooth and automated fashion. In the past two years we've been doing a lot less of this, but it still happens. We really have to stop this entirely. Warren Togami wtogami at redhat.com From wtogami at redhat.com Fri Jan 5 01:18:24 2007 From: wtogami at redhat.com (Warren Togami) Date: Thu, 04 Jan 2007 20:18:24 -0500 Subject: why I'm using Ubuntu instead of Fedora ATM In-Reply-To: <200701031231.09906.jkeating@redhat.com> References: <2cb10c440701030059l49323ecaqeae7c7041f735d63@mail.gmail.com> <2cb10c440701030858g4a0c84cdi3cc1f67186b3258f@mail.gmail.com> <459BE586.9080509@leemhuis.info> <200701031231.09906.jkeating@redhat.com> Message-ID: <459DA760.3010200@redhat.com> Jesse Keating wrote: > On Wednesday 03 January 2007 12:19, Thorsten Leemhuis wrote: >> Actually I'd like to see this, too, but Fedora people seem to be so >> unwilling to support something like that so I tried the "better ask for >> something that's not that hard to realize and lies between the current >> solution and the best/ideal solution; maybe then the chances to get what >> we ask for are higher and it make one big part of the problem go away, >> even if parts of the problem remain"-approach > > Well, we could do this, but it would involve an explosion of epochs :/ Our > ability to roll back package versions that were 'mistakes' is very very > useful. The only way to keep this is to use epochs to "force" a lower > version package to win. Once epoch, always epoch. Epoch is a necessary evil, when the alternative is people lose the ability to automatically upgrade and be useful to your testing efforts. It has been a bigger mistake for us to rollback these "mistakes" in this way. Warren Togami wtogami at redhat.com From kwade at redhat.com Fri Jan 5 03:09:09 2007 From: kwade at redhat.com (Karsten Wade) Date: Thu, 04 Jan 2007 19:09:09 -0800 Subject: The QA Problem In-Reply-To: <1167915947.4299.7.camel@localhost.localdomain> References: <2cb10c440701030059l49323ecaqeae7c7041f735d63@mail.gmail.com> <459B7B49.8040101@leemhuis.info> <20070103163903.GA22376@nostromo.devel.redhat.com> <20070103170805.GA23011@nostromo.devel.redhat.com> <1167853296.7767.87.camel@metroid.rdu.redhat.com> <1167915947.4299.7.camel@localhost.localdomain> Message-ID: <1167966549.25662.497.camel@erato.phig.org> On Thu, 2007-01-04 at 08:05 -0500, Paul W. Frields wrote: > On Wed, 2007-01-03 at 14:41 -0500, Will Woods wrote: > > Problem #1: Testing currently requires a lot of skill, which reduces > > usable manpower. > > > > We lack how-to-test documents, so each tester must know how to set up > > and test any given package/feature on his own. New features don't > > necessarily come with much documentation (e.g. iSCSI). > > > > - Possible solution(s): More docs would lower the barrier to entry. > > Setting up an official Fedora QA group will help keep track of team > > strength and help everyone work together. > > FWIW, the Docs team is willing to help with this part to the extent we > can. Since we don't know your methodology, it's significantly harder to > draft this ourselves. :-) Getting the stuff into maintainable, easily > digestible form and publishing/advertising it, though, we can and should > help do. +1 The size and capability of the Documentation Project has grown a bit recently. A call to arms on a high priority item such as supporting QA documentation efforts is sure to raise some good resources. Otherwise, we have the tools, processes, localization, and editorial help you'll need. - Karsten -- Karsten Wade, RHCE, 108 Editor ^ Fedora Documentation Project Sr. Developer Relations Mgr. | fedoraproject.org/wiki/DocsProject quaid.108.redhat.com | gpg key: AD0E0C41 ////////////////////////////////// \\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\ -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 189 bytes Desc: This is a digitally signed message part URL: From deisenst at gtw.net Fri Jan 5 07:06:36 2007 From: deisenst at gtw.net (David Eisenstein) Date: Fri, 05 Jan 2007 01:06:36 -0600 Subject: Legacy's Success; Re: why I'm using Ubuntu instead of Fedora ATM In-Reply-To: <2cb10c440701030858g4a0c84cdi3cc1f67186b3258f@mail.gmail.com> References: <2cb10c440701030059l49323ecaqeae7c7041f735d63@mail.gmail.com> <459B7B49.8040101@leemhuis.info> <2cb10c440701030858g4a0c84cdi3cc1f67186b3258f@mail.gmail.com> Message-ID: <459DF8FC.4020805@gtw.net> So many people now seem to want to point to Fedora Legacy and use it as a Free-Software Whipping Boy to use its eventual demise as some kind of example of failure to point to. For example, Luis Villa wrote: > On 1/3/07, Thorsten Leemhuis wrote: >> Hi! >> A Fedora LTS (two years? maybe the server parts ever three?) now and >> then (every second or third release?) from a new Fedora Legacy (needs a >> different name) would IMHO a nice solution. > > > > (FWIW, I think it is unreasaonble to expect a true-community distro to > do real LTS-y stuff- most volunteers don't have the patience to do the > necessary backporting for the necessary length of time. (See Fedora > Legacy.) As a long-time contributor to and advocate for the Fedora Legacy Project, I have to say that, over most of its life, Legacy did not fail its mission, if one were to consider Legacy's mission to provide security updates to packages that people really cared about. Why? Because it was those packages that folks cared about either (a) that squeaked the wheel on the project's email list or (b) that motivated people to dig in and get themselves dirty doing onerous, boring, but important work for the community of Legacy users. For the longest time, I personally cared about Fedora Core 1, and also cared about the old Red Hat Linux releases 7.3 and 9.0. The project cared too. Fedora Core 1 came out in Fall of 2003, and was essentially supported until May or June of this year -- which is a lifetime of two-and-a-half years -- covering security updates for those packages that the folks who volunteered wanted or that users squawked loudly for (like sendmail, glibc, mozilla, and others). And what about Red Hat Linux 7.3 and 9? Even longer! For these three releases, and also perhaps FC2, this project was more successful than perhaps the founders of Fedora Legacy had hoped or dreamed it would be. A lot of the work towards the end of the useful life of Fedora Legacy was done by one man: Marc Deslauriers, to which all Fedora Legacy users owe a LOT (and I mean a *LOT*) of thank-you's! He was the one builder brave enough to go in and do kernel security updates for the (at one time) FIVE Linux releases that Legacy was supporting; and for many other packages, Marc did much or most the work of the steps we had in place to assure sanity, quality, and security in the creation of updated (backported) packages for our end-users. Thank you from the bottom of my heart, Marc!!! Your example is one we should all be committed enough to follow and emulate! And what were Marc and the other contributors paid for this often onerous work? Not one penny. Often we were paid more complaints than compliments. It became utterly too thankless of a task (and too little interest from the community in even doing the QA work we had outlined in our documentation) for me to continue, and probably the same goes for Marc. I believe the few who did most of the work finally burned out. There are still people who want to help out and don't know where to begin to help to keep some kind of Legacy alive for the releases they care about. Is this failure? My assessment is this: If legacy failed it did so in these areas: * Management of contributor resources * Devotion of people who knew how to motivate and cause people in the contributing community to feel valued, motivated and special, and to give a voice to those who cared. Legacy rarely had meetings, had no board to speak of, and therefore no clear mechanism of accountability. I hope the good folks of Legacy remember Legacy *not* as a failed experiment, but as one that lasted longer and did better than folks had any right to expect. Warm regards, David Eisenstein From wwoods at redhat.com Fri Jan 5 15:57:55 2007 From: wwoods at redhat.com (Will Woods) Date: Fri, 05 Jan 2007 10:57:55 -0500 Subject: The QA Problem In-Reply-To: <1167966549.25662.497.camel@erato.phig.org> References: <2cb10c440701030059l49323ecaqeae7c7041f735d63@mail.gmail.com> <459B7B49.8040101@leemhuis.info> <20070103163903.GA22376@nostromo.devel.redhat.com> <20070103170805.GA23011@nostromo.devel.redhat.com> <1167853296.7767.87.camel@metroid.rdu.redhat.com> <1167915947.4299.7.camel@localhost.localdomain> <1167966549.25662.497.camel@erato.phig.org> Message-ID: <1168012675.7767.148.camel@metroid.rdu.redhat.com> On Thu, 2007-01-04 at 19:09 -0800, Karsten Wade wrote: > On Thu, 2007-01-04 at 08:05 -0500, Paul W. Frields wrote: > > On Wed, 2007-01-03 at 14:41 -0500, Will Woods wrote: > > > Problem #1: Testing currently requires a lot of skill, which reduces > > > usable manpower. > > > > > > We lack how-to-test documents, so each tester must know how to set up > > > and test any given package/feature on his own. New features don't > > > necessarily come with much documentation (e.g. iSCSI). > > > > > > - Possible solution(s): More docs would lower the barrier to entry. > > > Setting up an official Fedora QA group will help keep track of team > > > strength and help everyone work together. > > > > FWIW, the Docs team is willing to help with this part to the extent we > > can. Since we don't know your methodology, it's significantly harder to > > draft this ourselves. :-) Getting the stuff into maintainable, easily > > digestible form and publishing/advertising it, though, we can and should > > help do. > > +1 > > The size and capability of the Documentation Project has grown a bit > recently. A call to arms on a high priority item such as supporting QA > documentation efforts is sure to raise some good resources. Otherwise, > we have the tools, processes, localization, and editorial help you'll > need. This is quite exciting! One big item is writing a general Tester HOWTO that details the processes of tracking down bugs and filing really good bug reports. Another idea I had was to start writing per-package docs for testing things in updates-testing. Those docs would use the template laid out here: http://fedoraproject.org/wiki/QA/HowToTestTemplate Similarly, when we get closer to Test1, we would need to have per-feature docs for how to test New Stuff that goes into F7. (I don't have a template for that yet.) I would *love* to have help beautifying (and clarifying) the templates, making the Tester HOWTO available in different formats and languages, and I'd also love to hear any ideas you guys might have for tester's docs. Did you have anything in mind, or was that just a general offer for help? -w -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 189 bytes Desc: This is a digitally signed message part URL: From jkeating at redhat.com Fri Jan 5 16:15:43 2007 From: jkeating at redhat.com (Jesse Keating) Date: Fri, 5 Jan 2007 11:15:43 -0500 Subject: The QA Problem In-Reply-To: <1168012675.7767.148.camel@metroid.rdu.redhat.com> References: <2cb10c440701030059l49323ecaqeae7c7041f735d63@mail.gmail.com> <1167966549.25662.497.camel@erato.phig.org> <1168012675.7767.148.camel@metroid.rdu.redhat.com> Message-ID: <200701051115.43429.jkeating@redhat.com> On Friday 05 January 2007 10:57, Will Woods wrote: > Another idea I had was to start writing per-package docs for testing > things in updates-testing. Those docs would use the template laid out > here: http://fedoraproject.org/wiki/QA/HowToTestTemplate > Similarly, when we get closer to Test1, we would need to have > per-feature docs for how to test New Stuff that goes into F7. (I don't > have a template for that yet.) > > I would *love* to have help beautifying (and clarifying) the templates, > making the Tester HOWTO available in different formats and languages, > and I'd also love to hear any ideas you guys might have for tester's > docs. One thing that has been suggested before is that each package gets a wiki page, or some other sort of page, that has in formation regarding who maintains it, where is the upstream, Summary/Description all that, and also it could have information about how to use/test that package. I certainly think much of this info should be automatically generated out of the package database, and some of it user contributed. How to get there, I have no idea. -- Jesse Keating Release Engineer: Fedora -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 189 bytes Desc: not available URL: From tbm at cyrius.com Fri Jan 5 16:29:41 2007 From: tbm at cyrius.com (Martin Michlmayr) Date: Fri, 5 Jan 2007 17:29:41 +0100 Subject: The QA Problem In-Reply-To: <200701051115.43429.jkeating@redhat.com> References: <2cb10c440701030059l49323ecaqeae7c7041f735d63@mail.gmail.com> <1167966549.25662.497.camel@erato.phig.org> <1168012675.7767.148.camel@metroid.rdu.redhat.com> <200701051115.43429.jkeating@redhat.com> Message-ID: <20070105162941.GT15109@deprecation.cyrius.com> * Jesse Keating [2007-01-05 11:15]: > One thing that has been suggested before is that each package gets a > wiki page, or some other sort of page, that has in formation > regarding who maintains it, where is the upstream, > Summary/Description all that, and also it could have information > about how to use/test that package. FWIW, Debian has two sites which provide such information. Maybe they're a good source of ideas for Fedora. An overview of packages maintained by a package: http://qa.debian.org/developer.php?login=tbm%40cyrius.com&comaint=yes Some information about an individual package: http://packages.qa.debian.org/c/colo.html -- Martin Michlmayr http://www.cyrius.com/ From fedora at leemhuis.info Fri Jan 5 17:00:38 2007 From: fedora at leemhuis.info (Thorsten Leemhuis) Date: Fri, 05 Jan 2007 18:00:38 +0100 Subject: The QA Problem In-Reply-To: <200701051115.43429.jkeating@redhat.com> References: <2cb10c440701030059l49323ecaqeae7c7041f735d63@mail.gmail.com> <1167966549.25662.497.camel@erato.phig.org> <1168012675.7767.148.camel@metroid.rdu.redhat.com> <200701051115.43429.jkeating@redhat.com> Message-ID: <459E8436.5060202@leemhuis.info> Jesse Keating schrieb: > On Friday 05 January 2007 10:57, Will Woods wrote: > One thing that has been suggested before is that each package gets a wiki > page, or some other sort of page, that has in formation regarding who > maintains it, where is the upstream, Summary/Description all that, and also > it could have information about how to use/test that package. > > I certainly think much of this info should be automatically generated out of > the package database, and some of it user contributed. How to get there, I > have no idea. My preferred solution: let some script/tool generate static pages with a url that does not change, e.g. fedoraproject.org/packages/foo/ Some informations for those pages could come from the package database, and some others from the SRPM (maybe a slightly enhanced or modified version of repoview could do that). When that scripts runs let it check for the existence of fedoraproject.org/wiki/Packages/foo and integrate those parts and/or simply link to that page. The maintainer or interested users could then maintain some further informations (important bugs, known problems, ...) in the wiki easily. Cu thl From wwoods at redhat.com Fri Jan 5 17:08:04 2007 From: wwoods at redhat.com (Will Woods) Date: Fri, 05 Jan 2007 12:08:04 -0500 Subject: The QA Problem In-Reply-To: <459E8436.5060202@leemhuis.info> References: <2cb10c440701030059l49323ecaqeae7c7041f735d63@mail.gmail.com> <1167966549.25662.497.camel@erato.phig.org> <1168012675.7767.148.camel@metroid.rdu.redhat.com> <200701051115.43429.jkeating@redhat.com> <459E8436.5060202@leemhuis.info> Message-ID: <1168016884.7767.156.camel@metroid.rdu.redhat.com> On Fri, 2007-01-05 at 18:00 +0100, Thorsten Leemhuis wrote: > Jesse Keating schrieb: > > On Friday 05 January 2007 10:57, Will Woods wrote: > > One thing that has been suggested before is that each package gets a wiki > > page, or some other sort of page, that has in formation regarding who > > maintains it, where is the upstream, Summary/Description all that, and also > > it could have information about how to use/test that package. > > > > I certainly think much of this info should be automatically generated out of > > the package database, and some of it user contributed. How to get there, I > > have no idea. > > My preferred solution: let some script/tool generate static pages with a > url that does not change, e.g. > > fedoraproject.org/packages/foo/ > > Some informations for those pages could come from the package database, > and some others from the SRPM (maybe a slightly enhanced or modified > version of repoview could do that). > > When that scripts runs let it check for the existence of > fedoraproject.org/wiki/Packages/foo > and integrate those parts and/or simply link to that page. The > maintainer or interested users could then maintain some further > informations (important bugs, known problems, ...) in the wiki easily. Right, my working theory was that there'd be a site (say qa.fedoraproject.org) to hold static test results and autogenerated info for each package and such, with links to the package page on the wiki. I've been meaning to talk to the infrastructure folks about getting the virtual host set up and a couple shell accounts for the automated tools to use.. -w -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 189 bytes Desc: This is a digitally signed message part URL: From kwade at redhat.com Fri Jan 5 17:15:09 2007 From: kwade at redhat.com (Karsten Wade) Date: Fri, 05 Jan 2007 09:15:09 -0800 Subject: The QA Problem In-Reply-To: <1168012675.7767.148.camel@metroid.rdu.redhat.com> References: <2cb10c440701030059l49323ecaqeae7c7041f735d63@mail.gmail.com> <459B7B49.8040101@leemhuis.info> <20070103163903.GA22376@nostromo.devel.redhat.com> <20070103170805.GA23011@nostromo.devel.redhat.com> <1167853296.7767.87.camel@metroid.rdu.redhat.com> <1167915947.4299.7.camel@localhost.localdomain> <1167966549.25662.497.camel@erato.phig.org> <1168012675.7767.148.camel@metroid.rdu.redhat.com> Message-ID: <1168017310.25662.532.camel@erato.phig.org> On Fri, 2007-01-05 at 10:57 -0500, Will Woods wrote: > One big item is writing a general Tester HOWTO that details the > processes of tracking down bugs and filing really good bug reports. > > Another idea I had was to start writing per-package docs for testing > things in updates-testing. Those docs would use the template laid out > here: http://fedoraproject.org/wiki/QA/HowToTestTemplate > Similarly, when we get closer to Test1, we would need to have > per-feature docs for how to test New Stuff that goes into F7. (I don't > have a template for that yet.) > > I would *love* to have help beautifying (and clarifying) the templates, > making the Tester HOWTO available in different formats and languages, > and I'd also love to hear any ideas you guys might have for tester's > docs. > > Did you have anything in mind, or was that just a general offer for > help? :) We can move this over to fedora-docs-list for hashing out, but here is a quick overview so others are aware and can follow along there, if they wish: * Start by touching one or more Wiki templates; make sure they use the markup[1] and writing styles[2] from the start that support modularity. * Discuss what we would like a larger "Testing Guide" to look like; what parts should be in it, etc. - This includes deciding if it is sub-project-specific documentation (so should live in the QA/ Wiki namespace) or is more general Fedora documentation (so should start in Docs/Drafts/ and move to Docs/). - Break the document down to modules that one or a few people can quickly write, edit, and complete; easy success == more likely to come back for more abuse^H writing * Create a skeleton in a draft location[3]. * Find people to own each of the modules; cf. the release notes beat process[4] * When we have a complete document in the Wiki, we release it with these steps[5]: 1. Copy to a canonical location (e.g. Docs/TestingGuide) - Then content development could continue in the e.g. Docs/Drafts/TestingGuide location, or in CVS (see next step) 2. Convert the Wiki content to DocBook XML 2.1 Create module in CVS (/cvs/docs/ or /cvs/qa/) 2.2 Assign bugzilla module 3. From this CVS module we create PO files for translation 3.1 Upload into L10n CVS (or whatever the current process is) 4. Publish original and translations as links from Docs/ [1] http://fedoraproject.org/wiki/WikiEditing#Marking_Technical_Terms [2] http://fedoraproject.org/wiki/DocsProject/WritingUsingTheWiki [3] for ex. http://fedoraproject.org/wiki/Docs/Drafts/TestingGuide [4] http://fedoraproject.org/wiki/DocsProject/ReleaseNotes/Beats [5] http://fedoraproject.org/wiki/DocsProject/WorkFlow -- Karsten Wade, RHCE, 108 Editor ^ Fedora Documentation Project Sr. Developer Relations Mgr. | fedoraproject.org/wiki/DocsProject quaid.108.redhat.com | gpg key: AD0E0C41 ////////////////////////////////// \\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\ -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 189 bytes Desc: This is a digitally signed message part URL: From fedora at leemhuis.info Fri Jan 5 20:48:56 2007 From: fedora at leemhuis.info (Thorsten Leemhuis) Date: Fri, 05 Jan 2007 21:48:56 +0100 Subject: mailing-list reorganisation, round 3 on this list In-Reply-To: <45896DDE.3080102@leemhuis.info> References: <45896DDE.3080102@leemhuis.info> Message-ID: <459EB9B8.90005@leemhuis.info> Hi again! Thorsten Leemhuis schrieb: > second round in the process to solve the "we have to many mailinglists" > problem (first round starts at > https://www.redhat.com/archives/fedora-advisory-board/2006-December/msg00104.html > ; diff against first proposal: > http://www.fedoraproject.org/wiki/ThorstenLeemhuis/MailingListReorganization?action=diff&rev2=2&rev1=1 > ). Okay, here we go with round 3: http://www.fedoraproject.org/wiki/ThorstenLeemhuis/MailingListReorganization Diff to previous: http://www.fedoraproject.org/wiki/ThorstenLeemhuis/MailingListReorganization?action=diff&rev2=5&rev1=2 Sorry, the whole think got a bit longer. I'll send this out to some lists for public discussions early next week. I'll try to incorporate feedback afterwards. The I should be able to trim the list down again a bit. = Mailing list reorganization = A lot of people yelled "there are to many mailing lists; that is contra-productive and leads to a lot of confusion" (this is the short explanation without all the boring details). Thus the idea came up to revisit the Fedora mailing lists that got created after Fedora was born (that's years ago now!) and do some adjustments -- shut some down mainly. That's needed in parts for the Core and Extras merge in any case. The Fedora Board asked me, ThorstenLeemhuis (thl), to drive this task. So here we go, comments appreciated. ---- == Shut down == Meaning: Send a EOL message with some informations to the list and disallow further postings. Achieves remain accessible. === fedora-packaging-list === Packaging is important to all package maintainers. Everyone should get involved and see the discussions and at least see the topics being discussed, thus do it directly on fedora-devel (e.g. in the public where everybody can post). Use a special tag like "[packaging]" to mark the mails if there really is a need to differentiate. === fedora-extras-list === Well, the "Core packages gets merged into the Extras framework" is official -- thus discuss the stuff that was on fedora-extras-list until now on fedora-devel in the future, as there won#t be a "Extras" anymore soon. === fedora-test-list === A lot of people don't get the difference between fedora-devel and fedora-test list. And testing is a crucial part of the devel process, thus lets drop the test-list. === fedora-triage-list === Not used much anymore; fedora-qa-list will take over the job (see below) === fedora-extras-commits === No extras anymore; send the stuff to fedora-cvs-commits in the future. ---- == Create == We need some new lists, too. === fedora-qa-list === Rename fedora-triage-list to fedora-qa-list and use it for wwoods efforts and his recruits. Will be created quite soon, probably before this discussion is finished, because its badly needed. === fedora-project-list === We until now have no real list where Ambassadors, packagers, programmers, art-people and other contributors can talk to each other about general stuff that's important to the project as a whole without getting lost in the noise or scared away with "this is off topic on this list" calls. fedora-advisory-board somehow was this list until now in parts, but it's moderated and thus even some project contributors that were not subscribed feel excluded (bad). === fedora-maintainers-announce === Created already, but not uses until now. This is not a real mailing list for discussion, it's rather meant as a way to get really important information out to all maintainers -- e.g. low traffic, normally less then 3 mails a months. The plan is to subscribe users semi-automatically from the accounts system. Some people question the use of this list; they want people subscribed to fedora-maintainers instead. But important informations can get lost there in the noise easily, and that would be bad. And some of our maintainers (e.g. upstream maintainers that co-maintain the fedora package) are not interested in all those discussions that happened on fedora-maintainers in the past. Reply-to of this list will be set to fedora-maintainers. No monthly mailinglist-reminder. fedora-maintainers gets subscribed to fedora-maintainers-announce == Change policy == === fedora-advisory-board === The list remains for now, but we want to move most of the discussion into the border public. fedora-project should thus be used for most of the stuff that got discussed on fedora-advisory-board in the past. But the list will remain for the purpose it was created for: a communication channel between important project contributors. We'll close it later if we notice that this is not needed anymore (that what some people hope) ---- == Not sure == There are some list where I'm not sure if we still need them and/or need to adjust their policy. === fedora-desktop-list === Now traffic. Do we really need it? === fedora-games-list === Does it really make sense to have a mailing list for this particular group of packages? === fedora-devel-foo === There are several; some of them lists === fedora-maintainers === Open fedora-maintainers to all project contributors -- e.g. those from other projects like arts, docs, infrastructure, ambassadors, ...? Or will fedora-project serve this purpose? Some people requested a way to have a mailing list where only packagers discuss stuff. But is it really a big help? It leads to fragmentation again as it might easily happen that we discuss stuff on fedora-maintainers in a semi-official way that would be more suitable for fedora-devel *because* everyone can participate there. Comments? === fedora-list === This is mostly a list where users help he other. fedora-users-list or fedora-help-list would thus be a proper name that would make its use obvious. But renaming this list is probably not worth the trouble. === fedora-devel-help === A lot of developers get annoyed if users ask on the devel list for help (even if the questions are specific to the devel-tree. Most users of the devel tree on the other hand probably are not much interested in the discussions on fedora-list and won#t find help there. Maybe try to separate those questions to their own list? We would need to enforce that (e.g. never answer those question on fedora-devel). === fedora-cvs-commits === Could we use that for all commits and seperate the stuff (web, docs, F6, devel, ...) with channels? Similar to how it is done for fedora-packaging-announce? === fedora-desktop-list === Not much traffic. Do we still need it? === fedora-r-devel-list === Not much traffic. Do we still need it? ---- == Suggestions from the discussions == * rsc wants to get the daily developments reports, but does not what to subscribe to fedora-devel; can we create another channel in fedora-package-announce that mails the reports in addition to fedora-devel? == Miscellaneous == === Mailman guidelines === While at it let's define a standard look-and-feel for mailing lists, too. (FIXME: some mailman experts around? Is this sane?). Mailing list admins are strongly encouraged to follow this guidelines, but they don't have to if there are good reason for it (no, "I don't like it" is not a good reason!). Suggested setting (can all be found on the first mailman settings page) * the reply-to should not be modified by mailman (first_strip_reply_to = no) * the reply-to should point to the list (reply_goes_to_list = this list) * no explicit Reply-to-Address (reply_to_address = ) * no tagging (subject_prefix = ) == Legacy == Jesse will close fedora-legacy and fedora-legacy-annouce in the near future. === Notes === Ubutu has similar problems; they splitted ubuntu-devel into ubuntu-devel and ubuntu-devel-discuss: https://lists.ubuntu.com/archives/ubuntu-devel/2006-December/023022.html We have a similar concept with fedora-devel and fedora-maintainers already. EOF CU thl From wtogami at redhat.com Sat Jan 6 19:37:21 2007 From: wtogami at redhat.com (Warren Togami) Date: Sat, 06 Jan 2007 14:37:21 -0500 Subject: mailing-list reorganisation, round 3 on this list In-Reply-To: <459EB9B8.90005@leemhuis.info> References: <45896DDE.3080102@leemhuis.info> <459EB9B8.90005@leemhuis.info> Message-ID: <459FFA71.9040808@redhat.com> Thorsten Leemhuis wrote: > > == Shut down == > > Meaning: Send a EOL message with some informations to the list and > disallow further postings. Achieves remain accessible. > > === fedora-packaging-list === > > Packaging is important to all package maintainers. Everyone should get > involved and see the discussions and at least see the topics being > discussed, thus do it directly on fedora-devel (e.g. in the public where > everybody can post). Use a special tag like "[packaging]" to mark the > mails if there really is a need to differentiate. How was the signal to noise ratio on this list? If it remained on-topic without noise, then it may be effective to keep it. Mixing it into fedora-devel-list might make discussion on this topic more difficult due to all the noise. > > === fedora-extras-list === > > Well, the "Core packages gets merged into the Extras framework" is > official -- thus discuss the stuff that was on fedora-extras-list until > now on fedora-devel in the future, as there won#t be a "Extras" anymore > soon. fedora-extras-list historically had a pretty good signal to noise ratio. It will be heavily diluted if merged. But yes, logically it makes no sense to keep it separate anymore given Fedora 7. Design of the policy and enforcement of "development discussion only" will need to be the top priority if fedora-devel-list and fedora-extras-list are merged. > > === fedora-test-list === > > A lot of people don't get the difference between fedora-devel and > fedora-test list. And testing is a crucial part of the devel process, > thus lets drop the test-list. However... we don't want to redirect all of the end-users test discussion to fedora-devel-list. fedora-devel-list + fedora-extras-list would be pretty heavy even if it stuck to only development related discussions. I advise strongly against doing this. Please leave this one alone for now. > > === fedora-maintainers-announce === > > Created already, but not uses until now. This is not a real mailing list > for discussion, it's rather meant as a way to get really important > information out to all maintainers -- e.g. low traffic, normally less > then 3 mails a months. The plan is to subscribe users semi-automatically > from the accounts system. > > Some people question the use of this list; they want people subscribed > to fedora-maintainers instead. But important informations can get lost > there in the noise easily, and that would be bad. And some of our > maintainers (e.g. upstream maintainers that co-maintain the fedora > package) are not interested in all those discussions that happened on > fedora-maintainers in the past. > > Reply-to of this list will be set to fedora-maintainers. No monthly > mailinglist-reminder. fedora-maintainers gets subscribed to > fedora-maintainers-announce FESCO made a decision to use it. Just do it. Allow people a way to opt-out if they follow fedora-maintainers, or make sure there aren't duplicates (parent/child list), and things should be fine. > > === fedora-desktop-list === > > > Now traffic. Do we really need it? Attitudes about how it is run and where leadership & direction must come up need to change in order for this list to become useful. > > === fedora-games-list === > > Does it really make sense to have a mailing list for this particular > group of packages? Is it active? Are they doing useful work? If so, let them do it. Otherwise ask them if they think it should be shut down. > > === fedora-maintainers === > > Open fedora-maintainers to all project contributors -- e.g. those from > other projects like arts, docs, infrastructure, ambassadors, ...? Or > will fedora-project serve this purpose? > > Some people requested a way to have a mailing list where only packagers > discuss stuff. But is it really a big help? It leads to fragmentation > again as it might easily happen that we discuss stuff on > fedora-maintainers in a semi-official way that would be more suitable > for fedora-devel *because* everyone can participate there. Comments? If fedora-devel-list managed to keep 100% on-topic without end-user complaints, it might be manageable to use exclusively that list. But the signal/noise ratio there is now very bad even without fedora-extras-list. > > === fedora-list === > > This is mostly a list where users help he other. fedora-users-list or > fedora-help-list would thus be a proper name that would make its use > obvious. > > But renaming this list is probably not worth the trouble. This list has long ago been so heavy and full of crap that its usefulness is poor to end-users. A new user learning Linux can be easily scared by the huge deluge of mail. We would be better off redirecting people to a better end-user mutual support site like fedoraforum.org. The web board medium is much better suited to most end users. > > === fedora-devel-help === > > A lot of developers get annoyed if users ask on the devel list for help > (even if the questions are specific to the devel-tree. Most users of the > devel tree on the other hand probably are not much interested in the > discussions on fedora-list and won#t find help there. > > Maybe try to separate those questions to their own list? We would need > to enforce that (e.g. never answer those question on fedora-devel). Will developers pay attention? > > === fedora-cvs-commits === > > Could we use that for all commits and seperate the stuff (web, docs, F6, > devel, ...) with channels? Similar to how it is done for > fedora-packaging-announce? This should be easy. > == Suggestions from the discussions == > > * rsc wants to get the daily developments reports, but does not what to > subscribe to fedora-devel; can we create another channel in > fedora-package-announce that mails the reports in addition to fedora-devel? Daily development reports should be mailed to individuals if it requires their personal attention. Otherwise it should be something individuals can subscribe to separately if they don't want to follow the big & noisy list. Warren Togami wtogami at redhat.com From tibbs at math.uh.edu Sat Jan 6 19:53:22 2007 From: tibbs at math.uh.edu (Jason L Tibbitts III) Date: 06 Jan 2007 13:53:22 -0600 Subject: mailing-list reorganisation, round 3 on this list In-Reply-To: <459FFA71.9040808@redhat.com> References: <45896DDE.3080102@leemhuis.info> <459EB9B8.90005@leemhuis.info> <459FFA71.9040808@redhat.com> Message-ID: >>>>> "WT" == Warren Togami writes: [Regarding fedora-packaging] WT> How was the signal to noise ratio on this list? fedora-packaging has always had a very high signal to noise ratio, since it's used primarily for Packaging Committee discussions. If we lose it and it becomes difficult to perform committee business due to additional discussion on fedora-devel, there's a good chance that someone will just set up a private list elsewhere. Many PC members are short on time as it is. I believe that if you need a tag like "[packaging]" to differentiate traffic that the packaging committee needs to follow, then you've crammed too much stuff onto one list. - J< From Axel.Thimm at ATrpms.net Sat Jan 6 20:09:42 2007 From: Axel.Thimm at ATrpms.net (Axel Thimm) Date: Sat, 6 Jan 2007 21:09:42 +0100 Subject: mailing-list reorganisation, round 3 on this list In-Reply-To: References: <45896DDE.3080102@leemhuis.info> <459EB9B8.90005@leemhuis.info> <459FFA71.9040808@redhat.com> Message-ID: <20070106200942.GA14558@neu.nirvana> On Sat, Jan 06, 2007 at 01:53:22PM -0600, Jason L Tibbitts III wrote: > >>>>> "WT" == Warren Togami writes: > > [Regarding fedora-packaging] > WT> How was the signal to noise ratio on this list? > > fedora-packaging has always had a very high signal to noise ratio, > since it's used primarily for Packaging Committee discussions. If we > lose it and it becomes difficult to perform committee business due to > additional discussion on fedora-devel, there's a good chance that > someone will just set up a private list elsewhere. Many PC members > are short on time as it is. > > I believe that if you need a tag like "[packaging]" to differentiate > traffic that the packaging committee needs to follow, then you've > crammed too much stuff onto one list. I'd second keeping fedora-packaging as is. The charter is discussing about packaging, not (specific) packages. But Thorsten has a point: All packagers need to know what happens over there - same is true for some other lists, too. So I suggest to have a developer/packager "announce" list, where various boards, sigs and so on can *tersly* communicate *results*, e.g. important board/fesco decisions, packaging guidelines changes, infrastructure/buildsystem changes etc., but keep the discussions in separate lists as happens now. fedora-devel-news? But before continuing on thinking about mailing lists structure, perhaps one should wait to see what the upcoming organisational structures will look like to start mapping them to lists, or to deduce some list structure out of it. I guess the most important and communicative organs do need to keep dedicated lists for a good SNR. -- Axel.Thimm at ATrpms.net -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 189 bytes Desc: not available URL: From fedora at leemhuis.info Sun Jan 7 10:15:19 2007 From: fedora at leemhuis.info (Thorsten Leemhuis) Date: Sun, 07 Jan 2007 11:15:19 +0100 Subject: mailing-list reorganisation, round 4 on this list In-Reply-To: <459EB9B8.90005@leemhuis.info> References: <45896DDE.3080102@leemhuis.info> <459EB9B8.90005@leemhuis.info> Message-ID: <45A0C837.7010601@leemhuis.info> Hi! Me again with the next round! Thorsten Leemhuis schrieb: > Thorsten Leemhuis schrieb: >> second round in the process to solve the "we have to many mailinglists" >> problem (first round starts at >> https://www.redhat.com/archives/fedora-advisory-board/2006-December/msg00104.html >> ; diff against first proposal: >> http://www.fedoraproject.org/wiki/ThorstenLeemhuis/MailingListReorganization?action=diff&rev2=2&rev1=1 >> ). > http://www.fedoraproject.org/wiki/ThorstenLeemhuis/MailingListReorganization > Diff to previous: > http://www.fedoraproject.org/wiki/ThorstenLeemhuis/MailingListReorganization?action=diff&rev2=5&rev1=2 And the next round -- I got a bit more audacious with this one and propose to get rid of fedora-maintainers again. I know some people won't like that, but I think it's the best. Diff to last round: http://www.fedoraproject.org/wiki/ThorstenLeemhuis/MailingListReorganization?action=diff&rev2=6&rev1=5 Some replies to the mails from warren, jason and axel can be found at the end. ######################################################################## [[TableOfContents]] = Mailing list reorganization = A lot of people yelled "there are to many mailing lists; that is contra-productive and leads to a lot of confusion" (this is the short explanation without all the boring details). Thus the idea came up to revisit the Fedora mailing lists that got created after Fedora was born (that's years ago now!) and do some adjustments -- shut some down or adjust there usage model mainly. That's needed in parts for the Core and Extras merge in any case. Site note: this is a complex task and we can't make everybody happy -- but we nevertheless need to adjust at least something with the new Core and Extras merged world. The Fedora Board asked me, ThorstenLeemhuis (thl), to drive this task. You can find my proposal below, comments appreciated. Feel free to write a counter proposal if you think I did stuff wrong or if you think I ignored your comments ( I read them all, but as I wrote: we can't make everyone happy -- but I'm trying my best to find a compromise thats better than what we have today and is acceptable for most of us). ---- == Shut down == Meaning: Send a EOL message with some informations to the list and disallow further postings. Achieves remain accessible. === fedora-extras-list === Well, the "Core packages gets merged into the Extras framework" is official -- thus discuss most of the stuff that was on fedora-extras-list until now on fedora-packaging in the future, as there won't be a "Extras" anymore soon. Some stuff might be suitable for fedora-devel-list, too === fedora-test-list === A lot of people don't get the difference between fedora-devel and fedora-test list and ask users support questions on fedora-devel. One of the reasons for this is that it's not obvious what "testing" means. Another reasons: we answered those users question on fedora-devel, even if they were not on topic (and thus users asked again and again). And the topics for -devel and -test are to similar; sometimes the same things get discussed in parallel on both lists. === fedora-triage-list === Not used much anymore; fedora-qa-list will take over the job (see below) === fedora-extras-commits === No extras anymore; send the stuff to fedora-cvs-commits in the future. (see below) ---- == Create == We need some new lists, too. === fedora-qa-list === Rename fedora-triage-list to fedora-qa-list and use it for wwoods efforts and his recruits. Will be created quite soon, probably before this discussion is finished, because its badly needed. === fedora-project-list === We until now have no real list where Ambassadors, packagers, programmers, art-people and other contributors can talk to each other about general stuff that's important to the project as a whole without getting lost in the noise or scared away with "this is off topic on this list" calls. fedora-advisory-board somehow was this list until now in parts, but it's moderated and thus even some project contributors that were not subscribed feel excluded (bad). Summaries from the project meetings (Ambassadors, Infrastructure, FESCo successor) shall get send to the list so people know what's going on in the project as a whole. === fedora-maintainers-announce === Created already, but not uses until now. This is not a real mailing list for discussion, it's rather meant as a way to get really important information (policy changes, announcements for mass rebuilds, ...) out to all maintainers -- e.g. low traffic, normally less then 3-5 mails a months. The plan is to subscribe users semi-automatically from the accounts system to make sure we reach everyone. Some people question the use of this list; they want people subscribed to fedora-maintainers instead. But important informations can get lost there in the noise easily, and that would be bad. And some of our maintainers (e.g. upstream maintainers that co-maintain the fedora package) are not interested in all those discussions that happened on fedora-maintainers in the past. Reply-to of this list will be set to fedora-devel; that list will also get subscribed to fedora-maintainers-announce. Users can't unsubscribe manually from fedora-maintainers-announce list -- being on it is mandatory for all package contributors. == Change policy == === fedora-advisory-board === The list remains for now, but we want to move most of the discussion into the border public. fedora-project should thus be used for most of the stuff that got discussed on fedora-advisory-board in the past. But the list will remain for the purpose it was created for: a communication channel between important project contributors. We'll close it later if we notice that this is not needed anymore (that what some people hope) === fedora-packaging-list === Until it was used for the Packaging Committee discussions mainly. But at the same time the packaging committee sometimes missed things that were brought up on fedora-extras-list; and cross-posting was sometimes necessary, too, and created confusion. Thus with the dropping of fedora-extras-list let's move its discussions here, to cover both packaging in practice and the guidelines for it on one list. === fedora-cvs-commits === All commits should go here, but separated into channels (web, docs, F7, devel, ...), similar how we do it for how we do for fedora-packaging-announce ---- == Rename == === fedora-devel-list === Rename to fedora-devel-users. It shall to a major part replace fedora-test-list. Users can help each other here (and developers should keep any eye on there, too) if they have problems with running development or testing-channels. They can also discuss Fedora development there freely. === fedora-maintainers === Rename to fedora-devel. Free to post for all fedora contributors (being in ambassadors, mentors, packagers, ...). Moderated for the rest of the world (we need a whole team of moderators for it). Be liberal and allow good-guys to post without being moderated, even if they are no fedora-contributors; that includes: * people that are well know and active in open-source land somehow * people who work on upstream apps we package * people that have shown to be constructive after their posts got moderated for a while Enforce the devel topics and never ever answer support questions on that list -> those shall are on topic for fedora-devel-users. ---- == Not sure == There are some list where I'm not sure if we still need them and/or need to adjust their policy. === fedora-games-list === Does it really make sense to have a mailing list for this particular group of packages? The group itself shall please question that. IMHO their stuff should be discussed on fedora-packaging; they can use a [games] tag if they want to mark their stuff. === fedora-list === This is mostly a list where users help he other. fedora-users would thus be a proper name that would make its use obvious. === fedora-desktop-list === Not much traffic. Either close or rename to fedora-devel-desktop, to make its usage obvious. === fedora-r-devel-list === Not much traffic. Do we still need it? The group itself shall please question the need for this list. ---- == Suggestions from the discussions == == Miscellaneous == === Mailman guidelines === While at it let's define a standard look-and-feel for mailing lists, too. (FIXME: some mailman experts around? Is this sane?). Mailing list admins are strongly encouraged to follow this guidelines, but they don't have to if there are good reason for it (no, "I don't like it" is not a good reason!). Suggested setting (can all be found on the first mailman settings page) * the reply-to should not be modified by mailman (first_strip_reply_to = no) * the reply-to should point to the list (reply_goes_to_list = this list) * no explicit Reply-to-Address (reply_to_address = ) * no tagging (subject_prefix = ) === Legacy === Jesse will close fedora-legacy and fedora-legacy-announce in the near future. === To -list or not to list === Some of our mailing lists have the "-list" postfix, other not. I think all new lists should be created without it. === rawhide reports === We create another channel in fedora-package-announce where those kind of reports get mailed to. Don't send them to any other lists. Side note: A channel for the important packages might be helpful, because the reports will get quite long in the future after the extras and core merge. How about a channel that lists only changes in packages that are part of our most important spins Gnome, KDE and Server? == Notes == Ubuntu has similar problems; they splitted ubuntu-devel into ubuntu-devel and ubuntu-devel-discuss: https://lists.ubuntu.com/archives/ubuntu-devel/2006-December/023022.html . ######################################################################## Warren Togami wrote: > Thorsten Leemhuis wrote: >> > === fedora-test-list === >> > >> > A lot of people don't get the difference between fedora-devel and >> > fedora-test list. And testing is a crucial part of the devel process, >> > thus lets drop the test-list. > However... we don't want to redirect all of the end-users test > discussion to fedora-devel-list. fedora-devel-list + fedora-extras-list > would be pretty heavy even if it stuck to only development related > discussions. > > I advise strongly against doing this. Please leave this one alone for now. And I think the -devel-list and test-list confusion really needs to be sorted out. I'd IMHO one of the biggest problem we have. Maybe my new model with fedora-devel-users matches your expectations better. >> > === fedora-desktop-list === >> > >> > Now traffic. Do we really need it? > Attitudes about how it is run and where leadership & direction must come > up need to change in order for this list to become useful. Agreed, I adjusted the text slightly. This is low priority IMHO. >> > === fedora-games-list === >> > Does it really make sense to have a mailing list for this particular >> > group of packages? > Is it active? Are they doing useful work? If so, let them do it. > Otherwise ask them if they think it should be shut down. Text adjusted as well. Members of the list should comment. >> > === fedora-maintainers === >> > Open fedora-maintainers to all project contributors -- e.g. those from >> > other projects like arts, docs, infrastructure, ambassadors, ...? Or >> > will fedora-project serve this purpose? >> > >> > Some people requested a way to have a mailing list where only packagers >> > discuss stuff. But is it really a big help? It leads to fragmentation >> > again as it might easily happen that we discuss stuff on >> > fedora-maintainers in a semi-official way that would be more suitable >> > for fedora-devel *because* everyone can participate there. Comments? > If fedora-devel-list managed to keep 100% on-topic without end-user > complaints, it might be manageable to use exclusively that list. Agreed, because I think it's otherwise hightly confusing what gets discussed on what list. > But > the signal/noise ratio there is now very bad even without > fedora-extras-list. See proposal. >> > === fedora-list === >> > This is mostly a list where users help he other. fedora-users-list or >> > fedora-help-list would thus be a proper name that would make its use >> > obvious. >> > But renaming this list is probably not worth the trouble. > > This list has long ago been so heavy and full of crap that its > usefulness is poor to end-users. A new user learning Linux can be > easily scared by the huge deluge of mail. Well, nevertheless it seems there are still a lot of people active there :) > We would be better off redirecting people to a better end-user mutual > support site like fedoraforum.org. The web board medium is much better > suited to most end users. There are people that want to use a mailing-list -- I for example hate web-forums and would strongly prefer a list. >> > === fedora-devel-help === >> > A lot of developers get annoyed if users ask on the devel list for help >> > (even if the questions are specific to the devel-tree. Most users of the >> > devel tree on the other hand probably are not much interested in the >> > discussions on fedora-list and won#t find help there. >> > >> > Maybe try to separate those questions to their own list? We would need >> > to enforce that (e.g. never answer those question on fedora-devel). > Will developers pay attention? We can't force that in any case. Some will, and a lot of users will probably help each other. That should suffice IMHO [...] Axel Thimm wrote: > On Sat, Jan 06, 2007 at 01:53:22PM -0600, Jason L Tibbitts III wrote: >>>>>>> > >>>>> "WT" == Warren Togami writes: >> > [Regarding fedora-packaging] >> > WT> How was the signal to noise ratio on this list? >> > >> > fedora-packaging has always had a very high signal to noise ratio, >> > since it's used primarily for Packaging Committee discussions. If we >> > lose it and it becomes difficult to perform committee business due to >> > additional discussion on fedora-devel, there's a good chance that >> > someone will just set up a private list elsewhere. Many PC members >> > are short on time as it is. >> > >> > I believe that if you need a tag like "[packaging]" to differentiate >> > traffic that the packaging committee needs to follow, then you've >> > crammed too much stuff onto one list. > I'd second keeping fedora-packaging as is. The charter is discussing > about packaging, not (specific) packages. > But Thorsten has a point: All packagers need to know what happens over > there - same is true for some other lists, too. Axel, tibbs, does the new idea (fedora-extras traffic mostly goes to fedora-packaging to cover both packaging in practice and the guidelines on one list) suite your needs better? > So I suggest to have a developer/packager "announce" list, where > various boards, sigs and so on can *tersly* communicate *results*, > e.g. important board/fesco decisions, packaging guidelines changes, > infrastructure/buildsystem changes etc., but keep the discussions in > separate lists as happens now. fedora-devel-news? packaging guidelines changes -> fedora-maintainers-annouce All the other stuff IMHO should be okay for fedora-devel. > But before continuing on thinking about mailing lists structure, > perhaps one should wait to see what the upcoming organisational > structures will look like [...] Yeah, might be a good idea. CU thl From fedora at leemhuis.info Sun Jan 7 11:07:05 2007 From: fedora at leemhuis.info (Thorsten Leemhuis) Date: Sun, 07 Jan 2007 12:07:05 +0100 Subject: how to govern and manage the new combined repository Message-ID: <45A0D459.1050807@leemhuis.info> Hi all! Quoting Bill from: https://www.redhat.com/archives/fedora-devel-list/2007-January/msg00091.html > What's left of the Core Steering Committee > is going to work with the Fedora Board and FESCO to figure out just > how this new combined repository is going to be governed and managed. Yeah, that becomes more an more urgent afaics. The idea was raised to sort this out in a special IRC-meeting and/or conference call with people from the Core Cabal, FESCo and the Board. That's probably a good idea, but that needs a bit of preparation, too. Thus (and because a board members ask me to) I sat down and will roughly outline my thoughts on the issue with this mail. I don't want to start yet another endless discussion with this mail/proposal (I know, I'm good with that, but I'm not proud of it) -- rather I'd like to see similar mails/proposals from other people where they describe how this whole thing should be governed. Then we can mix the best ideas into each others scheme and then discuss one or two proposals further to find the final one eventually. Find attached a small image that roughly outlines how I think the organization structure should look like. Description: The Fedora Board sits on the top (hey, that was easy!). It's job is to govern the Fedora as a whole; as such it has to coordinate the different projects like Ambassadors, Infrastructure, the merged Core and Extras repo and other sub-projects I just hinted in the drawing. The Board should further set the big long-term strategies and thus has to decide or at least ACK important things that the sub-projects worked out. The Board should normally *not* have to handle technical decisions. That IMHO should be the job of the FESCo/Core Cabal successor. We need a name for it -- Jesse Suggested "FTC - Fedora Technical Commitee". I like that in general and can live with it, but it's IMHO way to easily confused with the Federal Trade Commission ( http://www.ftc.gov ). Maybe someone else has a better idea. I'll nevertheless use the term now for the rest of this mail. The FTC will have a lot to do. Thus I think it should only work out the really the big decisions -- the roadmap for example (such big things would need a ACK from the Board). The other work should be handed over to sub-groups, that are each responsible for parts of the whole picture and report to the FTC periodically so it can coordinate the different groups. >From the top of my head and with help from Jesse (he told me about some groups that currently exist internally at Fedora / Red Hat) I came up with this rough list of groups we could have: * Packaging Guidelines Group -- The Packaging Committee (create the guidelines for packaging and reviewing, but don't enforce those rules) * Packages/Repo Group -- a kind of FESCo successor that actually takes care of maintaining the repo, reviewing new packages and enforcing what the Packaging Committee decided. (?); has sub-groups (missing in the drawing) in the form of SIGs (special interest groups) that take care of specific type of packages (games, python for example) or archs (PPC and x86_64, ...) * QA -- wwoods and his recruits (also (?)) * Desktop * base OS (kernel, rpm, initscripts) * core tools (gcc, glibc, ...) * server software (apache, bind, samba, sendmail ...) * others like ET (emerging Technologies like xen and kvm), cluster stuff, selinux, java, ... (needs to be discussed what we really need) * Release Team -- a kind of core cabal successor that create the distributions spins. This group might need sub-groups again for each spin in the long term: * GNOME * KDE * Server * more to come over time Note: there are interdependencies between those groups sometimes (Desktop Group and Gnome/KDE Spin; Server-Apps and Server Spin; Packaging Guidelines and Packages/Repo Group, which itself need to coordinate with the the Release Team), but that should be big problem afaics. Sure, some people will say "that's way to complicated/to many groups" -- but if you look closer you'll notice that all of those groups (expect the FTC and those for the Distributions spins, witch are more a long-term idea) exist in some form already today. Each of those group should have its own Steering Committee with 3 to 10 people (7 normally). It should normally be a balanced mix of community members and red hat employees -- especially the FTC (?). Having a committee with mostly red hat employees is fine if those do all the work. But even in that case the committees are strongly urged to make sure the community feels heard, respected and integrated -- thus it should have at least some community members and public meetings. The desktop-group is probably an area where that did not work in the past (see the recent discussions on fedora-devel). In other words: "Fedora is a meritocracy" -- that should be valid for the committees outlined, too. I'm not yet sure how to best make sure we reach that goal when it comes to the constitution of those committees. Maybe a mix of elected and appointed (the FTC could appoint them or the groups could try to do it them selfs) members? Further: The committee's should work similar to the scheme that's was used by FESCo up to know. To be more precise: The committee's must meet at least every two weeks (bigger groups weekly) in the public (IRC normally); additional meetings in private (e.g. release engineers sit down on a table or in the pub and work out how to plan the next big thing) are acceptable, but the stuff that got decided should at least be revisited once in the public. Community contributors must be able to participate in the (bi-)weekly meetings, and all the important stuff should also be discussed on the mailinglists, too, to give people that can't or don't like to participate in the meetings have a chance to influence the decisions to make sure they feel heard. A schedule needs to be maintained in the wiki; the meeting planing and the topics that will come up should be send out to a public mailing list 24 hours before the meeting. A summary has to be send to the list *and* to the FTC after the meeting. The above rules are of course valid for the FTC itself, too; it will collect the reports, discuss issues if needed and will send a trimmed report to the Board. Closing words: Yes, the things described in the last two paras (?) and the organization structure is a bit more overhead then we had in the past -- especially for red hat folks. But remember: that stuff gives the community a chance to get involved -- and if we get the community involved properly it will be in the end a better product for everyone and less work for each of us as the load gets spread over more shoulders. Thanks for reading. CU thl (?) -- FIXME -- the security team either needs to become a separate group or gets under the hood of the Package/Repo Group or the QA group (?) -- I'd prefer if we could have a defined 50% community and 50% red hat mix for the FTC (at least in the beginning), as it will be a crucial committee for the whole game. But it seems I'm the only one that want something like this :-/ (?) -- We should probably sooner or later define a "how does a ideal committee work" template, to avoid that each of the groups has to work out the generic rules on its own. FESCo and it's structure in parts should act as a starting point; https://www.redhat.com/archives/fedora-advisory-board/2006-November/msg00258.html has some ideas, too. -------------- next part -------------- A non-text attachment was scrubbed... Name: fesco-successor.png Type: image/png Size: 25028 bytes Desc: not available URL: From Axel.Thimm at ATrpms.net Sun Jan 7 12:59:46 2007 From: Axel.Thimm at ATrpms.net (Axel Thimm) Date: Sun, 7 Jan 2007 13:59:46 +0100 Subject: mailing-list reorganisation, round 4 on this list In-Reply-To: <45A0C837.7010601@leemhuis.info> References: <45896DDE.3080102@leemhuis.info> <459EB9B8.90005@leemhuis.info> <45A0C837.7010601@leemhuis.info> Message-ID: <20070107125946.GA26973@neu.nirvana> On Sun, Jan 07, 2007 at 11:15:19AM +0100, Thorsten Leemhuis wrote: > Axel Thimm wrote: > > On Sat, Jan 06, 2007 at 01:53:22PM -0600, Jason L Tibbitts III wrote: > >>>>>>> > >>>>> "WT" == Warren Togami writes: > >> > [Regarding fedora-packaging] > >> > WT> How was the signal to noise ratio on this list? > >> > > >> > fedora-packaging has always had a very high signal to noise ratio, > >> > since it's used primarily for Packaging Committee discussions. If we > >> > lose it and it becomes difficult to perform committee business due to > >> > additional discussion on fedora-devel, there's a good chance that > >> > someone will just set up a private list elsewhere. Many PC members > >> > are short on time as it is. > > I'd second keeping fedora-packaging as is. The charter is discussing > > about packaging, not (specific) packages. > > But Thorsten has a point: All packagers need to know what happens over > > there - same is true for some other lists, too. > > Axel, tibbs, does the new idea (fedora-extras traffic mostly goes to > fedora-packaging to cover both packaging in practice and the guidelines > on one list) suite your needs better? On the contrary, what tibbs and I didn't like is that committee discussions get mangled with additional discussion. Your previous suggestion had this list merge with fedora-devel, now it's fedora-extras, in both cases you get very different content killing the other's SNR. I still think most organizational bodies need a list primary for their daily work. That is not to say that other people should not subscribe and discuss there, too, but the topic should be defined in that way. I liked the way it was until now: Packagers would consult each other on fedora-extras or other lists and if some issue escalated for the packaging committee to look at it would do so. That's why we have such a good SNR. Please keep fedora-packaging as is - it is one of the split off lists that served its purpose rather well IMHO. -- Axel.Thimm at ATrpms.net -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 189 bytes Desc: not available URL: From fedora at leemhuis.info Sun Jan 7 13:28:12 2007 From: fedora at leemhuis.info (Thorsten Leemhuis) Date: Sun, 07 Jan 2007 14:28:12 +0100 Subject: mailing-list reorganisation, round 4 on this list In-Reply-To: <20070107125946.GA26973@neu.nirvana> References: <45896DDE.3080102@leemhuis.info> <459EB9B8.90005@leemhuis.info> <45A0C837.7010601@leemhuis.info> <20070107125946.GA26973@neu.nirvana> Message-ID: <45A0F56C.8000501@leemhuis.info> Axel Thimm schrieb: > On Sun, Jan 07, 2007 at 11:15:19AM +0100, Thorsten Leemhuis wrote: >> Axel Thimm wrote: >>> On Sat, Jan 06, 2007 at 01:53:22PM -0600, Jason L Tibbitts III wrote: >>>>>>>>>>>>>>> "WT" == Warren Togami writes: >>>>> [Regarding fedora-packaging] >>>>> WT> How was the signal to noise ratio on this list? >>>>> >>>>> fedora-packaging has always had a very high signal to noise ratio, >>>>> since it's used primarily for Packaging Committee discussions. If we >>>>> lose it and it becomes difficult to perform committee business due to >>>>> additional discussion on fedora-devel, there's a good chance that >>>>> someone will just set up a private list elsewhere. Many PC members >>>>> are short on time as it is. > >>> I'd second keeping fedora-packaging as is. The charter is discussing >>> about packaging, not (specific) packages. >>> But Thorsten has a point: All packagers need to know what happens over >>> there - same is true for some other lists, too. >> Axel, tibbs, does the new idea (fedora-extras traffic mostly goes to >> fedora-packaging to cover both packaging in practice and the guidelines >> on one list) suite your needs better? > > On the contrary, what tibbs and I didn't like is that committee > discussions get mangled with additional discussion. Your previous > suggestion had this list merge with fedora-devel, now it's > fedora-extras, in both cases you get very different content killing > the other's SNR. > > I still think most organizational bodies need a list primary for their > daily work. > > That is not to say that other people should not subscribe > and discuss there, too, but the topic should be defined in that way. > > I liked the way it was until now: Packagers would consult each other > on fedora-extras or other lists and if some issue escalated for the > packaging committee to look at it would do so. That's why we have such > a good SNR. > > Please keep fedora-packaging as is - it is one of the split off lists > that served its purpose rather well IMHO. Well, I on the contrary heard complain that the split made everything way more complicated and that we have to many lists. And I want the packagers involved in the discussions about packaging -- a separate list makes that only more complicated. So I'll leave it there until I hear more complains (if I hear those I'll propose to leave fedora-packaging and will suggest fedora-packages or fedora-collective as alternative, but then this whole idea "let's get rid of some mailing list" becomes more and more a mailing list shuffle). I would strongly dislike to differentiate to much between theoretical (guidelines) and practical (packages in the repo that have to follow those guidelines) packaging. /me thinks a bit more about it some more Hmm, I send my "how to govern and manage the new combined repository" out already and proposed two groups (one that takes care of the repo and one that takes care of the guidelines) there. I now think that was wrong. The packaging committee was created because FESCo had a lot to do already and we needed the guidelines for both Core and Extras. That's history now. So why not have one Committee again that handles both theoretical and practical packaging. That's less overhead. CU thl From Axel.Thimm at ATrpms.net Sun Jan 7 13:52:38 2007 From: Axel.Thimm at ATrpms.net (Axel Thimm) Date: Sun, 7 Jan 2007 14:52:38 +0100 Subject: mailing-list reorganisation, round 4 on this list In-Reply-To: <45A0F56C.8000501@leemhuis.info> References: <45896DDE.3080102@leemhuis.info> <459EB9B8.90005@leemhuis.info> <45A0C837.7010601@leemhuis.info> <20070107125946.GA26973@neu.nirvana> <45A0F56C.8000501@leemhuis.info> Message-ID: <20070107135238.GD26973@neu.nirvana> On Sun, Jan 07, 2007 at 02:28:12PM +0100, Thorsten Leemhuis wrote: > Axel Thimm schrieb: > > On Sun, Jan 07, 2007 at 11:15:19AM +0100, Thorsten Leemhuis wrote: > >> Axel Thimm wrote: > >>> On Sat, Jan 06, 2007 at 01:53:22PM -0600, Jason L Tibbitts III wrote: > >>>>>>>>>>>>>>> "WT" == Warren Togami writes: > >>>>> [Regarding fedora-packaging] > >>>>> WT> How was the signal to noise ratio on this list? > >>>>> > >>>>> fedora-packaging has always had a very high signal to noise ratio, > >>>>> since it's used primarily for Packaging Committee discussions. If we > >>>>> lose it and it becomes difficult to perform committee business due to > >>>>> additional discussion on fedora-devel, there's a good chance that > >>>>> someone will just set up a private list elsewhere. Many PC members > >>>>> are short on time as it is. > > > >>> I'd second keeping fedora-packaging as is. The charter is discussing > >>> about packaging, not (specific) packages. > >>> But Thorsten has a point: All packagers need to know what happens over > >>> there - same is true for some other lists, too. > >> Axel, tibbs, does the new idea (fedora-extras traffic mostly goes to > >> fedora-packaging to cover both packaging in practice and the guidelines > >> on one list) suite your needs better? > > > > On the contrary, what tibbs and I didn't like is that committee > > discussions get mangled with additional discussion. Your previous > > suggestion had this list merge with fedora-devel, now it's > > fedora-extras, in both cases you get very different content killing > > the other's SNR. > > > > I still think most organizational bodies need a list primary for their > > daily work. > > > > That is not to say that other people should not subscribe > > and discuss there, too, but the topic should be defined in that way. > > > > I liked the way it was until now: Packagers would consult each other > > on fedora-extras or other lists and if some issue escalated for the > > packaging committee to look at it would do so. That's why we have such > > a good SNR. > > > > Please keep fedora-packaging as is - it is one of the split off lists > > that served its purpose rather well IMHO. > > Well, I on the contrary heard complain that the split made everything > way more complicated and that we have to many lists. Where there complaints about fedora-packaging in particular? Could the people with these complaints perhaps speak up and detail them? If there are some issues I'd rather deal differently than merging all lists into a list soup. To be honest, I'm not really aware that there were any such issue (at least compared to other list fragmentation), but maybe that's part of the problem (although I'm on almost every list and should have picked up any larger dissatisfaction about fedora-packaging). > So I'll leave it there until I hear more complains You mean, you will leave the list as is and wait for more complaints about merging it into something else, or that you will leave sour suggestion to merge it into something and wait for counter-complaints. In the later case you already have two. > Hmm, I send my "how to govern and manage the new combined > repository" out already and proposed two groups (one that takes care > of the repo and one that takes care of the guidelines) there. I now > think that was wrong. > > The packaging committee was created because FESCo had a lot to do > already and we needed the guidelines for both Core and > Extras. That's history now. So why not have one Committee again that > handles both theoretical and practical packaging. That's less > overhead. Even if the FPC were dissolved and the resposibilities moved back to higher level organizational entities you still have different topics. Discussing packaging guidelines and repo structure has very little in common, what is important for one group is the noise for the other. You have a dillema: On the one hand you want to off-load work from fesco/cabal or any successing entity at their places, but on the other you are thinking of merging the groups back together, which may lead to everyone being unsatisfied with the SNR as their perceive it. -- Axel.Thimm at ATrpms.net -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 189 bytes Desc: not available URL: From fedora at leemhuis.info Sun Jan 7 14:15:34 2007 From: fedora at leemhuis.info (Thorsten Leemhuis) Date: Sun, 07 Jan 2007 15:15:34 +0100 Subject: mailing-list reorganisation, round 4 on this list In-Reply-To: <20070107135238.GD26973@neu.nirvana> References: <45896DDE.3080102@leemhuis.info> <459EB9B8.90005@leemhuis.info> <45A0C837.7010601@leemhuis.info> <20070107125946.GA26973@neu.nirvana> <45A0F56C.8000501@leemhuis.info> <20070107135238.GD26973@neu.nirvana> Message-ID: <45A10086.8080903@leemhuis.info> Axel Thimm schrieb: >> So I'll leave it there until I hear more complains > You mean, you will leave the list as is [...] No, I meant "I don't change my proposal" for now. CU thl From lmacken at redhat.com Sun Jan 7 14:28:30 2007 From: lmacken at redhat.com (Luke Macken) Date: Sun, 7 Jan 2007 09:28:30 -0500 Subject: how to govern and manage the new combined repository In-Reply-To: <45A0D459.1050807@leemhuis.info> References: <45A0D459.1050807@leemhuis.info> Message-ID: <20070107142830.GA16958@tomservo.rh.rit.edu> On Sun, Jan 07, 2007 at 12:07:05PM +0100, Thorsten Leemhuis wrote: > * QA -- wwoods and his recruits (also (?)) [...] > (?) -- FIXME -- the security team either needs to become a separate > group or gets under the hood of the Package/Repo Group or the QA group separate_group++; The only reason I've heard in the past for not wanting to actively pursue a Fedora Security Team was because if Fedora is pushing out security updates faster than RHEL, it makes RHEL look bad to its customers. This is not a good enough reason IMO. At the moment, all security updates require approval from the Red Hat Security Response Team. With the new update system, this should most likely change, and it would be nice to have our own security team to do the approvals. With the proper infrastructure, and overlapping security teams, coordinating security fixes between the distros can be made trivial. Some good reasons off the top of my head to have our own security team: - allow Fedora to be an active contributor in the security world. How many liaisons do we have on vendor-sec? How many hackers do we have auditing our packages and infrastructure? Probably none in both cases -- which should change. - provide useful security advisories as opposed to an RPM changelog and whatever notes the overworked developer feels like writing up. Every other distro out there sends security advisories to bugtraq and such. We push out plenty of security updates already, why not make it known? - help make sure security issues are fixed timely, by staying on top of the devs and cracking the whip. luke From notting at redhat.com Mon Jan 8 19:29:33 2007 From: notting at redhat.com (Bill Nottingham) Date: Mon, 8 Jan 2007 14:29:33 -0500 Subject: mailing-list reorganisation, round 3 on this list In-Reply-To: <20070106200942.GA14558@neu.nirvana> References: <45896DDE.3080102@leemhuis.info> <459EB9B8.90005@leemhuis.info> <459FFA71.9040808@redhat.com> <20070106200942.GA14558@neu.nirvana> Message-ID: <20070108192933.GD20857@nostromo.devel.redhat.com> Axel Thimm (Axel.Thimm at ATrpms.net) said: > But Thorsten has a point: All packagers need to know what happens over > there - same is true for some other lists, too. > > So I suggest to have a developer/packager "announce" list, where > various boards, sigs and so on can *tersly* communicate *results*, > e.g. important board/fesco decisions, packaging guidelines changes, > infrastructure/buildsystem changes etc., but keep the discussions in > separate lists as happens now. fedora-devel-news? > > But before continuing on thinking about mailing lists structure, > perhaps one should wait to see what the upcoming organisational > structures will look like to start mapping them to lists, or to deduce > some list structure out of it. I guess the most important and > communicative organs do need to keep dedicated lists for a good SNR. Makes sense to me. Bill From notting at redhat.com Mon Jan 8 19:33:24 2007 From: notting at redhat.com (Bill Nottingham) Date: Mon, 8 Jan 2007 14:33:24 -0500 Subject: mailing-list reorganisation, round 4 on this list In-Reply-To: <45A0C837.7010601@leemhuis.info> References: <45896DDE.3080102@leemhuis.info> <459EB9B8.90005@leemhuis.info> <45A0C837.7010601@leemhuis.info> Message-ID: <20070108193324.GE20857@nostromo.devel.redhat.com> Thorsten Leemhuis (fedora at leemhuis.info) said: > And the next round -- I got a bit more audacious with this one and > propose to get rid of fedora-maintainers again. I know some people won't > like that, but I think it's the best. OK, so I'd like to start off by saying: - Renaming is bad and should be avoided at all costs. If we can fix things without having to rename or move people from list to list, it will work much better. That being said: > === fedora-maintainers-announce === > > Created already, but not uses until now. This is not a real mailing list > for discussion, it's rather meant as a way to get really important > information (policy changes, announcements for mass rebuilds, ...) out > to all maintainers -- e.g. low traffic, normally less then 3-5 mails a > months. The plan is to subscribe users semi-automatically from the > accounts system to make sure we reach everyone. > > Some people question the use of this list; they want people subscribed > to fedora-maintainers instead. But important informations can get lost > there in the noise easily, and that would be bad. And some of our > maintainers (e.g. upstream maintainers that co-maintain the fedora > package) are not interested in all those discussions that happened on > fedora-maintainers in the past. > > Reply-to of this list will be set to fedora-devel; that list will also > get subscribed to fedora-maintainers-announce. Users can't unsubscribe > manually from fedora-maintainers-announce list -- being on it is > mandatory for all package contributors. If you want to do this, repurpose fedora-maintainers, move fedora-maintainers discussion to fedora-devel list. However, I think something like Axel's suggestion of a list where things like packaging committee *results*, policy changes, etc. would fit in here, but it would be more than 3-5 mails a month. > === fedora-devel-list === > > Rename to fedora-devel-users. It shall to a major part replace > fedora-test-list. Users can help each other here (and developers should > keep any eye on there, too) if they have problems with running > development or testing-channels. They can also discuss Fedora > development there freely. I don't see the point here. If you want a test list for test releases, or users testing, *use fedora-test-list*. Don't do pointless renaming. > === fedora-maintainers === > > Rename to fedora-devel. Free to post for all fedora contributors (being > in ambassadors, mentors, packagers, ...). Moderated for the rest of the > world (we need a whole team of moderators for it). Be liberal and allow > good-guys to post without being moderated, even if they are no > fedora-contributors; that includes: > > * people that are well know and active in open-source land somehow > > * people who work on upstream apps we package > > * people that have shown to be constructive after their posts got > moderated for a while > > Enforce the devel topics and never ever answer support questions on that > list -> those shall are on topic for fedora-devel-users. Just leave fedora-devel, recharter it. Don't bother with renaming. Bill From max at spevack.org Tue Jan 9 14:53:54 2007 From: max at spevack.org (Max Spevack) Date: Tue, 9 Jan 2007 09:53:54 -0500 (EST) Subject: fedora board meeting today In-Reply-To: <1168348590.1868.28.camel@localhost.localdomain> References: <1168348590.1868.28.camel@localhost.localdomain> Message-ID: Today's Fedora Board meeting will be at 5:00 PM Eastern/9:00 PM GMT. We have basically one topic today, which is Fedora Documentation. Karsten Wade, who is the Docs project leader, will be joining us. #fedora-board for a running summary of our phone call. Thanks as always to mdomsch's bot for providing us with the logs. Agenda below: On Tue, 9 Jan 2007, Paul W. Frields wrote: > Here's a possible agenda for the initial Docs segment of the Board > meeting today at 5pm Eastern. Karsten Wade, FDSCo chair, will be > attending the telephone conference call and hopefully will also check in > at #fedora-board on IRC concomitantly. Please make sure Karsten is CC'd > on any other messages in this thread as appropriate. > > We've had a thread going for a few days gathering some of these issues > on fedora-docs-list. I would expect Karsten will take the lead > especially on item 3 below, since he has more knowledge of who is > responsible for the issues in question. Karsten, you may not be able to > reply to this thread given the way the ML works, but I'll forward any of > your email responses to this list. > > 1. Wrap-up of last 6-12 months of advances for those who don't know > anything about Docs, seeing as how it is a little less obvious/sexy > than things like metadata parsers, stats tracking, and .gnu-hash. > ;-) I don't want to belabor this section so it will be a quick 2 > minutes. > > 2. Future plans, mainly a discussion of some other Fedora initiatives > that are going to impact the Docs Project and how we intend to deal > with them, as well as some plans of our own (wiki click-thru, new > FLOSS community outreach efforts). This may take 5-10 minutes. > > 3. What agenda would be complete without the gripe session? OK, not so > much gripes as some shortfalls that are really causing us pain and > costing us contributor follow-through and community goodwill. -- Max Spevack + http://spevack.org + gpg key -- http://spevack.org/max.asc + fingerprint -- CD52 5E72 369B B00D 9E9A 773E 2FDB CB46 5A17 CF21 From wwoods at redhat.com Tue Jan 9 16:25:57 2007 From: wwoods at redhat.com (Will Woods) Date: Tue, 09 Jan 2007 11:25:57 -0500 Subject: mailing-list reorganisation, round 4 on this list In-Reply-To: <45A0C837.7010601@leemhuis.info> References: <45896DDE.3080102@leemhuis.info> <459EB9B8.90005@leemhuis.info> <45A0C837.7010601@leemhuis.info> Message-ID: <1168359957.15360.11.camel@metroid.rdu.redhat.com> On Sun, 2007-01-07 at 11:15 +0100, Thorsten Leemhuis wrote: > Hi! > > Me again with the next round! > === fedora-triage-list === > > Not used much anymore; fedora-qa-list will take over the job (see below) > === To -list or not to list === > > Some of our mailing lists have the "-list" postfix, other not. I think > all new lists should be created without it. So, should this just be fedora-qa at redhat.com then? I'm fine with closing fedora-triage-list - the fedora-qa list should serve for all discussion of qa-related stuff (bug triaging, test tools, testing status, etc.) I'm also fine with fedora-test-list being repurposed for user discussion of rawhide and other test releases. But where should the automated reports that are currently sent to fedora-test-list go? I haven't seen a consensus on that. I'd prefer they stay out of the way of the discussion, so they'd have to be on a different list (or, at least, really easy to filter out of fedora-qa..) -w -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 189 bytes Desc: This is a digitally signed message part URL: From matt at domsch.com Tue Jan 9 17:11:42 2007 From: matt at domsch.com (Matt Domsch) Date: Tue, 9 Jan 2007 11:11:42 -0600 Subject: fedora board meeting today In-Reply-To: References: <1168348590.1868.28.camel@localhost.localdomain> Message-ID: <20070109171142.GA28942@domsch.com> On Tue, Jan 09, 2007 at 09:53:54AM -0500, Max Spevack wrote: > Today's Fedora Board meeting will be at 5:00 PM Eastern/9:00 PM GMT. > > We have basically one topic today, which is Fedora Documentation. Karsten > Wade, who is the Docs project leader, will be joining us. > > #fedora-board for a running summary of our phone call. Thanks as always > to mdomsch's bot for providing us with the logs. http://fedoraproject.org/wiki/Board/Meetings/2007-01-02 has last week's minutes. Please review and edit per your recollection. Thanks, Matt From fedora at leemhuis.info Tue Jan 9 17:31:51 2007 From: fedora at leemhuis.info (Thorsten Leemhuis) Date: Tue, 09 Jan 2007 18:31:51 +0100 Subject: mailing-list reorganisation, round 4 on this list In-Reply-To: <1168359957.15360.11.camel@metroid.rdu.redhat.com> References: <45896DDE.3080102@leemhuis.info> <459EB9B8.90005@leemhuis.info> <45A0C837.7010601@leemhuis.info> <1168359957.15360.11.camel@metroid.rdu.redhat.com> Message-ID: <45A3D187.6030207@leemhuis.info> Will Woods schrieb: > On Sun, 2007-01-07 at 11:15 +0100, Thorsten Leemhuis wrote: >> === fedora-triage-list === >> >> Not used much anymore; fedora-qa-list will take over the job (see below) > >> === To -list or not to list === >> >> Some of our mailing lists have the "-list" postfix, other not. I think >> all new lists should be created without it. > > So, should this just be fedora-qa at redhat.com then? BTW, I can also live with having "-list" behind all of them, but I'd prefer to have it (mostly) every or nowhere. > > I'm fine with closing fedora-triage-list - the fedora-qa list should > serve for all discussion of qa-related stuff (bug triaging, test tools, > testing status, etc.) > > I'm also fine with fedora-test-list being repurposed for user discussion > of rawhide and other test releases. k, thx for feedback > But where should the automated reports that are currently sent to > fedora-test-list go? I haven't seen a consensus on that. I'd prefer they > stay out of the way of the discussion, so they'd have to be on a > different list (or, at least, really easy to filter out of fedora-qa..) A separate channel on fedora-package announce? CU thl From notting at redhat.com Thu Jan 11 05:33:00 2007 From: notting at redhat.com (Bill Nottingham) Date: Thu, 11 Jan 2007 00:33:00 -0500 Subject: how to govern and manage the new combined repository In-Reply-To: <45A0D459.1050807@leemhuis.info> References: <45A0D459.1050807@leemhuis.info> Message-ID: <20070111053300.GA2586@nostromo.devel.redhat.com> Thorsten Leemhuis (fedora at leemhuis.info) said: > That's probably a good idea, but that needs a bit of preparation, too. > Thus (and because a board members ask me to) I sat down and will roughly > outline my thoughts on the issue with this mail. I don't want to start > yet another endless discussion with this mail/proposal (I know, I'm good > with that, but I'm not proud of it) -- rather I'd like to see similar > mails/proposals from other people where they describe how this whole > thing should be governed. OK, here's a proposal. Not sure it will match up to yours well... :) Governance in the post Core/Extras world Q: What are we trying to accomplish? A: To enable people to do Cool Stuff with Fedora. To enable people to make Fedora better. Q: How do we best accomplish this? A: Empower people, and get out of their way. So, what sort of structure should we have to enact this? On top is the Fedora board - the directing organization, the big picture thinkers, and the resolution point of last resort. Under them are various subprojects - we have the infrastructure project, the docs project, etc. So, what new structures do we *need*? 1) Fedora Packaging Project (or committee, or what have you) Charter: - set packaging standards - set packager standards - enforce those standards - encourage new contributors/contributions Structure? I'm of the opinion that it doesn't really matter - find people willing to do the work, *and do it*. But I could certainly see how the current FESCo model can work here, especially since FESCo handles most (if not all) of these areas. 2) Fedora Release Team Charter: - defines the schedule - defines the feature list (?) - enforces the freezes - spins such releases that we see fit (pushes the button, pushes to site, etc.) Structure? One lead, and a team of people. Frankly, election seems to be the decidedly wrong way of staffing this. Both of these 'report' to the board, if you're drawing a chart. That's *all* that I think we need that doesn't exist now. Sure, there are more groups, but those are all existing now, whether they be SIGs (KDE, Games), random people (LiveCD), projects with leaders and a group of contributors (Art, QA, Security). I think trying to mandate specific structures or forms for these groups isn't the best plan for now - just leave it simple, and up to *these groups* how they organize. Bill From fedora at leemhuis.info Thu Jan 11 17:55:32 2007 From: fedora at leemhuis.info (Thorsten Leemhuis) Date: Thu, 11 Jan 2007 18:55:32 +0100 Subject: how to govern and manage the new combined repository In-Reply-To: <20070111053300.GA2586@nostromo.devel.redhat.com> References: <45A0D459.1050807@leemhuis.info> <20070111053300.GA2586@nostromo.devel.redhat.com> Message-ID: <45A67A14.3010604@leemhuis.info> Bill Nottingham schrieb: > Thorsten Leemhuis (fedora at leemhuis.info) said: > Q: What are we trying to accomplish? > A: To enable people to do Cool Stuff with Fedora. To enable people > to make Fedora better. +1 > Q: How do we best accomplish this? > A: Empower people, and get out of their way. Well, your have a point, but I don't agree fully. One reasons for it: A clear infrastructure will actually help getting the community involved. Otherwise some contributors might say "I did not know where to ask to get involved, thus I move along to something else" (see the recent discussion about fedora-desktop on fedora-devel -- that shows the problem nicely ). That's what I'd like to avoid. > So, what sort of structure should we have to enact this? > > On top is the Fedora board - the directing organization, the big > picture thinkers, and the resolution point of last resort. +1 > Under them are various subprojects - we have the infrastructure > project, the docs project, etc. So, what new structures do we > *need*? Hear I to disagree. I think it does not work well if we have to much projects on the same level if they have to interact a lot (and those two your porpose have to interact a lot). An reasons for my opinion: the communication and "who does foo" between the Packaging Committee and FESCo sometimes sucked (the recent "conflicts" issue is a good example). Having something like a FTC (Fedora Technical Committee) at the top might help as it can say "Either you work out something until X or we do it, as we and/or group 'b' need a solution *really soon*". > 1) Fedora Packaging Project (or committee, or what have you) > > Charter: > - set packaging standards > - set packager standards > - enforce those standards > - encourage new contributors/contributions I like the parts to have one committee that takes care of both packaging standards and enforcing them.. > Structure? I'm of the opinion that it doesn't really matter - find > people willing to do the work, *and do it*. But I could certainly > see how the current FESCo model can work here, especially since > FESCo handles most (if not all) of these areas. Well, > - set packaging standards > - set packager standards are the job of the Packaging Committee. > 2) Fedora Release Team > > Charter: > - defines the schedule > - defines the feature list (?) > - enforces the freezes I think that could lead to problems if this group handles the freezes if all the other repo work falls into the area of group "1" > - spins such releases that we see fit (pushes the button, pushes > to site, etc.) > > Structure? [...] See above. Just my 2 cent . CU thl From notting at redhat.com Thu Jan 11 18:02:25 2007 From: notting at redhat.com (Bill Nottingham) Date: Thu, 11 Jan 2007 13:02:25 -0500 Subject: how to govern and manage the new combined repository In-Reply-To: <45A67A14.3010604@leemhuis.info> References: <45A0D459.1050807@leemhuis.info> <20070111053300.GA2586@nostromo.devel.redhat.com> <45A67A14.3010604@leemhuis.info> Message-ID: <20070111180225.GD9785@nostromo.devel.redhat.com> Thorsten Leemhuis (fedora at leemhuis.info) said: > > Q: How do we best accomplish this? > > A: Empower people, and get out of their way. > > Well, your have a point, but I don't agree fully. One reasons for it: A > clear infrastructure will actually help getting the community involved. > Otherwise some contributors might say "I did not know where to ask to > get involved, thus I move along to something else" (see the recent > discussion about fedora-desktop on fedora-devel -- that shows the > problem nicely ). That's what I'd like to avoid. There should be mechanisms for people to contribute, pages on the wiki, logs, etc. But I don't see what good mandating that you need a committee, made up of X, Y, or Z does for you. Most of the projects I know about now (art, the SIGs, QA, Infrastrucute) don't really fall into this category. Essentially, communities should provide their own governance. AFAICT, the fedora-desktop thread was just a bunch of mutual namecalling. Maybe I missed something in the noise. Realistically, I want to hear - what are people trying to do that they can't do now? What are they trying to *ask* about getting involved? > > Under them are various subprojects - we have the infrastructure > > project, the docs project, etc. So, what new structures do we > > *need*? > > Hear I to disagree. I think it does not work well if we have to much > projects on the same level if they have to interact a lot (and those two > your porpose have to interact a lot). > > An reasons for my opinion: the communication and "who does foo" between > the Packaging Committee and FESCo sometimes sucked (the recent > "conflicts" issue is a good example). Having something like a FTC > (Fedora Technical Committee) at the top might help as it can say "Either > you work out something until X or we do it, as we and/or group 'b' need > a solution *really soon*". This implies to me that the packaging committee shouldn't have been separated from FESCo in the first place - that's why I put it back together in my proposal. ;) > > 2) Fedora Release Team > > > > Charter: > > - defines the schedule > > - defines the feature list (?) > > - enforces the freezes > > I think that could lead to problems if this group handles the freezes if > all the other repo work falls into the area of group "1" I don't see how it would be that bad - realistically, if people can't work together, we have problems. Bill From rdieter at math.unl.edu Thu Jan 11 18:04:47 2007 From: rdieter at math.unl.edu (Rex Dieter) Date: Thu, 11 Jan 2007 12:04:47 -0600 Subject: how to govern and manage the new combined repository In-Reply-To: <45A67A14.3010604@leemhuis.info> References: <45A0D459.1050807@leemhuis.info> <20070111053300.GA2586@nostromo.devel.redhat.com> <45A67A14.3010604@leemhuis.info> Message-ID: <45A67C3F.3080706@math.unl.edu> Thorsten Leemhuis wrote: > Bill Nottingham schrieb: >> Under them are various subprojects - we have the infrastructure >> project, the docs project, etc. So, what new structures do we >> *need*? > > Hear I to disagree. I think it does not work well if we have to much > projects on the same level if they have to interact a lot (and those two > your porpose have to interact a lot). > > An reasons for my opinion: the communication and "who does foo" between > the Packaging Committee and FESCo sometimes sucked IMO, the problem(s) experienced haven't been that these projects were on the same level, but rather that there wasn't a clear description/mandate of authority and bounds of said projects. In this instance (FESCo/FPC), it was unclear whose job it was to enforce packaging policy. -- Rex From fedora at leemhuis.info Thu Jan 11 18:52:39 2007 From: fedora at leemhuis.info (Thorsten Leemhuis) Date: Thu, 11 Jan 2007 19:52:39 +0100 Subject: how to govern and manage the new combined repository In-Reply-To: <45A67C3F.3080706@math.unl.edu> References: <45A0D459.1050807@leemhuis.info> <20070111053300.GA2586@nostromo.devel.redhat.com> <45A67A14.3010604@leemhuis.info> <45A67C3F.3080706@math.unl.edu> Message-ID: <45A68777.5090706@leemhuis.info> Rex Dieter schrieb: > Thorsten Leemhuis wrote: >> Bill Nottingham schrieb: >>> Under them are various subprojects - we have the infrastructure >>> project, the docs project, etc. So, what new structures do we >>> *need*? >> Hear I to disagree. I think it does not work well if we have to much >> projects on the same level if they have to interact a lot (and those two >> your porpose have to interact a lot). >> An reasons for my opinion: the communication and "who does foo" between >> the Packaging Committee and FESCo sometimes sucked > IMO, the problem(s) experienced haven't been that these projects were on > the same level, but rather that there wasn't a clear description/mandate > of authority and bounds of said projects. In this instance (FESCo/FPC), > it was unclear whose job it was to enforce packaging policy. Maybe that part of the problem. The PC afaics said that they don't want to enforce stuff. Thus FESCo sometimes hit problems, had to move them over to PC for solving them in the guidlines. The PC then had to move the issue back to FESCO to actually enforce what the PC decided. That was fine when we had Core and Extras, but soon it will be to much overhead for no real gain. CU thl From fedora at leemhuis.info Fri Jan 12 15:03:01 2007 From: fedora at leemhuis.info (Thorsten Leemhuis) Date: Fri, 12 Jan 2007 16:03:01 +0100 Subject: move mailinglists to a new server (for example something like lists.fedoraproject.org) Message-ID: <45A7A325.7080906@leemhuis.info> Hi, during the "mailing list reorganization" discussion on fedora-devel the idea came up again (?) to move the fedora lists (?) off from redhat.com to a separate machine that's under our control. E.g. something like lists.fedoraproject.org, as that would make it obvious - that it's a list you are posting to - that it's "Fedora" The sub-thread about it starts here: https://www.redhat.com/archives/fedora-devel-list/2007-January/msg00826.html skvidal and mether already indicated on the list that they would like such a move. mmcgrath and his recruits are currently evaluating a little bit if such a move is possible and how much work it would make (site note: three well known extras contributors already volunteered to help with that effort). Before we invest more time in that idea I'd like to ask on this list: would such a move be fine for everybody (especially those from the board)? If not please yell now! tia ;-) @wwoods: what do we do with fedora-qa(-list)? How urgent do you need it? CU thl (?) http://www.redhat.com/archives/fedora-marketing-list/2005-December/msg00003.html (?) The move would be slowly -- details would need to be worked out. From bpepple at fedoraproject.org Fri Jan 12 16:00:53 2007 From: bpepple at fedoraproject.org (Brian Pepple) Date: Fri, 12 Jan 2007 11:00:53 -0500 Subject: move mailinglists to a new server (for example something like lists.fedoraproject.org) In-Reply-To: <45A7A325.7080906@leemhuis.info> References: <45A7A325.7080906@leemhuis.info> Message-ID: <1168617653.8625.1.camel@Chuck> On Fri, 2007-01-12 at 16:03 +0100, Thorsten Leemhuis wrote: > > skvidal and mether already indicated on the list that they would like > such a move. mmcgrath and his recruits are currently evaluating a > little bit if such a move is possible and how much work it would make > (site note: three well known extras contributors already volunteered to > help with that effort). Before we invest more time in that idea I'd like > to ask on this list: > > would such a move be fine for everybody (especially those from the > board)? If not please yell now! tia ;-) I think it would be a good idea. /B -- Brian Pepple gpg --keyserver pgp.mit.edu --recv-keys 810CC15E BD5E 6F9E 8688 E668 8F5B CBDE 326A E936 810C C15E -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 189 bytes Desc: This is a digitally signed message part URL: From jwboyer at jdub.homelinux.org Fri Jan 12 16:22:41 2007 From: jwboyer at jdub.homelinux.org (Josh Boyer) Date: Fri, 12 Jan 2007 10:22:41 -0600 Subject: move mailinglists to a new server (for example something like lists.fedoraproject.org) In-Reply-To: <45A7A325.7080906@leemhuis.info> References: <45A7A325.7080906@leemhuis.info> Message-ID: <1168618961.3696.24.camel@zod.rchland.ibm.com> On Fri, 2007-01-12 at 16:03 +0100, Thorsten Leemhuis wrote: > Hi, > > during the "mailing list reorganization" discussion on fedora-devel the > idea came up again (?) to move the fedora lists (?) off from redhat.com > to a separate machine that's under our control. E.g. something like > lists.fedoraproject.org, as that would make it obvious > - that it's a list you are posting to > - that it's "Fedora" > > The sub-thread about it starts here: > https://www.redhat.com/archives/fedora-devel-list/2007-January/msg00826.html > > skvidal and mether already indicated on the list that they would like > such a move. mmcgrath and his recruits are currently evaluating a > little bit if such a move is possible and how much work it would make > (site note: three well known extras contributors already volunteered to > help with that effort). Before we invest more time in that idea I'd like > to ask on this list: > > would such a move be fine for everybody (especially those from the > board)? If not please yell now! tia ;-) This sounds like a good idea to me [1]. josh [1] As long as we have a big fat warning first [2] [2] If I was as smart as Thorsten, I would have real superscript numbers. From fedora at leemhuis.info Fri Jan 12 16:29:30 2007 From: fedora at leemhuis.info (Thorsten Leemhuis) Date: Fri, 12 Jan 2007 17:29:30 +0100 Subject: how to govern and manage the new combined repository (second proposal from thl) In-Reply-To: <45A0D459.1050807@leemhuis.info> References: <45A0D459.1050807@leemhuis.info> Message-ID: <45A7B76A.3060301@leemhuis.info> Thorsten Leemhuis schrieb: > Quoting Bill from: > https://www.redhat.com/archives/fedora-devel-list/2007-January/msg00091.html > >> What's left of the Core Steering Committee >> is going to work with the Fedora Board and FESCO to figure out just >> how this new combined repository is going to be governed and managed. > > Yeah, that becomes more an more urgent afaics. The idea was raised to > sort this out in a special IRC-meeting and/or conference call with > people from the Core Cabal, FESCo and the Board. [...] Here is my second proposal: New proposal, again with a small graphic attached: - FESCo and the Core Cabal (f13 and notting [jeremy is a member of FESCO already]) become FTC -- Fedora Technical Committee (the final name still needs to be discussed, there are some alternatives floating around). That makes 15 Members for now. - the FTC sits below the Board and handles all the day to day work and the details around the packages and the releases; some of that work get delegated to be done by SIGs (Extras has some already that simply continue to exist; further create SIGs [even if that means only one or two people] for each Spin). Big or long term decisions (roadmaps for example) get worked out together by the FTC and the Board. - Those four members that are in the Packaging Committee and in FESCo now will represent the Packaging Committee in FTC, too. The Packaging Committee at the same time becomes a SIG. It will work as before; similar how FESCo and the Core Cabal had veto power it's now FTC that can block decisions. - the FTC will run with a similar scheme how FESCo ran until now (e.g. in the open, public meetings, ...) - the FTC in the current form has a lifespan until six weeks after F7 was released. The FTC until then has to work out a plan how FTC will be constituted after that (e.g. size, seats [some elected, other appointed?], fixed level of community/red hat members, ...) together with the Board. Comments? CU thl -------------- next part -------------- A non-text attachment was scrubbed... Name: thlproposal2.png Type: image/png Size: 19691 bytes Desc: not available URL: From tibbs at math.uh.edu Fri Jan 12 16:34:19 2007 From: tibbs at math.uh.edu (Jason L Tibbitts III) Date: 12 Jan 2007 10:34:19 -0600 Subject: move mailinglists to a new server (for example something like lists.fedoraproject.org) In-Reply-To: <45A7A325.7080906@leemhuis.info> References: <45A7A325.7080906@leemhuis.info> Message-ID: >>>>> "TL" == Thorsten Leemhuis writes: TL> would such a move be fine for everybody (especially those from the TL> board)? If not please yell now! tia ;-) I think it's an excellent idea if it gives us more flexibility in how we manage the lists and more speed in their creation. If not, then I'm not sure if it would be worth the trouble. As I understand things, list creation currently has to do through some Red Hat department and involves various marketing and branding issues as well. If that's true, then getting away from that would certainly be an improvement. - J< From wwoods at redhat.com Fri Jan 12 16:36:24 2007 From: wwoods at redhat.com (Will Woods) Date: Fri, 12 Jan 2007 11:36:24 -0500 Subject: move mailinglists to a new server (for example something like lists.fedoraproject.org) In-Reply-To: <45A7A325.7080906@leemhuis.info> References: <45A7A325.7080906@leemhuis.info> Message-ID: <1168619784.3033.7.camel@metroid.rdu.redhat.com> On Fri, 2007-01-12 at 16:03 +0100, Thorsten Leemhuis wrote: > would such a move be fine for everybody (especially those from the > board)? If not please yell now! tia ;-) For what it's worth, it's always struck me as a bit odd that the fedora lists are still @redhat.com. Moving them to fedoraproject.org means more work for the infrastructure folks, though, so I can't support the idea unless they're willing and able. > @wwoods: what do we do with fedora-qa(-list)? How urgent do you need it? It's not "OMG HEAD ON FIRE" urgent. We can keep using fedora-test-list until the board et. al. figure out where mailing lists should live. I'd prefer not to move to fedora-qa at rh.com, just to be forced to move again two weeks later. -w -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 189 bytes Desc: This is a digitally signed message part URL: From notting at redhat.com Fri Jan 12 16:36:22 2007 From: notting at redhat.com (Bill Nottingham) Date: Fri, 12 Jan 2007 11:36:22 -0500 Subject: move mailinglists to a new server (for example something like lists.fedoraproject.org) In-Reply-To: References: <45A7A325.7080906@leemhuis.info> Message-ID: <20070112163622.GD27230@nostromo.devel.redhat.com> Jason L Tibbitts III (tibbs at math.uh.edu) said: > I think it's an excellent idea if it gives us more flexibility in how > we manage the lists and more speed in their creation. If not, then > I'm not sure if it would be worth the trouble. > > As I understand things, list creation currently has to do through some > Red Hat department Yes... > and involves various marketing and branding issues > as well. ... and not to my knowledge. What I'd be concerned about is simply adding to the admin load, making sure the servers have the capacity, dealing with spam/postmaster/etc noise. Bill From skvidal at linux.duke.edu Fri Jan 12 16:44:11 2007 From: skvidal at linux.duke.edu (seth vidal) Date: Fri, 12 Jan 2007 11:44:11 -0500 Subject: move mailinglists to a new server (for example something like lists.fedoraproject.org) In-Reply-To: <20070112163622.GD27230@nostromo.devel.redhat.com> References: <45A7A325.7080906@leemhuis.info> <20070112163622.GD27230@nostromo.devel.redhat.com> Message-ID: <1168620251.11006.25.camel@cutter> On Fri, 2007-01-12 at 11:36 -0500, Bill Nottingham wrote: > Jason L Tibbitts III (tibbs at math.uh.edu) said: > > I think it's an excellent idea if it gives us more flexibility in how > > we manage the lists and more speed in their creation. If not, then > > I'm not sure if it would be worth the trouble. > > > > As I understand things, list creation currently has to do through some > > Red Hat department > > Yes... > > > and involves various marketing and branding issues > > as well. > > ... and not to my knowledge. > > What I'd be concerned about is simply adding to the admin load, making > sure the servers have the capacity, dealing with spam/postmaster/etc > noise. This is what we brought up yesterday in the infrastructure meeting. Capacity issues, and available resources. Mike said he would talk to rh IS folks about what sort of capacity and mailman tweaks are required, now. -sv From mmcgrath at fedoraproject.org Fri Jan 12 16:45:16 2007 From: mmcgrath at fedoraproject.org (Mike McGrath) Date: Fri, 12 Jan 2007 10:45:16 -0600 Subject: how to govern and manage the new combined repository (second proposal from thl) In-Reply-To: <45A7B76A.3060301@leemhuis.info> References: <45A0D459.1050807@leemhuis.info> <45A7B76A.3060301@leemhuis.info> Message-ID: <3237e4410701120845k10735412rc35fc153bff055f9@mail.gmail.com> On 1/12/07, Thorsten Leemhuis wrote: > Thorsten Leemhuis schrieb: > > Quoting Bill from: > > https://www.redhat.com/archives/fedora-devel-list/2007-January/msg00091.html > > New proposal, again with a small graphic attached: > > - FESCo and the Core Cabal (f13 and notting [jeremy is a member of FESCO > already]) become FTC -- Fedora Technical Committee (the final name still > needs to be discussed, there are some alternatives floating around). > That makes 15 Members for now. > - the FTC sits below the Board and handles all the day to day work and > the details around the packages and the releases; some of that work get > delegated to be done by SIGs (Extras has some already that simply > continue to exist; further create SIGs [even if that means only one or > two people] for each Spin). Big or long term decisions (roadmaps for > example) get worked out together by the FTC and the Board. > - Those four members that are in the Packaging Committee and in FESCo > now will represent the Packaging Committee in FTC, too. The Packaging > Committee at the same time becomes a SIG. It will work as before; > similar how FESCo and the Core Cabal had veto power it's now FTC that > can block decisions. > - the FTC will run with a similar scheme how FESCo ran until now (e.g. > in the open, public meetings, ...) > - the FTC in the current form has a lifespan until six weeks after F7 > was released. The FTC until then has to work out a plan how FTC will be > constituted after that (e.g. size, seats [some elected, other > appointed?], fixed level of community/red hat members, ...) together > with the Board. > > Comments? > I see the FTC as 3 different groups: Technical 'vision' aspects of fedora (Like Gnome being our default wm) Packaging (Our packaging guys rock BTW) Technical Integration (More on the dev side) I see these entities as being and acting differently and independently of each other, especially for day to day things. It might be worth it to make them somewhat mutually exclusive, not allowing any one individual to be in all 3. Though in its "6 weeks after F7" form I think this proposal looks great. Lets see how it goes and start discussion after F7 is released. -Mike From notting at redhat.com Fri Jan 12 17:21:28 2007 From: notting at redhat.com (Bill Nottingham) Date: Fri, 12 Jan 2007 12:21:28 -0500 Subject: how to govern and manage the new combined repository (second proposal from thl) In-Reply-To: <45A7B76A.3060301@leemhuis.info> References: <45A0D459.1050807@leemhuis.info> <45A7B76A.3060301@leemhuis.info> Message-ID: <20070112172128.GE27230@nostromo.devel.redhat.com> Thorsten Leemhuis (fedora at leemhuis.info) said: > - FESCo and the Core Cabal (f13 and notting [jeremy is a member of FESCO > already]) become FTC -- Fedora Technical Committee (the final name still > needs to be discussed, there are some alternatives floating around). I'd suggest Fedora Distribution Committee (FEDCo?), as really, this is all about the distributions we do. I'm open to more suggestions, but I think FTC 1) has acronym collisions 2) is somewhat broad (infrastructure is fairly technical, after all) > Comments? Seems OK to me. Bill From notting at redhat.com Fri Jan 12 17:24:29 2007 From: notting at redhat.com (Bill Nottingham) Date: Fri, 12 Jan 2007 12:24:29 -0500 Subject: how to govern and manage the new combined repository (second proposal from thl) In-Reply-To: <3237e4410701120845k10735412rc35fc153bff055f9@mail.gmail.com> References: <45A0D459.1050807@leemhuis.info> <45A7B76A.3060301@leemhuis.info> <3237e4410701120845k10735412rc35fc153bff055f9@mail.gmail.com> Message-ID: <20070112172429.GF27230@nostromo.devel.redhat.com> Mike McGrath (mmcgrath at fedoraproject.org) said: > I see the FTC as 3 different groups: > > Technical 'vision' aspects of fedora (Like Gnome being our default wm) > > Packaging (Our packaging guys rock BTW) > > Technical Integration (More on the dev side) > > > I see these entities as being and acting differently and independently > of each other, especially for day to day things. It might be worth it > to make them somewhat mutually exclusive, not allowing any one > individual to be in all 3. I can see having them separate, but I'm not sure about mutually exclusive. Might be something to put off in the name of simplicity for now. Bill From fedora at leemhuis.info Fri Jan 12 17:28:47 2007 From: fedora at leemhuis.info (Thorsten Leemhuis) Date: Fri, 12 Jan 2007 18:28:47 +0100 Subject: how to govern and manage the new combined repository (second proposal from thl) In-Reply-To: <3237e4410701120845k10735412rc35fc153bff055f9@mail.gmail.com> References: <45A0D459.1050807@leemhuis.info> <45A7B76A.3060301@leemhuis.info> <3237e4410701120845k10735412rc35fc153bff055f9@mail.gmail.com> Message-ID: <45A7C54F.8040403@leemhuis.info> Mike McGrath schrieb: > On 1/12/07, Thorsten Leemhuis wrote: >> Thorsten Leemhuis schrieb: >>> Quoting Bill from: >>> https://www.redhat.com/archives/fedora-devel-list/2007-January/msg00091.html >> New proposal, again with a small graphic attached: >>[...] >> Comments? > > I see the FTC as 3 different groups: > > Technical 'vision' aspects of fedora (Like Gnome being our default wm) 'visions' IMHO have to come from the Board (or come from other Committees/Groups/the community and get then get ACKed by the Board). > Packaging (Our packaging guys rock BTW) > Technical Integration (More on the dev side) > > I see these entities as being and acting differently and independently > of each other, especially for day to day things. And I more and more think we should keep those two things as close together as possible to avoid interaction problems between different groups -- e.g. have a strong group at the top (below the Board) that's interested in day-to-day work and sub-groups that takes care of the different things. If a problems comes up that is relevant for two group that strong group can handle. I fear that otherwise group "a" will say "b, that you problem" while "b" says "a, no, you have to solve that". Both in such situations often have good reasons for their standpoint. But it does not help to get problems solved quickly (read: if often does not get fixed properly or only slow). > [...] CU thl From fedora at leemhuis.info Fri Jan 12 17:37:32 2007 From: fedora at leemhuis.info (Thorsten Leemhuis) Date: Fri, 12 Jan 2007 18:37:32 +0100 Subject: how to govern and manage the new combined repository (second proposal from thl) In-Reply-To: <20070112172128.GE27230@nostromo.devel.redhat.com> References: <45A0D459.1050807@leemhuis.info> <45A7B76A.3060301@leemhuis.info> <20070112172128.GE27230@nostromo.devel.redhat.com> Message-ID: <45A7C75C.1070804@leemhuis.info> Bill Nottingham schrieb: > Thorsten Leemhuis (fedora at leemhuis.info) said: >> - FESCo and the Core Cabal (f13 and notting [jeremy is a member of FESCO >> already]) become FTC -- Fedora Technical Committee (the final name still >> needs to be discussed, there are some alternatives floating around). > I'd suggest Fedora Distribution Committee (FEDCo?), as really, this is > all about the distributions we do. I'm fine with a FESDco *as long* as it's position is below the FTC > I'm open to more suggestions, but I > think FTC 1) has acronym collisions +1 > 2) is somewhat broad (infrastructure > is fairly technical, after all) See my reply to mmcgrath. CU thl From mmcgrath at fedoraproject.org Fri Jan 12 17:44:09 2007 From: mmcgrath at fedoraproject.org (Mike McGrath) Date: Fri, 12 Jan 2007 11:44:09 -0600 Subject: how to govern and manage the new combined repository (second proposal from thl) In-Reply-To: <45A7C75C.1070804@leemhuis.info> References: <45A0D459.1050807@leemhuis.info> <45A7B76A.3060301@leemhuis.info> <20070112172128.GE27230@nostromo.devel.redhat.com> <45A7C75C.1070804@leemhuis.info> Message-ID: <3237e4410701120944w21d1fafcr7a7512b77eab08ec@mail.gmail.com> On 1/12/07, Thorsten Leemhuis wrote: > Bill Nottingham schrieb: > > Thorsten Leemhuis (fedora at leemhuis.info) said: > >> - FESCo and the Core Cabal (f13 and notting [jeremy is a member of FESCO > >> already]) become FTC -- Fedora Technical Committee (the final name still > >> needs to be discussed, there are some alternatives floating around). > > I'd suggest Fedora Distribution Committee (FEDCo?), as really, this is > > all about the distributions we do. > > I'm fine with a FESDco *as long* as it's position is below the FTC > > > I'm open to more suggestions, but I > > think FTC 1) has acronym collisions > > +1 > > > 2) is somewhat broad (infrastructure > > is fairly technical, after all) > > See my reply to mmcgrath. In the end I guess my concern is making sure that the groups are able to focus on their task while avoiding noise. "Do one thing well, Play well with others" -Mike From fedora at leemhuis.info Fri Jan 12 17:59:41 2007 From: fedora at leemhuis.info (Thorsten Leemhuis) Date: Fri, 12 Jan 2007 18:59:41 +0100 Subject: how to govern and manage the new combined repository (second proposal from thl) In-Reply-To: <3237e4410701120944w21d1fafcr7a7512b77eab08ec@mail.gmail.com> References: <45A0D459.1050807@leemhuis.info> <45A7B76A.3060301@leemhuis.info> <20070112172128.GE27230@nostromo.devel.redhat.com> <45A7C75C.1070804@leemhuis.info> <3237e4410701120944w21d1fafcr7a7512b77eab08ec@mail.gmail.com> Message-ID: <45A7CC8D.6070502@leemhuis.info> Mike McGrath schrieb: > On 1/12/07, Thorsten Leemhuis wrote: >> Bill Nottingham schrieb: >>> Thorsten Leemhuis (fedora at leemhuis.info) said: >>>> - FESCo and the Core Cabal (f13 and notting [jeremy is a member of FESCO >>>> already]) become FTC -- Fedora Technical Committee (the final name still >>>> needs to be discussed, there are some alternatives floating around). >>> I'd suggest Fedora Distribution Committee (FEDCo?), as really, this is >>> all about the distributions we do. >> I'm fine with a FESDco *as long* as it's position is below the FTC >>> I'm open to more suggestions, but I >>> think FTC 1) has acronym collisions >> +1 >>> 2) is somewhat broad (infrastructure >>> is fairly technical, after all) >> See my reply to mmcgrath. > In the end I guess my concern is making sure that the groups are able > to focus on their task while avoiding noise. "Do one thing well, Play > well with others" And I think we tried that already with FESCo and PC in the past and it worked, but I was not to happy with it (it was not really bad either). That's why I want a strong groups that's involved in day-to-day things (the Board afaics does not want to get involved in those) an coordinates the work that is relevant for multiple groups. Or even overrule them if needed. CU thl From notting at redhat.com Fri Jan 12 18:33:28 2007 From: notting at redhat.com (Bill Nottingham) Date: Fri, 12 Jan 2007 13:33:28 -0500 Subject: how to govern and manage the new combined repository (second proposal from thl) In-Reply-To: <45A7C75C.1070804@leemhuis.info> References: <45A0D459.1050807@leemhuis.info> <45A7B76A.3060301@leemhuis.info> <20070112172128.GE27230@nostromo.devel.redhat.com> <45A7C75C.1070804@leemhuis.info> Message-ID: <20070112183327.GA2935@nostromo.devel.redhat.com> Thorsten Leemhuis (fedora at leemhuis.info) said: > >> - FESCo and the Core Cabal (f13 and notting [jeremy is a member of FESCO > >> already]) become FTC -- Fedora Technical Committee (the final name still > >> needs to be discussed, there are some alternatives floating around). > > I'd suggest Fedora Distribution Committee (FEDCo?), as really, this is > > all about the distributions we do. > > I'm fine with a FESDco *as long* as it's position is below the FTC Huh? I'm suggesting that what you call FTC *be* called the distribution committee - really, that's the 'technical' aspect they're involved in, not infrastructure, web sites, art, etc. Bill From fedora at leemhuis.info Fri Jan 12 18:50:23 2007 From: fedora at leemhuis.info (Thorsten Leemhuis) Date: Fri, 12 Jan 2007 19:50:23 +0100 Subject: how to govern and manage the new combined repository (second proposal from thl) In-Reply-To: <20070112183327.GA2935@nostromo.devel.redhat.com> References: <45A0D459.1050807@leemhuis.info> <45A7B76A.3060301@leemhuis.info> <20070112172128.GE27230@nostromo.devel.redhat.com> <45A7C75C.1070804@leemhuis.info> <20070112183327.GA2935@nostromo.devel.redhat.com> Message-ID: <45A7D86F.1020207@leemhuis.info> Bill Nottingham schrieb: > Thorsten Leemhuis (fedora at leemhuis.info) said: >>>> - FESCo and the Core Cabal (f13 and notting [jeremy is a member of FESCO >>>> already]) become FTC -- Fedora Technical Committee (the final name still >>>> needs to be discussed, there are some alternatives floating around). >>> I'd suggest Fedora Distribution Committee (FEDCo?), as really, this is >>> all about the distributions we do. >> I'm fine with a FESDco *as long* as it's position is below the FTC > Huh? I'm suggesting that what you call FTC *be* called the distribution > committee - really, that's the 'technical' aspect they're involved in, > not infrastructure, web sites, art, etc. Ohh, sorry, I read that as if you wanted another committee. Okay, FESDco might be a good idea as well. Cu thl From fedora at leemhuis.info Sat Jan 13 17:30:37 2007 From: fedora at leemhuis.info (Thorsten Leemhuis) Date: Sat, 13 Jan 2007 18:30:37 +0100 Subject: names for the fesco successor (was: Re: how to govern and manage the new combined repository (second proposal from thl)) In-Reply-To: <45A7B76A.3060301@leemhuis.info> References: <45A0D459.1050807@leemhuis.info> <45A7B76A.3060301@leemhuis.info> Message-ID: <45A9173D.9050905@leemhuis.info> Thorsten Leemhuis schrieb: > Thorsten Leemhuis schrieb: >> Quoting Bill from: >> https://www.redhat.com/archives/fedora-devel-list/2007-January/msg00091.html >> >>> What's left of the Core Steering Committee >>> is going to work with the Fedora Board and FESCO to figure out just >>> how this new combined repository is going to be governed and managed. >> Yeah, that becomes more an more urgent afaics. The idea was raised to >> sort this out in a special IRC-meeting and/or conference call with >> people from the Core Cabal, FESCo and the Board. [...] > Here is my second proposal: > New proposal, again with a small graphic attached: [...] So, what's still mostly undecided is the name for the FESCo/Core Cabal and Packaging Committee successor. Suggested in the last days were afaik: FTC -- Fedora Technical Committee [1]? FTT -- Fedora Technical Team [2] FET -- Fedora Engineering Team [3] FEDCo -- Fedora Distribution Committee [4] Has anybody any better ideas? Cu thl [1] -- name clash with Federal Trade Commission. Really bad IMHO. 39 meanings in total on http://www.acronymfinder.com/af-query.asp?Acronym=FTC&Find=find&string=exact [2] -- "Failure to thrive (FTT) refers to a baby or child that is not developing as well as desired." 12 meanings in total on http://www.acronymfinder.com/af-query.asp?Acronym=FTT&Find=find&string=exact [3] -- "Field-Effect Transistor" 24 meanings in total on http://www.acronymfinder.com/af-query.asp?Acronym=FET&Find=find&string=exact [4] -- "Federal Employees Distributing Company (Co-Op)" 2 meanings on http://www.acronymfinder.com/af-query.asp?Acronym=FEDCo&Find=find&string=exact ? -- jwb, fyi, it seems my keyboard generates those nice UTF-8 superscript numbers only up to 3 -- after that it starts with ? :-( From Axel.Thimm at ATrpms.net Mon Jan 15 00:13:02 2007 From: Axel.Thimm at ATrpms.net (Axel Thimm) Date: Mon, 15 Jan 2007 01:13:02 +0100 Subject: move mailinglists to a new server (for example something like lists.fedoraproject.org) In-Reply-To: <45A7A325.7080906@leemhuis.info> References: <45A7A325.7080906@leemhuis.info> Message-ID: <20070115001302.GA7715@neu.nirvana> On Fri, Jan 12, 2007 at 04:03:01PM +0100, Thorsten Leemhuis wrote: > during the "mailing list reorganization" discussion on fedora-devel the > idea came up again (?) to move the fedora lists (?) off from redhat.com > to a separate machine that's under our control. E.g. something like > lists.fedoraproject.org, as that would make it obvious If things are going to be moved to fedoraproject.org (which would be nice), I'd suggest to rename the lists from fedora-foo[-list]@redhat.com to foo at fedoraproject.org E.g. drop the prefixing fedora- from the local part and the "lists." from the domain name. Using lists.fedoraproject.org as the web frontend for mailman is OK, though, that's the usual practice anyway (list addresses are plain @example.org and mailman's web interface sits on lists.example.org). You don't really need different hardware for just splitting the lists. -- Axel.Thimm at ATrpms.net -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 189 bytes Desc: not available URL: From fedora at leemhuis.info Mon Jan 15 05:48:15 2007 From: fedora at leemhuis.info (Thorsten Leemhuis) Date: Mon, 15 Jan 2007 06:48:15 +0100 Subject: move mailinglists to a new server (for example something like lists.fedoraproject.org) In-Reply-To: <20070115001302.GA7715@neu.nirvana> References: <45A7A325.7080906@leemhuis.info> <20070115001302.GA7715@neu.nirvana> Message-ID: <45AB159F.1020100@leemhuis.info> On 15.01.2007 01:13, Axel Thimm wrote: > On Fri, Jan 12, 2007 at 04:03:01PM +0100, Thorsten Leemhuis wrote: >> during the "mailing list reorganization" discussion on fedora-devel the >> idea came up again (?) to move the fedora lists (?) off from redhat.com >> to a separate machine that's under our control. E.g. something like >> lists.fedoraproject.org, as that would make it obvious > If things are going to be moved to fedoraproject.org (which would be > nice), I'd suggest > [...] > E.g. drop the prefixing fedora- Agreed -- I asked for opinions regarding dropping the "fedora-" prefix on fedora-devel already. Only one comment yet ("drop it"), with yours that makes two now. I tend to drop the fedora- prefix, too. > from the local part and the "lists." > from the domain name. Using lists.fedoraproject.org as the web > frontend for mailman is OK, though, that's the usual practice anyway > (list addresses are plain @example.org and mailman's web interface > sits on lists.example.org). Seems a lot of people prefer to have a host like lists.fedoraproject.org as that makes it obvious that you write to a list. I agree with those people. > You don't really need different hardware for just splitting the lists. You mean we could continue to use the red hat mainman? Sure, but we'd sill would need to route all our admin requests trough red hat IS, and it seems that has been painful sometimes in the past. CU thl From sundaram at fedoraproject.org Mon Jan 15 07:52:21 2007 From: sundaram at fedoraproject.org (Rahul Sundaram) Date: Mon, 15 Jan 2007 13:22:21 +0530 Subject: move mailinglists to a new server (for example something like lists.fedoraproject.org) In-Reply-To: <45AB159F.1020100@leemhuis.info> References: <45A7A325.7080906@leemhuis.info> <20070115001302.GA7715@neu.nirvana> <45AB159F.1020100@leemhuis.info> Message-ID: <45AB32B5.7070008@fedoraproject.org> Thorsten Leemhuis wrote: > On 15.01.2007 01:13, Axel Thimm wrote: >> On Fri, Jan 12, 2007 at 04:03:01PM +0100, Thorsten Leemhuis wrote: >>> during the "mailing list reorganization" discussion on fedora-devel the >>> idea came up again (?) to move the fedora lists (?) off from redhat.com >>> to a separate machine that's under our control. E.g. something like >>> lists.fedoraproject.org, as that would make it obvious >> If things are going to be moved to fedoraproject.org (which would be >> nice), I'd suggest > > [...] >> E.g. drop the prefixing fedora- > > Agreed -- I asked for opinions regarding dropping the "fedora-" prefix > on fedora-devel already. Only one comment yet ("drop it"), with yours > that makes two now. I tend to drop the fedora- prefix, too. I suspect that dropping the prefix would make it harder to search. There is a reason why other projects don't do this. Rahul From skvidal at linux.duke.edu Mon Jan 15 08:08:31 2007 From: skvidal at linux.duke.edu (seth vidal) Date: Mon, 15 Jan 2007 03:08:31 -0500 Subject: move mailinglists to a new server (for example something like lists.fedoraproject.org) In-Reply-To: <45AB32B5.7070008@fedoraproject.org> References: <45A7A325.7080906@leemhuis.info> <20070115001302.GA7715@neu.nirvana> <45AB159F.1020100@leemhuis.info> <45AB32B5.7070008@fedoraproject.org> Message-ID: <1168848511.21489.38.camel@cutter> On Mon, 2007-01-15 at 13:22 +0530, Rahul Sundaram wrote: > Thorsten Leemhuis wrote: > > On 15.01.2007 01:13, Axel Thimm wrote: > >> On Fri, Jan 12, 2007 at 04:03:01PM +0100, Thorsten Leemhuis wrote: > >>> during the "mailing list reorganization" discussion on fedora-devel the > >>> idea came up again (?) to move the fedora lists (?) off from redhat.com > >>> to a separate machine that's under our control. E.g. something like > >>> lists.fedoraproject.org, as that would make it obvious > >> If things are going to be moved to fedoraproject.org (which would be > >> nice), I'd suggest > > > [...] > >> E.g. drop the prefixing fedora- > > > > Agreed -- I asked for opinions regarding dropping the "fedora-" prefix > > on fedora-devel already. Only one comment yet ("drop it"), with yours > > that makes two now. I tend to drop the fedora- prefix, too. > > I suspect that dropping the prefix would make it harder to search. There > is a reason why other projects don't do this. > why would it make it harder to search? site:lists.fedoraproject.org at google. trivial to search. -sv From sundaram at fedoraproject.org Mon Jan 15 08:05:37 2007 From: sundaram at fedoraproject.org (Rahul Sundaram) Date: Mon, 15 Jan 2007 13:35:37 +0530 Subject: move mailinglists to a new server (for example something like lists.fedoraproject.org) In-Reply-To: <1168848511.21489.38.camel@cutter> References: <45A7A325.7080906@leemhuis.info> <20070115001302.GA7715@neu.nirvana> <45AB159F.1020100@leemhuis.info> <45AB32B5.7070008@fedoraproject.org> <1168848511.21489.38.camel@cutter> Message-ID: <45AB35D1.8010302@fedoraproject.org> seth vidal wrote: > > why would it make it harder to search? > > site:lists.fedoraproject.org at google. > > trivial to search. People dont search like that usually though. They type either search terms or the name of the mailing lists in the search box. We have a few years of established conventions around what they would expect the names to be. Rahul From Axel.Thimm at ATrpms.net Mon Jan 15 11:09:23 2007 From: Axel.Thimm at ATrpms.net (Axel Thimm) Date: Mon, 15 Jan 2007 12:09:23 +0100 Subject: move mailinglists to a new server (for example something like lists.fedoraproject.org) In-Reply-To: <45AB159F.1020100@leemhuis.info> References: <45A7A325.7080906@leemhuis.info> <20070115001302.GA7715@neu.nirvana> <45AB159F.1020100@leemhuis.info> Message-ID: <20070115110923.GC7715@neu.nirvana> On Mon, Jan 15, 2007 at 06:48:15AM +0100, Thorsten Leemhuis wrote: > On 15.01.2007 01:13, Axel Thimm wrote: > >You don't really need different hardware for just splitting the lists. > > You mean we could continue to use the red hat mainman? Sure, but we'd > sill would need to route all our admin requests trough red hat IS, and > it seems that has been painful sometimes in the past. That's a different story, of course, I was just referring to the technical part of. Politics, policies and resources are another matter altogether. But if the Red Hat mailman-in-person wouldn't mind to run lists with "fedoraproject.org" virtual domain it would have the benefit of running with the known availability and w/o reinventing the anti-spam-wheel again. It would also ease on Fedora's own resources in infrastructure. -- Axel.Thimm at ATrpms.net -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 189 bytes Desc: not available URL: From Axel.Thimm at ATrpms.net Mon Jan 15 11:14:33 2007 From: Axel.Thimm at ATrpms.net (Axel Thimm) Date: Mon, 15 Jan 2007 12:14:33 +0100 Subject: move mailinglists to a new server (for example something like lists.fedoraproject.org) In-Reply-To: <45AB35D1.8010302@fedoraproject.org> References: <45A7A325.7080906@leemhuis.info> <20070115001302.GA7715@neu.nirvana> <45AB159F.1020100@leemhuis.info> <45AB32B5.7070008@fedoraproject.org> <1168848511.21489.38.camel@cutter> <45AB35D1.8010302@fedoraproject.org> Message-ID: <20070115111433.GD7715@neu.nirvana> On Mon, Jan 15, 2007 at 01:35:37PM +0530, Rahul Sundaram wrote: > seth vidal wrote: > > > > >why would it make it harder to search? > > > >site:lists.fedoraproject.org at google. > > > >trivial to search. > > People dont search like that usually though. They type either search > terms or the name of the mailing lists in the search box. foo at fedoraproject.org would also pick up the proper mailing list. > We have a few years of established conventions around what they > would expect the names to be. Well, we're about to break allmost all established conventions by removing -list suffix, moving the domain name and renaming/merging lists. If people don't notice they'll search for fedora-extras at redhat.com anyway and wonder that this list's last posts were in mid 2007. -- Axel.Thimm at ATrpms.net -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 189 bytes Desc: not available URL: From skvidal at linux.duke.edu Mon Jan 15 15:47:41 2007 From: skvidal at linux.duke.edu (seth vidal) Date: Mon, 15 Jan 2007 10:47:41 -0500 Subject: move mailinglists to a new server (for example something like lists.fedoraproject.org) In-Reply-To: <45AB35D1.8010302@fedoraproject.org> References: <45A7A325.7080906@leemhuis.info> <20070115001302.GA7715@neu.nirvana> <45AB159F.1020100@leemhuis.info> <45AB32B5.7070008@fedoraproject.org> <1168848511.21489.38.camel@cutter> <45AB35D1.8010302@fedoraproject.org> Message-ID: <1168876061.21489.44.camel@cutter> On Mon, 2007-01-15 at 13:35 +0530, Rahul Sundaram wrote: > seth vidal wrote: > People dont search like that usually though. They type either search > terms or the name of the mailing lists in the search box. We have a few > years of established conventions around what they would expect the names > to be. "People don't search like that"?? I'd love to know where you determine that from. if I'm looking for a list I type in the email address of the list. if listname at lists.fedoraproject.org isn't specific enough I don't know what is. -sv From jwboyer at jdub.homelinux.org Mon Jan 15 16:10:00 2007 From: jwboyer at jdub.homelinux.org (Josh Boyer) Date: Mon, 15 Jan 2007 10:10:00 -0600 Subject: move mailinglists to a new server (for example something like lists.fedoraproject.org) In-Reply-To: <1168876061.21489.44.camel@cutter> References: <45A7A325.7080906@leemhuis.info> <20070115001302.GA7715@neu.nirvana> <45AB159F.1020100@leemhuis.info> <45AB32B5.7070008@fedoraproject.org> <1168848511.21489.38.camel@cutter> <45AB35D1.8010302@fedoraproject.org> <1168876061.21489.44.camel@cutter> Message-ID: <1168877400.5102.33.camel@zod.rchland.ibm.com> On Mon, 2007-01-15 at 10:47 -0500, seth vidal wrote: > On Mon, 2007-01-15 at 13:35 +0530, Rahul Sundaram wrote: > > seth vidal wrote: > > > People dont search like that usually though. They type either search > > terms or the name of the mailing lists in the search box. We have a few > > years of established conventions around what they would expect the names > > to be. > > "People don't search like that"?? I'd love to know where you determine > that from. > > if I'm looking for a list I type in the email address of the list. And if you are looking for a list that you don't know the name of, what do you do then? Think newbie. josh From skvidal at linux.duke.edu Mon Jan 15 16:36:50 2007 From: skvidal at linux.duke.edu (seth vidal) Date: Mon, 15 Jan 2007 11:36:50 -0500 Subject: move mailinglists to a new server (for example something like lists.fedoraproject.org) In-Reply-To: <1168877400.5102.33.camel@zod.rchland.ibm.com> References: <45A7A325.7080906@leemhuis.info> <20070115001302.GA7715@neu.nirvana> <45AB159F.1020100@leemhuis.info> <45AB32B5.7070008@fedoraproject.org> <1168848511.21489.38.camel@cutter> <45AB35D1.8010302@fedoraproject.org> <1168876061.21489.44.camel@cutter> <1168877400.5102.33.camel@zod.rchland.ibm.com> Message-ID: <1168879010.21489.48.camel@cutter> On Mon, 2007-01-15 at 10:10 -0600, Josh Boyer wrote: > On Mon, 2007-01-15 at 10:47 -0500, seth vidal wrote: > > On Mon, 2007-01-15 at 13:35 +0530, Rahul Sundaram wrote: > > > seth vidal wrote: > > > > > People dont search like that usually though. They type either search > > > terms or the name of the mailing lists in the search box. We have a few > > > years of established conventions around what they would expect the names > > > to be. > > > > "People don't search like that"?? I'd love to know where you determine > > that from. > > > > if I'm looking for a list I type in the email address of the list. > > And if you are looking for a list that you don't know the name of, what > do you do then? > > Think newbie. I am - you go to fedoraproject.org you see the 'lists link' entry as referencing where the mailing lists are. you go to lists.fedoraproject.org - which then shows you a listing of 20 or so lists as opposed to the 8 trillion on red hat's mailman server. think newbie. -sv From sundaram at fedoraproject.org Mon Jan 15 16:33:27 2007 From: sundaram at fedoraproject.org (Rahul Sundaram) Date: Mon, 15 Jan 2007 22:03:27 +0530 Subject: move mailinglists to a new server (for example something like lists.fedoraproject.org) In-Reply-To: <1168879010.21489.48.camel@cutter> References: <45A7A325.7080906@leemhuis.info> <20070115001302.GA7715@neu.nirvana> <45AB159F.1020100@leemhuis.info> <45AB32B5.7070008@fedoraproject.org> <1168848511.21489.38.camel@cutter> <45AB35D1.8010302@fedoraproject.org> <1168876061.21489.44.camel@cutter> <1168877400.5102.33.camel@zod.rchland.ibm.com> <1168879010.21489.48.camel@cutter> Message-ID: <45ABACD7.6010606@fedoraproject.org> seth vidal wrote: > I am - you go to fedoraproject.org > > you see the 'lists link' entry as referencing where the mailing lists > are. > > you go to lists.fedoraproject.org - which then shows you a listing of 20 > or so lists as opposed to the 8 trillion on red hat's mailman server. > > think newbie. To be fair, http://fedoraproject.org/wiki/Communicate which has a prominent link on the front page has a list of Fedora mailing lists. Rahul From skvidal at linux.duke.edu Mon Jan 15 17:07:33 2007 From: skvidal at linux.duke.edu (seth vidal) Date: Mon, 15 Jan 2007 12:07:33 -0500 Subject: move mailinglists to a new server (for example something like lists.fedoraproject.org) In-Reply-To: <45ABACD7.6010606@fedoraproject.org> References: <45A7A325.7080906@leemhuis.info> <20070115001302.GA7715@neu.nirvana> <45AB159F.1020100@leemhuis.info> <45AB32B5.7070008@fedoraproject.org> <1168848511.21489.38.camel@cutter> <45AB35D1.8010302@fedoraproject.org> <1168876061.21489.44.camel@cutter> <1168877400.5102.33.camel@zod.rchland.ibm.com> <1168879010.21489.48.camel@cutter> <45ABACD7.6010606@fedoraproject.org> Message-ID: <1168880853.21489.50.camel@cutter> On Mon, 2007-01-15 at 22:03 +0530, Rahul Sundaram wrote: > seth vidal wrote: > > > I am - you go to fedoraproject.org > > > > you see the 'lists link' entry as referencing where the mailing lists > > are. > > > > you go to lists.fedoraproject.org - which then shows you a listing of 20 > > or so lists as opposed to the 8 trillion on red hat's mailman server. > > > > think newbie. > > To be fair, http://fedoraproject.org/wiki/Communicate which has a > prominent link on the front page has a list of Fedora mailing lists. > which we have to manually update. -sv From sundaram at fedoraproject.org Mon Jan 15 17:02:24 2007 From: sundaram at fedoraproject.org (Rahul Sundaram) Date: Mon, 15 Jan 2007 22:32:24 +0530 Subject: move mailinglists to a new server (for example something like lists.fedoraproject.org) In-Reply-To: <1168880853.21489.50.camel@cutter> References: <45A7A325.7080906@leemhuis.info> <20070115001302.GA7715@neu.nirvana> <45AB159F.1020100@leemhuis.info> <45AB32B5.7070008@fedoraproject.org> <1168848511.21489.38.camel@cutter> <45AB35D1.8010302@fedoraproject.org> <1168876061.21489.44.camel@cutter> <1168877400.5102.33.camel@zod.rchland.ibm.com> <1168879010.21489.48.camel@cutter> <45ABACD7.6010606@fedoraproject.org> <1168880853.21489.50.camel@cutter> Message-ID: <45ABB3A0.20304@fedoraproject.org> seth vidal wrote: > On Mon, 2007-01-15 at 22:03 +0530, Rahul Sundaram wrote: >> To be fair, http://fedoraproject.org/wiki/Communicate which has a >> prominent link on the front page has a list of Fedora mailing lists. >> > > which we have to manually update. > Yes. Simply pointing to lists.fedoraproject.org is not very useful for newbie's either. We need to categorize them and have pointers like I currently maintain in the wiki page above. Rahul From fedora at leemhuis.info Mon Jan 15 18:28:27 2007 From: fedora at leemhuis.info (Thorsten Leemhuis) Date: Mon, 15 Jan 2007 19:28:27 +0100 Subject: how to govern and manage the new combined repository (second proposal from thl) In-Reply-To: <45A7B76A.3060301@leemhuis.info> References: <45A0D459.1050807@leemhuis.info> <45A7B76A.3060301@leemhuis.info> Message-ID: <45ABC7CB.8060202@leemhuis.info> Thorsten Leemhuis schrieb: > Thorsten Leemhuis schrieb: >> Quoting Bill from: >> https://www.redhat.com/archives/fedora-devel-list/2007-January/msg00091.html >> >>> What's left of the Core Steering Committee >>> is going to work with the Fedora Board and FESCO to figure out just >>> how this new combined repository is going to be governed and managed. >> Yeah, that becomes more an more urgent afaics. The idea was raised to >> sort this out in a special IRC-meeting and/or conference call with >> people from the Core Cabal, FESCo and the Board. [...] > Here is my second proposal: > New proposal, again with a small graphic attached: > - FESCo and the Core Cabal (f13 and notting [jeremy is a member of FESCO > already]) become FTC -- Fedora Technical Committee (the final name still > needs to be discussed, there are some alternatives floating around). > That makes 15 Members for now. > - the FTC sits below the Board and handles all the day to day work and > the details around the packages and the releases; some of that work get > delegated to be done by SIGs (Extras has some already that simply > continue to exist; further create SIGs [even if that means only one or > two people] for each Spin). Big or long term decisions (roadmaps for > example) get worked out together by the FTC and the Board. > - Those four members that are in the Packaging Committee and in FESCo > now will represent the Packaging Committee in FTC, too. The Packaging > Committee at the same time becomes a SIG. It will work as before; > similar how FESCo and the Core Cabal had veto power it's now FTC that > can block decisions. > - the FTC will run with a similar scheme how FESCo ran until now (e.g. > in the open, public meetings, ...) > - the FTC in the current form has a lifespan until six weeks after F7 > was released. The FTC until then has to work out a plan how FTC will be > constituted after that (e.g. size, seats [some elected, other > appointed?], fixed level of community/red hat members, ...) together > with the Board. FYI: FESCo in its meeting today agreed that this is the route it wants to take (notting and jesse were around, too, and did not yell). Could the board please look at the proposal, discuss it in its tomorrows meeting and ACK it or if it dislikes parts give feedback if it wants to see something changed? BTW, we did not agree on a final name yet. I'll ask fedora-devel-list and fedora-extras-list for name suggestions. FESCo will discuss the name issue in this weeks meeting and push hard to find the final name in next weeks meeting. CU thl From mmcgrath at fedoraproject.org Mon Jan 15 17:52:23 2007 From: mmcgrath at fedoraproject.org (Mike McGrath) Date: Mon, 15 Jan 2007 11:52:23 -0600 Subject: move mailinglists to a new server (for example something like lists.fedoraproject.org) In-Reply-To: <45ABB3A0.20304@fedoraproject.org> References: <45A7A325.7080906@leemhuis.info> <45AB32B5.7070008@fedoraproject.org> <1168848511.21489.38.camel@cutter> <45AB35D1.8010302@fedoraproject.org> <1168876061.21489.44.camel@cutter> <1168877400.5102.33.camel@zod.rchland.ibm.com> <1168879010.21489.48.camel@cutter> <45ABACD7.6010606@fedoraproject.org> <1168880853.21489.50.camel@cutter> <45ABB3A0.20304@fedoraproject.org> Message-ID: <3237e4410701150952w2f192126o93fa23546430eee1@mail.gmail.com> On 1/15/07, Rahul Sundaram wrote: > seth vidal wrote: > > On Mon, 2007-01-15 at 22:03 +0530, Rahul Sundaram wrote: Just so I can keep up with this, is there a pretty strong motivation to move forward with this or is the FAB in general just sort of opposed or sort of in favor of it? -Mike From sundaram at fedoraproject.org Mon Jan 15 21:16:52 2007 From: sundaram at fedoraproject.org (Rahul Sundaram) Date: Tue, 16 Jan 2007 02:46:52 +0530 Subject: move mailinglists to a new server (for example something like lists.fedoraproject.org) In-Reply-To: <3237e4410701150952w2f192126o93fa23546430eee1@mail.gmail.com> References: <45A7A325.7080906@leemhuis.info> <45AB32B5.7070008@fedoraproject.org> <1168848511.21489.38.camel@cutter> <45AB35D1.8010302@fedoraproject.org> <1168876061.21489.44.camel@cutter> <1168877400.5102.33.camel@zod.rchland.ibm.com> <1168879010.21489.48.camel@cutter> <45ABACD7.6010606@fedoraproject.org> <1168880853.21489.50.camel@cutter> <45ABB3A0.20304@fedoraproject.org> <3237e4410701150952w2f192126o93fa23546430eee1@mail.gmail.com> Message-ID: <45ABEF44.7030801@fedoraproject.org> Mike McGrath wrote: > On 1/15/07, Rahul Sundaram wrote: >> seth vidal wrote: >> > On Mon, 2007-01-15 at 22:03 +0530, Rahul Sundaram wrote: > > Just so I can keep up with this, is there a pretty strong motivation > to move forward with this or is the FAB in general just sort of > opposed or sort of in favor of it? I am very much in favor of it but we need to know who in the infrastructure team is going to administrate all these mailing lists and associated issues - moderation, spam etc. Make sure you talk to Red Hat team managing this to know much work it takes. I know it is very much a non trivial amount of requirement and we would be reinventing a lot of the infrastructure. If the infrastructure team can't manage the additional workload better than Red Hat team, we should not jump into this. It might be better to just change the domain and let Red Hat manage it then. Rahul From mmcgrath at fedoraproject.org Mon Jan 15 21:49:45 2007 From: mmcgrath at fedoraproject.org (Mike McGrath) Date: Mon, 15 Jan 2007 15:49:45 -0600 Subject: move mailinglists to a new server (for example something like lists.fedoraproject.org) In-Reply-To: <45ABEF44.7030801@fedoraproject.org> References: <45A7A325.7080906@leemhuis.info> <45AB35D1.8010302@fedoraproject.org> <1168876061.21489.44.camel@cutter> <1168877400.5102.33.camel@zod.rchland.ibm.com> <1168879010.21489.48.camel@cutter> <45ABACD7.6010606@fedoraproject.org> <1168880853.21489.50.camel@cutter> <45ABB3A0.20304@fedoraproject.org> <3237e4410701150952w2f192126o93fa23546430eee1@mail.gmail.com> <45ABEF44.7030801@fedoraproject.org> Message-ID: <3237e4410701151349o43e4a218v1f84389468291a97@mail.gmail.com> On 1/15/07, Rahul Sundaram wrote: > Mike McGrath wrote: > > On 1/15/07, Rahul Sundaram wrote: > >> seth vidal wrote: > >> > On Mon, 2007-01-15 at 22:03 +0530, Rahul Sundaram wrote: > > > > Just so I can keep up with this, is there a pretty strong motivation > > to move forward with this or is the FAB in general just sort of > > opposed or sort of in favor of it? > > I am very much in favor of it but we need to know who in the > infrastructure team is going to administrate all these mailing lists and > associated issues - moderation, spam etc. Make sure you talk to Red Hat > team managing this to know much work it takes. I know it is very much a > non trivial amount of requirement and we would be reinventing a lot of > the infrastructure. If the infrastructure team can't manage the > additional workload better than Red Hat team, we should not jump into > this. It might be better to just change the domain and let Red Hat > manage it then. > > Rahul Moderation: I'm hoping to get 'list sponsors' who are not in the infrastructure team to do this when the lists are created. This is similar to the current setup. Spam: we're actually hoping to have a spam filtering SMTP engine up very soon (its up not but not live). I wonder how the red hat folks do this. Workload: I think we can manage the workload no problem. I'm just worried that we're looking at doing it "because we can". With the exception of a few vocal people, everyone else I've heard from has kind of said 'meh'. -Mike Also does anyone know the direct email address of any of the maillist guys (email me privately) I haven't heard back from the generic address. From notting at redhat.com Mon Jan 15 21:57:24 2007 From: notting at redhat.com (Bill Nottingham) Date: Mon, 15 Jan 2007 16:57:24 -0500 Subject: Fedora Board meeting, 2007-01-16, 10AM EST Message-ID: <20070115215724.GB18488@nostromo.devel.redhat.com> (on behalf of Max) The Fedora Board will be having a meeting tomorrow, January 16th, 10AM EST. Currently scheduled to be discussed: - FESCo successor proposal from FESCo - other issues as they warrant Bill From skvidal at linux.duke.edu Mon Jan 15 22:08:09 2007 From: skvidal at linux.duke.edu (seth vidal) Date: Mon, 15 Jan 2007 17:08:09 -0500 Subject: Fedora Board meeting, 2007-01-16, 10AM EST In-Reply-To: <20070115215724.GB18488@nostromo.devel.redhat.com> References: <20070115215724.GB18488@nostromo.devel.redhat.com> Message-ID: <1168898889.21489.60.camel@cutter> On Mon, 2007-01-15 at 16:57 -0500, Bill Nottingham wrote: > (on behalf of Max) > > The Fedora Board will be having a meeting tomorrow, January 16th, > 10AM EST. > > Currently scheduled to be discussed: > - FESCo successor proposal from FESCo > - other issues as they warrant > I have another meeting at work at 10:30am. For some reason I had this board meeting down for 5pm tomorrow, hence the scheduling conflict. I'll be available for the first half of the meeting tomorrow, then I'll have to go. -sv From mspevack at redhat.com Mon Jan 15 23:51:50 2007 From: mspevack at redhat.com (Max Spevack) Date: Mon, 15 Jan 2007 18:51:50 -0500 (EST) Subject: Fedora Board meeting, 2007-01-16, 10AM EST In-Reply-To: <1168898889.21489.60.camel@cutter> References: <20070115215724.GB18488@nostromo.devel.redhat.com> <1168898889.21489.60.camel@cutter> Message-ID: On Mon, 15 Jan 2007, seth vidal wrote: > On Mon, 2007-01-15 at 16:57 -0500, Bill Nottingham wrote: >> (on behalf of Max) >> >> The Fedora Board will be having a meeting tomorrow, January 16th, >> 10AM EST. >> >> Currently scheduled to be discussed: >> - FESCo successor proposal from FESCo >> - other issues as they warrant My guess is that we should use the time to talk about whatever "governance decisions" in general are needed in Fedora, starting with the FESCO successor and continuing with the "core/extras merge" stuff that has been bandied about on-list and in the FESCO meetings. Also want to pick a time during FUDCon for an in-person Board meeting. Won't that be so official! :-) --Max From mspevack at redhat.com Tue Jan 16 00:03:47 2007 From: mspevack at redhat.com (Max Spevack) Date: Mon, 15 Jan 2007 19:03:47 -0500 (EST) Subject: Fedora Board meeting, 2007-01-16, 10AM EST In-Reply-To: <20070115235758.GC433@zoidtechnologies.com> References: <20070115215724.GB18488@nostromo.devel.redhat.com> <1168898889.21489.60.camel@cutter> <20070115235758.GC433@zoidtechnologies.com> Message-ID: >> Also want to pick a time during FUDCon for an in-person Board meeting. >> Won't that be so official! :-) > > complete with a "...and all I got was this lousy t-shirt" shirt? ;) DENIED: lack of budget. You only get a tshirt when your board tenure is over. And it has to be a tank top. Elliot, yours is in the mail. :-) From gdk at redhat.com Tue Jan 16 00:01:38 2007 From: gdk at redhat.com (Greg Dekoenigsberg) Date: Mon, 15 Jan 2007 19:01:38 -0500 (EST) Subject: Fedora Board meeting, 2007-01-16, 10AM EST In-Reply-To: References: <20070115215724.GB18488@nostromo.devel.redhat.com> <1168898889.21489.60.camel@cutter> <20070115235758.GC433@zoidtechnologies.com> Message-ID: On Mon, 15 Jan 2007, Max Spevack wrote: >>> Also want to pick a time during FUDCon for an in-person Board meeting. >>> Won't that be so official! :-) >> >> complete with a "...and all I got was this lousy t-shirt" shirt? ;) > > DENIED: lack of budget. You only get a tshirt when your board tenure is > over. And it has to be a tank top. Elliot, yours is in the mail. :-) Don't worry. Big Daddy G still has his fat community budget. I've already purchased the Red Hat Official Bedazzler, and I will be putting sequins on the tanktops to spell out "Fedora Supahstah". --g ------------------------------------------------------------- Greg DeKoenigsberg || Fedora Project || fedoraproject.org Be an Ambassador || http://fedoraproject.org/wiki/Ambassadors ------------------------------------------------------------- From skvidal at linux.duke.edu Tue Jan 16 00:10:47 2007 From: skvidal at linux.duke.edu (seth vidal) Date: Mon, 15 Jan 2007 19:10:47 -0500 Subject: Fedora Board meeting, 2007-01-16, 10AM EST In-Reply-To: References: <20070115215724.GB18488@nostromo.devel.redhat.com> <1168898889.21489.60.camel@cutter> <20070115235758.GC433@zoidtechnologies.com> Message-ID: <1168906247.21489.68.camel@cutter> On Mon, 2007-01-15 at 19:01 -0500, Greg Dekoenigsberg wrote: > > On Mon, 15 Jan 2007, Max Spevack wrote: > > >>> Also want to pick a time during FUDCon for an in-person Board meeting. > >>> Won't that be so official! :-) > >> > >> complete with a "...and all I got was this lousy t-shirt" shirt? ;) > > > > DENIED: lack of budget. You only get a tshirt when your board tenure is > > over. And it has to be a tank top. Elliot, yours is in the mail. :-) > > Don't worry. Big Daddy G still has his fat community budget. I've > already purchased the Red Hat Official Bedazzler, and I will be putting > sequins on the tanktops to spell out "Fedora Supahstah". > might I suggest you find a hobby, or maybe a selection of pharmaceuticals? :) -sv From skvidal at linux.duke.edu Tue Jan 16 00:16:23 2007 From: skvidal at linux.duke.edu (seth vidal) Date: Mon, 15 Jan 2007 19:16:23 -0500 Subject: Fedora Board meeting, 2007-01-16, 10AM EST In-Reply-To: References: <20070115215724.GB18488@nostromo.devel.redhat.com> <1168898889.21489.60.camel@cutter> Message-ID: <1168906583.21489.71.camel@cutter> On Mon, 2007-01-15 at 18:51 -0500, Max Spevack wrote: > On Mon, 15 Jan 2007, seth vidal wrote: > > > On Mon, 2007-01-15 at 16:57 -0500, Bill Nottingham wrote: > >> (on behalf of Max) > >> > >> The Fedora Board will be having a meeting tomorrow, January 16th, > >> 10AM EST. > >> > >> Currently scheduled to be discussed: > >> - FESCo successor proposal from FESCo > >> - other issues as they warrant > > My guess is that we should use the time to talk about whatever "governance > decisions" in general are needed in Fedora, starting with the FESCO > successor and continuing with the "core/extras merge" stuff that has been > bandied about on-list and in the FESCO meetings. > > Also want to pick a time during FUDCon for an in-person Board meeting. > Won't that be so official! :-) > Also - we should announce which hotel has 'the rate' for fudcon. I'm still waiting to book my hotel and I've not heard anything yet. -sv From gdk at redhat.com Tue Jan 16 00:46:56 2007 From: gdk at redhat.com (Greg Dekoenigsberg) Date: Mon, 15 Jan 2007 19:46:56 -0500 (EST) Subject: Fedora Board meeting, 2007-01-16, 10AM EST In-Reply-To: <1168906583.21489.71.camel@cutter> References: <20070115215724.GB18488@nostromo.devel.redhat.com> <1168898889.21489.60.camel@cutter> <1168906583.21489.71.camel@cutter> Message-ID: On Mon, 15 Jan 2007, seth vidal wrote: > On Mon, 2007-01-15 at 18:51 -0500, Max Spevack wrote: >> On Mon, 15 Jan 2007, seth vidal wrote: >> >>> On Mon, 2007-01-15 at 16:57 -0500, Bill Nottingham wrote: >>>> (on behalf of Max) >>>> >>>> The Fedora Board will be having a meeting tomorrow, January 16th, >>>> 10AM EST. >>>> >>>> Currently scheduled to be discussed: >>>> - FESCo successor proposal from FESCo >>>> - other issues as they warrant >> >> My guess is that we should use the time to talk about whatever "governance >> decisions" in general are needed in Fedora, starting with the FESCO >> successor and continuing with the "core/extras merge" stuff that has been >> bandied about on-list and in the FESCO meetings. >> >> Also want to pick a time during FUDCon for an in-person Board meeting. >> Won't that be so official! :-) >> > > Also - we should announce which hotel has 'the rate' for fudcon. I'm > still waiting to book my hotel and I've not heard anything yet. Probably going to be the Fairmont Copley Plaza. Rate will be $119 or less. Should be finalized tomorrow AM -- stay tuned. Plenty of availability, so no sweat there. --g ------------------------------------------------------------- Greg DeKoenigsberg || Fedora Project || fedoraproject.org Be an Ambassador || http://fedoraproject.org/wiki/Ambassadors ------------------------------------------------------------- From skvidal at linux.duke.edu Tue Jan 16 02:37:54 2007 From: skvidal at linux.duke.edu (seth vidal) Date: Mon, 15 Jan 2007 21:37:54 -0500 Subject: Fedora Board meeting, 2007-01-16, 10AM EST In-Reply-To: References: <20070115215724.GB18488@nostromo.devel.redhat.com> <1168898889.21489.60.camel@cutter> <1168906583.21489.71.camel@cutter> Message-ID: <1168915074.21489.76.camel@cutter> On Mon, 2007-01-15 at 19:46 -0500, Greg Dekoenigsberg wrote: > On Mon, 15 Jan 2007, seth vidal wrote: > > > On Mon, 2007-01-15 at 18:51 -0500, Max Spevack wrote: > >> On Mon, 15 Jan 2007, seth vidal wrote: > >> > >>> On Mon, 2007-01-15 at 16:57 -0500, Bill Nottingham wrote: > >>>> (on behalf of Max) > >>>> > >>>> The Fedora Board will be having a meeting tomorrow, January 16th, > >>>> 10AM EST. > >>>> > >>>> Currently scheduled to be discussed: > >>>> - FESCo successor proposal from FESCo > >>>> - other issues as they warrant > >> > >> My guess is that we should use the time to talk about whatever "governance > >> decisions" in general are needed in Fedora, starting with the FESCO > >> successor and continuing with the "core/extras merge" stuff that has been > >> bandied about on-list and in the FESCO meetings. > >> > >> Also want to pick a time during FUDCon for an in-person Board meeting. > >> Won't that be so official! :-) > >> > > > > Also - we should announce which hotel has 'the rate' for fudcon. I'm > > still waiting to book my hotel and I've not heard anything yet. > > Probably going to be the Fairmont Copley Plaza. Rate will be $119 or > less. Should be finalized tomorrow AM -- stay tuned. Plenty of > availability, so no sweat there. > Cool. That sounds like a great rate! Thanks for getting that taken care of. -sv From fedora at leemhuis.info Tue Jan 16 07:34:51 2007 From: fedora at leemhuis.info (Thorsten Leemhuis) Date: Tue, 16 Jan 2007 08:34:51 +0100 Subject: Fedora Board meeting, 2007-01-16, 10AM EST In-Reply-To: <20070115215724.GB18488@nostromo.devel.redhat.com> References: <20070115215724.GB18488@nostromo.devel.redhat.com> Message-ID: <45AC801B.7010106@leemhuis.info> On 15.01.2007 22:57, Bill Nottingham wrote: > (on behalf of Max) > The Fedora Board will be having a meeting tomorrow, January 16th, > 10AM EST. > Currently scheduled to be discussed: > - FESCo successor proposal from FESCo > - other issues as they warrant - was the naming issue "Fedora Desktop Spin" (with Gnome) vs. "Fedora KDE" ever discussed by the Board? Could it do that in todays meeting please? (?) - I'd like to have some advice (not a final decision, as that depends on further investigation by mmcgrath) from the Board regarding the proposed move of the mailing lists to our own servers. CU thl (?) -- side note: I was against calling it "Fedora Desktop Spin" in the beginning, but I now tend a bit to "Lets call it Fedora Desktop Spin and put in the best software (firefox, Openoffice) around the Gnome desktop. At the same time *clearly communicate* that we call that Spin 'Fedora Desktop' to leave the way open for a real 'Fedora Gnome' spin, that focuses around Gnome-Software (epiphany, Gnome-Office)". The *clearly communicate* IMHO is important. From Christian.Iseli at licr.org Tue Jan 16 10:02:55 2007 From: Christian.Iseli at licr.org (Christian Iseli) Date: Tue, 16 Jan 2007 11:02:55 +0100 Subject: Fedora Board meeting, 2007-01-16, 10AM EST In-Reply-To: <45AC801B.7010106@leemhuis.info> References: <20070115215724.GB18488@nostromo.devel.redhat.com> <45AC801B.7010106@leemhuis.info> Message-ID: <20070116110255.297fb977@ludwig-alpha.unil.ch> On Tue, 16 Jan 2007 08:34:51 +0100, Thorsten Leemhuis wrote: > I was against calling it "Fedora Desktop Spin" in the > beginning, but I now tend a bit to "Lets call it Fedora Desktop Spin and > put in the best software (firefox, Openoffice) around the Gnome desktop. > At the same time *clearly communicate* that we call that Spin 'Fedora > Desktop' to leave the way open for a real 'Fedora Gnome' spin, that > focuses around Gnome-Software (epiphany, Gnome-Office)". The *clearly > communicate* IMHO is important. +1 C From notting at redhat.com Tue Jan 16 15:16:21 2007 From: notting at redhat.com (Bill Nottingham) Date: Tue, 16 Jan 2007 10:16:21 -0500 Subject: Fedora Board meeting, 2007-01-16, 10AM EST In-Reply-To: <20070115215724.GB18488@nostromo.devel.redhat.com> References: <20070115215724.GB18488@nostromo.devel.redhat.com> Message-ID: <20070116151621.GA28194@nostromo.devel.redhat.com> Bill Nottingham (notting at redhat.com) said: > (on behalf of Max) > > The Fedora Board will be having a meeting tomorrow, January 16th, > 10AM EST. > > Currently scheduled to be discussed: > - FESCo successor proposal from FESCo > - other issues as they warrant Due to lack of quorum, we will attempt to reconvene at 5PM EST, today. Bill From gdk at redhat.com Tue Jan 16 16:06:36 2007 From: gdk at redhat.com (Greg Dekoenigsberg) Date: Tue, 16 Jan 2007 11:06:36 -0500 (EST) Subject: Fedora Board meeting, 2007-01-16, 10AM EST In-Reply-To: <20070116151621.GA28194@nostromo.devel.redhat.com> References: <20070115215724.GB18488@nostromo.devel.redhat.com> <20070116151621.GA28194@nostromo.devel.redhat.com> Message-ID: I'll definitely be there. --g On Tue, 16 Jan 2007, Bill Nottingham wrote: > Bill Nottingham (notting at redhat.com) said: >> (on behalf of Max) >> >> The Fedora Board will be having a meeting tomorrow, January 16th, >> 10AM EST. >> >> Currently scheduled to be discussed: >> - FESCo successor proposal from FESCo >> - other issues as they warrant > > Due to lack of quorum, we will attempt to reconvene at 5PM EST, today. > > Bill > > _______________________________________________ > fedora-advisory-board mailing list > fedora-advisory-board at redhat.com > http://www.redhat.com/mailman/listinfo/fedora-advisory-board > -- ------------------------------------------------------------- Greg DeKoenigsberg || Fedora Project || fedoraproject.org Be an Ambassador || http://fedoraproject.org/wiki/Ambassadors ------------------------------------------------------------- From mspevack at redhat.com Tue Jan 16 18:08:13 2007 From: mspevack at redhat.com (Max Spevack) Date: Tue, 16 Jan 2007 13:08:13 -0500 (EST) Subject: Fedora Board meeting, 2007-01-16, 10AM EST In-Reply-To: <20070116110255.297fb977@ludwig-alpha.unil.ch> References: <20070115215724.GB18488@nostromo.devel.redhat.com> <45AC801B.7010106@leemhuis.info> <20070116110255.297fb977@ludwig-alpha.unil.ch> Message-ID: On Tue, 16 Jan 2007, Christian Iseli wrote: > On Tue, 16 Jan 2007 08:34:51 +0100, Thorsten Leemhuis wrote: >> I was against calling it "Fedora Desktop Spin" in the >> beginning, but I now tend a bit to "Lets call it Fedora Desktop Spin and >> put in the best software (firefox, Openoffice) around the Gnome desktop. >> At the same time *clearly communicate* that we call that Spin 'Fedora >> Desktop' to leave the way open for a real 'Fedora Gnome' spin, that >> focuses around Gnome-Software (epiphany, Gnome-Office)". The *clearly >> communicate* IMHO is important. > > +1 +1 also From sopwith at gmail.com Wed Jan 17 04:02:26 2007 From: sopwith at gmail.com (Elliot Lee) Date: Tue, 16 Jan 2007 23:02:26 -0500 Subject: Fedora Board meeting, 2007-01-16, 10AM EST In-Reply-To: References: <20070115215724.GB18488@nostromo.devel.redhat.com> <1168898889.21489.60.camel@cutter> <20070115235758.GC433@zoidtechnologies.com> Message-ID: On Jan 15, 2007, at 7:03 PM, Max Spevack wrote: >>> Also want to pick a time during FUDCon for an in-person Board >>> meeting. Won't that be so official! :-) >> >> complete with a "...and all I got was this lousy t-shirt" shirt? ;) > > DENIED: lack of budget. You only get a tshirt when your board > tenure is over. And it has to be a tank top. Elliot, yours is in > the mail. :-) Wow, I've always wanted a genuine wife-beater... :-) Now that I'm in the position to get one, here are random thoughts from a more distanced perspective: Fedora Extras amazes me with how much stuff I find already packaged in it. For Extras, I think we may overestimate the value of package quality and underestimate the value of just having a ton of up-to- date packages, but things are definitely going in the right direction. The biggest headache for me recently has seemed to be hardware support. So far, to get my new storage server working half-decently, I've had to recompile an lm_sensors chip driver with an experimental patch, and download an experimental network driver that works very poorly. I think in the past I underestimated the amount of pain people go through getting their hardware to work. I wonder if there is interest in creating a "Fedora Kernel" sub-project or something, to do things like: - package up 3rd party drivers - improve direct communications with driver developers - create a distributed hardware test grid (participants would download a nightly LiveCD image, boot it on a system with questionable or untested hardware to automatically run a test suite, and report the results back) - and help with the hardware reporting tool that Fedora Infrastructure keeps talking about. (On the other hand, it was so cool to plug in a UPS via USB and see the power applet magically show up on the GNOME panel.) The new updates applet & pup are killer, especially having update info in the pup list. In other words, there's still plenty of value to be added by perfecting the basic user experience as opposed to branching out in huge new directions. Best, -- Elliot From blizzard at redhat.com Wed Jan 17 09:33:09 2007 From: blizzard at redhat.com (Christopher Blizzard) Date: Wed, 17 Jan 2007 04:33:09 -0500 Subject: Fedora Board meeting, 2007-01-16, 10AM EST In-Reply-To: References: <20070115215724.GB18488@nostromo.devel.redhat.com> <1168898889.21489.60.camel@cutter> <20070115235758.GC433@zoidtechnologies.com> Message-ID: <45ADED55.7010902@redhat.com> Elliot Lee wrote: > Fedora Extras amazes me with how much stuff I find already packaged in > it. For Extras, I think we may overestimate the value of package quality > and underestimate the value of just having a ton of up-to-date packages, > but things are definitely going in the right direction. Yeah, it's huge. (Quantity, not quality?) But it's important to be able to say that we have high quality packages, too. If nothing else, but for marketing reasons. > The biggest headache for me recently has seemed to be hardware support. > So far, to get my new storage server working half-decently, I've had to > recompile an lm_sensors chip driver with an experimental patch, and > download an experimental network driver that works very poorly. I think > in the past I underestimated the amount of pain people go through > getting their hardware to work. I wonder if there is interest in > creating a "Fedora Kernel" sub-project or something, to do things like: > - package up 3rd party drivers > - improve direct communications with driver developers > - create a distributed hardware test grid (participants would > download a nightly LiveCD image, boot it on a system with questionable > or untested hardware to automatically run a test suite, and report the > results back) Yep. I've basically been pushing for two big things: 1. Awesome hardware reporting. Including statistics on suspend/resume rates. I attended mjg59's talk today about supporting suspend + resume. It was pretty eye opening and I think there are some interesting things we could do there. 2. Doing something interesting with how we submit patches, share information and generate packages. Use cases would be "I have the kernel with this patch and I need to know if it works for other people as well." So with one click they could generate an rpm for others and let them know they exist. Same with watching: that I could watch what davej is doing and try out his code very easily. Those are the two things that help move the needle, imho. > - and help with the hardware reporting tool that Fedora > Infrastructure keeps talking about. (On the other hand, it was so cool > to plug in a UPS via USB and see the power applet magically show up on > the GNOME panel.) Yeah, it's great, isn't it? I think DavidZ got a free UPS out of that deal. :) > > The new updates applet & pup are killer, especially having update info > in the pup list. > > In other words, there's still plenty of value to be added by perfecting > the basic user experience as opposed to branching out in huge new > directions. > Yep, but doesn't mean we shouldn't be investing in the big stuff down the road. --Chris From mmcgrath at fedoraproject.org Wed Jan 17 14:33:23 2007 From: mmcgrath at fedoraproject.org (Mike McGrath) Date: Wed, 17 Jan 2007 08:33:23 -0600 Subject: Fedora Board meeting, 2007-01-16, 10AM EST In-Reply-To: <45ADED55.7010902@redhat.com> References: <20070115215724.GB18488@nostromo.devel.redhat.com> <1168898889.21489.60.camel@cutter> <20070115235758.GC433@zoidtechnologies.com> <45ADED55.7010902@redhat.com> Message-ID: <3237e4410701170633m115fefd6u6ff577794e7ca3ec@mail.gmail.com> On 1/17/07, Christopher Blizzard wrote: > Elliot Lee wrote: > > The biggest headache for me recently has seemed to be hardware support. > > So far, to get my new storage server working half-decently, I've had to > > recompile an lm_sensors chip driver with an experimental patch, and > > download an experimental network driver that works very poorly. I think > > in the past I underestimated the amount of pain people go through > > getting their hardware to work. I wonder if there is interest in > > creating a "Fedora Kernel" sub-project or something, to do things like: > > - package up 3rd party drivers > > - improve direct communications with driver developers > > - create a distributed hardware test grid (participants would > > download a nightly LiveCD image, boot it on a system with questionable > > or untested hardware to automatically run a test suite, and report the > > results back) > > Yep. I've basically been pushing for two big things: > > 1. Awesome hardware reporting. Including statistics on suspend/resume > rates. I attended mjg59's talk today about supporting suspend + resume. > It was pretty eye opening and I think there are some interesting > things we could do there. I've got the hardware profiler (smolt) up and ready for review, we should be able to add tons to it: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=222959 If you come up with a clever way of detecting suspend / resume issues or successes let me know, we can get it in there. I'd assume it'd would go something like "write file, suspend, resume delete file" "on boot, look for written file, if it hasn't been deleted then the last resume failed" ? -Mike From dmalcolm at redhat.com Thu Jan 18 16:01:28 2007 From: dmalcolm at redhat.com (David Malcolm) Date: Thu, 18 Jan 2007 11:01:28 -0500 Subject: Fedora Board meeting, 2007-01-16, 10AM EST In-Reply-To: <3237e4410701170633m115fefd6u6ff577794e7ca3ec@mail.gmail.com> References: <20070115215724.GB18488@nostromo.devel.redhat.com> <1168898889.21489.60.camel@cutter> <20070115235758.GC433@zoidtechnologies.com> <45ADED55.7010902@redhat.com> <3237e4410701170633m115fefd6u6ff577794e7ca3ec@mail.gmail.com> Message-ID: <1169136088.2822.7.camel@localhost.localdomain> On Wed, 2007-01-17 at 08:33 -0600, Mike McGrath wrote: > On 1/17/07, Christopher Blizzard wrote: > > Elliot Lee wrote: > > > The biggest headache for me recently has seemed to be hardware support. > > > So far, to get my new storage server working half-decently, I've had to > > > recompile an lm_sensors chip driver with an experimental patch, and > > > download an experimental network driver that works very poorly. I think > > > in the past I underestimated the amount of pain people go through > > > getting their hardware to work. I wonder if there is interest in > > > creating a "Fedora Kernel" sub-project or something, to do things like: > > > - package up 3rd party drivers > > > - improve direct communications with driver developers > > > - create a distributed hardware test grid (participants would > > > download a nightly LiveCD image, boot it on a system with questionable > > > or untested hardware to automatically run a test suite, and report the > > > results back) > > > > Yep. I've basically been pushing for two big things: > > > > 1. Awesome hardware reporting. Including statistics on suspend/resume > > rates. I attended mjg59's talk today about supporting suspend + resume. > > It was pretty eye opening and I think there are some interesting > > things we could do there. > > I've got the hardware profiler (smolt) up and ready for review, we > should be able to add tons to it: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=222959 > > If you come up with a clever way of detecting suspend / resume issues > or successes let me know, we can get it in there. I'd assume it'd > would go something like "write file, suspend, resume delete file" "on > boot, look for written file, if it hasn't been deleted then the last > resume failed" ? > What's the relationship, if any, with the LHCP project also on hosted.fedoraproject.org? Compare: https://hosted.fedoraproject.org/projects/smolt with: https://hosted.fedoraproject.org/projects/LHCP Both seem to be attempting roughly the same task; though LHCP code seems to be purely client-side. LHCP has a way of "detecting" suspend/resume issues: it asks the user :-) Hope this helps Dave From stickster at gmail.com Thu Jan 18 16:09:17 2007 From: stickster at gmail.com (Paul W. Frields) Date: Thu, 18 Jan 2007 11:09:17 -0500 Subject: Fedora Board meeting, 2007-01-16, 10AM EST In-Reply-To: References: <20070115215724.GB18488@nostromo.devel.redhat.com> <1168898889.21489.60.camel@cutter> <1168906583.21489.71.camel@cutter> Message-ID: <1169136557.16938.5.camel@localhost.localdomain> On Mon, 2007-01-15 at 19:46 -0500, Greg Dekoenigsberg wrote: > Probably going to be the Fairmont Copley Plaza. Rate will be $119 or > less. Should be finalized tomorrow AM -- stay tuned. Plenty of > availability, so no sweat there. Ping: code? -- Paul W. Frields, RHCE http://paul.frields.org/ gpg fingerprint: 3DA6 A0AC 6D58 FEC4 0233 5906 ACDB C937 BD11 3717 Fedora Project Board: http://fedoraproject.org/wiki/Board Fedora Docs Project: http://fedoraproject.org/wiki/DocsProject -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 189 bytes Desc: This is a digitally signed message part URL: From gdk at redhat.com Thu Jan 18 16:11:11 2007 From: gdk at redhat.com (Greg Dekoenigsberg) Date: Thu, 18 Jan 2007 11:11:11 -0500 (EST) Subject: FUDCon hotel info Message-ID: ---------- Forwarded message ---------- Date: Wed, 17 Jan 2007 19:07:27 -0500 From: Sara Jones To: Greg DeKoenigsberg Subject: All set You are all set at the Fairmont Copley Plaza 105 is the rate and people call in and ask for FUDCON rate Let me know if you need anything SJ -- Learn. Network. Experience open source. Red Hat Summit San Diego | May 9-11, 2007 Learn more: http://www.redhat.com/promo/summit/2007 Sara B Jones Events Manager Red Hat dir. 919.754.3700 x44471 fax. 919.754.3705 cell.916.798.4920 sjones at redhat.com From matt at domsch.com Mon Jan 22 19:34:40 2007 From: matt at domsch.com (Matt Domsch) Date: Mon, 22 Jan 2007 13:34:40 -0600 Subject: Linux User & Developer magazine has FC6 DVD Message-ID: <20070122193440.GB30433@domsch.com> In case you missed it, Linux User and Developer magazine out of the UK has a 9GB DVD of Fedora Core 6 and Extras on its cover this month. http://www.linuxuser.co.uk/ -Matt From blizzard at redhat.com Mon Jan 22 19:41:47 2007 From: blizzard at redhat.com (Christopher Blizzard) Date: Mon, 22 Jan 2007 14:41:47 -0500 Subject: Linux User & Developer magazine has FC6 DVD In-Reply-To: <20070122193440.GB30433@domsch.com> References: <20070122193440.GB30433@domsch.com> Message-ID: <45B5137B.6000507@redhat.com> Matt Domsch wrote: > In case you missed it, Linux User and Developer magazine out of the UK > has a 9GB DVD of Fedora Core 6 and Extras on its cover this month. > > http://www.linuxuser.co.uk/ > Holy crap. 9G? Is that, like, everything? --Chris From jkeating at redhat.com Mon Jan 22 19:52:43 2007 From: jkeating at redhat.com (Jesse Keating) Date: Mon, 22 Jan 2007 14:52:43 -0500 Subject: Linux User & Developer magazine has FC6 DVD In-Reply-To: <45B5137B.6000507@redhat.com> References: <20070122193440.GB30433@domsch.com> <45B5137B.6000507@redhat.com> Message-ID: <200701221452.43651.jkeating@redhat.com> On Monday 22 January 2007 14:41, Christopher Blizzard wrote: > Holy crap. ?9G? ?Is that, like, everything? Most likely, yes. -- Jesse Keating Release Engineer: Fedora -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 189 bytes Desc: not available URL: From dennis at ausil.us Mon Jan 22 19:50:46 2007 From: dennis at ausil.us (Dennis Gilmore) Date: Mon, 22 Jan 2007 13:50:46 -0600 Subject: Linux User & Developer magazine has FC6 DVD In-Reply-To: <45B5137B.6000507@redhat.com> References: <20070122193440.GB30433@domsch.com> <45B5137B.6000507@redhat.com> Message-ID: <200701221350.46591.dennis@ausil.us> On Monday 22 January 2007 13:41, Christopher Blizzard wrote: > Matt Domsch wrote: > > In case you missed it, Linux User and Developer magazine out of the UK > > has a 9GB DVD of Fedora Core 6 and Extras on its cover this month. > > > > http://www.linuxuser.co.uk/ > > Holy crap. 9G? Is that, like, everything? its probably just i386 and x86_64 core dvd's and a small selection of extras packages. its certainly not everything sorry if this comes through twice -- ?,-._|\ ? ?Dennis Gilmore, RHCE /Aussie\ ? Proud Australian \_.--._/ ? | Aurora | Fedora | ? ? ? v ? ? From mspevack at redhat.com Tue Jan 23 16:46:04 2007 From: mspevack at redhat.com (Max Spevack) Date: Tue, 23 Jan 2007 11:46:04 -0500 (EST) Subject: fedora board meeting, 2007-01-24 Message-ID: This week's Fedora Board Meeting has been moved to Wednesday 1/24 so that the Board can serenade Max with "Happy Birthday". Actually, we've moved the meeting because we're having a conversation this week with Mark Webbink, Red Hat's uber-lawyer and the guy we ultimately go to with our Fedora legal questions. It will be at 5:00 PM Eastern, which is 22:00 UTC on 2007-01-24. It's going to be interesting to see how much of this we can discuss in real-time on the IRC channel, just because of the nature of some of the topics. We will do our best for the live IRC stuff, and following up afterward. But some of the topics that we will be discussing include: + CLA simplification, especially with regard to fedoraproject.org + OIN, especially regarding Mono and NTFS code + ipw* firmware + some general license questions, with regard to Tom Callaway's audit just an fyi for this list. --Max -- Max Spevack + http://fedoraproject.org/wiki/MaxSpevack + gpg key -- http://spevack.org/max.asc + fingerprint -- CD52 5E72 369B B00D 9E9A 773E 2FDB CB46 5A17 CF21 From wtogami at redhat.com Tue Jan 30 17:24:20 2007 From: wtogami at redhat.com (Warren Togami) Date: Tue, 30 Jan 2007 12:24:20 -0500 Subject: Fedora 7 will ship Nouveau? Message-ID: <45BF7F44.30303@redhat.com> http://www.phoronix.com/scan.php?page=article&item=634&num=1 "and the Fedora Project announcing its intent for including Nouveau in the upcoming Fedora 7 release." I've seen this mentioned in many news articles without references. Huh? Is Nouveau really ready for inclusion? My understanding is that it is nowhere near ready. Warren Togami wtogami at redhat.com From blizzard at redhat.com Tue Jan 30 17:48:51 2007 From: blizzard at redhat.com (Christopher Blizzard) Date: Tue, 30 Jan 2007 12:48:51 -0500 Subject: Fedora 7 will ship Nouveau? In-Reply-To: <45BF7F44.30303@redhat.com> References: <45BF7F44.30303@redhat.com> Message-ID: <45BF8503.9000906@redhat.com> Warren Togami wrote: > http://www.phoronix.com/scan.php?page=article&item=634&num=1 > "and the Fedora Project announcing its intent for including Nouveau in > the upcoming Fedora 7 release." > > I've seen this mentioned in many news articles without references. > > Huh? > > Is Nouveau really ready for inclusion? My understanding is that it is > nowhere near ready. > Dear god no. It's not ready by _any_ stretch. --Chris From tibbs at math.uh.edu Tue Jan 30 17:55:29 2007 From: tibbs at math.uh.edu (Jason L Tibbitts III) Date: 30 Jan 2007 11:55:29 -0600 Subject: Fedora 7 will ship Nouveau? In-Reply-To: <45BF7F44.30303@redhat.com> References: <45BF7F44.30303@redhat.com> Message-ID: >>>>> "WT" == Warren Togami writes: WT> http://www.phoronix.com/scan.php?page=article&item=634&num=1 "and WT> the Fedora Project announcing its intent for including Nouveau in WT> the upcoming Fedora 7 release." ajax has indicated that he wants to include it in some fashion. WT> I've seen this mentioned in many news articles without references. I do recall seeing it on IRC. WT> Is Nouveau really ready for inclusion? My understanding is that WT> it is nowhere near ready. It's nowhere near ready to be the default driver for Nvidia chipsets, of course. But ajax seems excited enough that he wants it to be in the distro. I imagine that it would probably be a very quick moving non-default experimental bit. I suppose it's an open question as to whether experimental, fast-moving packages have a place in a released version of the distro. It seems to me that Nouveau is the kind of thing we want to foster, but whether we actually put the packages in the main repo or just put pointers in the documentation (or something between) is another matter. - J< From jrb at redhat.com Tue Jan 30 18:06:04 2007 From: jrb at redhat.com (Jonathan Blandford) Date: Tue, 30 Jan 2007 13:06:04 -0500 Subject: Fedora 7 will ship Nouveau? In-Reply-To: <45BF8503.9000906@redhat.com> References: <45BF7F44.30303@redhat.com> <45BF8503.9000906@redhat.com> Message-ID: <1170180364.3124.27.camel@localhost.localdomain> On Tue, 2007-01-30 at 12:48 -0500, Christopher Blizzard wrote: > Dear god no. It's not ready by _any_ stretch. It's apparently as good as the nv driver for 2D for many cards. 3D support is not there yet. Adam is pretty enthusiastic about putting it in, and we will include the nv driver as well. Thanks, -Jonathan From jkeating at redhat.com Tue Jan 30 18:24:45 2007 From: jkeating at redhat.com (Jesse Keating) Date: Tue, 30 Jan 2007 13:24:45 -0500 Subject: Fedora 7 will ship Nouveau? In-Reply-To: References: <45BF7F44.30303@redhat.com> Message-ID: <200701301324.45525.jkeating@redhat.com> On Tuesday 30 January 2007 12:55, Jason L Tibbitts III wrote: > It's nowhere near ready to be the default driver for Nvidia chipsets, > of course. ?But ajax seems excited enough that he wants it to be in > the distro. ?I imagine that it would probably be a very quick moving > non-default experimental bit. Much like the i810/intel combo we have now. -- Jesse Keating Release Engineer: Fedora -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 189 bytes Desc: not available URL: From mspevack at redhat.com Tue Jan 30 20:24:04 2007 From: mspevack at redhat.com (Max Spevack) Date: Tue, 30 Jan 2007 15:24:04 -0500 (EST) Subject: Fedora 7 will ship Nouveau? In-Reply-To: References: <45BF7F44.30303@redhat.com> Message-ID: On Tue, 30 Jan 2007, Jason L Tibbitts III wrote: > It's nowhere near ready to be the default driver for Nvidia chipsets, of > course. But ajax seems excited enough that he wants it to be in the > distro. I imagine that it would probably be a very quick moving > non-default experimental bit. This seems reasonable to me. Everyone seems to agree that it's not ready to be the default, but that doesn't preclude us from having in the Fedora package set, and available for install. To me the question is between: 1) putting it on the Fedora Desktop ISO in a non-default capacity 2) just having it in the Fedora repository (what used to be Extras, what now is "all the packages that aren't included on the 'official' ISOs") and let people install it via pup/yum I don't necessarily have a strong opinion one way or the other. But I can't really think of any reason why it shouldn't be a part of Fedora -- it's Free in all the senses that Fedora cares about, and folks like ajax and jrb in our own community are behind it. --Max -- Max Spevack + http://fedoraproject.org/wiki/MaxSpevack + gpg key -- http://spevack.org/max.asc + fingerprint -- CD52 5E72 369B B00D 9E9A 773E 2FDB CB46 5A17 CF21 From davej at redhat.com Tue Jan 30 20:27:45 2007 From: davej at redhat.com (Dave Jones) Date: Tue, 30 Jan 2007 15:27:45 -0500 Subject: Fedora 7 will ship Nouveau? In-Reply-To: <45BF8503.9000906@redhat.com> References: <45BF7F44.30303@redhat.com> <45BF8503.9000906@redhat.com> Message-ID: <20070130202745.GC4793@redhat.com> On Tue, Jan 30, 2007 at 12:48:51PM -0500, Christopher Blizzard wrote: > Warren Togami wrote: > > http://www.phoronix.com/scan.php?page=article&item=634&num=1 > > "and the Fedora Project announcing its intent for including Nouveau in > > the upcoming Fedora 7 release." > > > > I've seen this mentioned in many news articles without references. > > > > Huh? > > > > Is Nouveau really ready for inclusion? My understanding is that it is > > nowhere near ready. > > > > Dear god no. It's not ready by _any_ stretch. Devicescape is also by no means 'ready', yet if we want half decent wireless, it's 'teh future'. nouveau shouldn't be any worse than the nv driver we already ship in most cases (My laptop triggers one edge case where the panel size is misdetected - we patch the nv driver in fedora, but nouveau lacks the same change [or better change] right now). Getting this into F7 is something we should definitly attempt IMO. Dave -- http://www.codemonkey.org.uk From jkeating at redhat.com Tue Jan 30 20:42:29 2007 From: jkeating at redhat.com (Jesse Keating) Date: Tue, 30 Jan 2007 15:42:29 -0500 Subject: Fedora 7 will ship Nouveau? In-Reply-To: References: <45BF7F44.30303@redhat.com> Message-ID: <200701301542.29310.jkeating@redhat.com> On Tuesday 30 January 2007 15:24, Max Spevack wrote: > This seems reasonable to me. ?Everyone seems to agree that it's not ready > to be the default, but that doesn't preclude us from having in the Fedora > package set, and available for install. > > To me the question is between: > > 1) putting it on the Fedora Desktop ISO in a non-default capacity > > 2) just having it in the Fedora repository (what used to be Extras, what > now is "all the packages that aren't included on the 'official' ISOs") and > let people install it via pup/yum > > I don't necessarily have a strong opinion one way or the other. > > But I can't really think of any reason why it shouldn't be a part of > Fedora -- it's Free in all the senses that Fedora cares about, and folks > like ajax and jrb in our own community are behind it. IIRC we can just make it fall out of the xorg-x11-drivers-nv package as just another file. 'nv' would be used by default, but one could change 'nv' to 'nouveau' to test it out. No installing extra packages. -- Jesse Keating Release Engineer: Fedora -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 189 bytes Desc: not available URL: From tibbs at math.uh.edu Tue Jan 30 20:47:08 2007 From: tibbs at math.uh.edu (Jason L Tibbitts III) Date: 30 Jan 2007 14:47:08 -0600 Subject: Fedora 7 will ship Nouveau? In-Reply-To: <200701301542.29310.jkeating@redhat.com> References: <45BF7F44.30303@redhat.com> <200701301542.29310.jkeating@redhat.com> Message-ID: >>>>> "JK" == Jesse Keating writes: JK> IIRC we can just make it fall out of the xorg-x11-drivers-nv JK> package as just another file. Yes, you could, but then if it's going to see frequent revisions (and I believe we all hope it will) then the poor nv users will be getting zero-content updates all the time. - J< From dennis at ausil.us Tue Jan 30 20:53:08 2007 From: dennis at ausil.us (Dennis Gilmore) Date: Tue, 30 Jan 2007 14:53:08 -0600 Subject: Fedora 7 will ship Nouveau? In-Reply-To: References: <45BF7F44.30303@redhat.com> <200701301542.29310.jkeating@redhat.com> Message-ID: <200701301453.09378.dennis@ausil.us> On Tuesday 30 January 2007 14:47, Jason L Tibbitts III wrote: > >>>>> "JK" == Jesse Keating writes: > > JK> IIRC we can just make it fall out of the xorg-x11-drivers-nv > JK> package as just another file. > > Yes, you could, but then if it's going to see frequent revisions (and > I believe we all hope it will) then the poor nv users will be getting > zero-content updates all the time. right now the way x is packaged xorg-x11-drivers would need to require it so all users will get frequent updates with no content for them. same as people who dont have other cards that have been seeing frequent updates. it makes zero difference if its in xorg-x11-drv-nv or xorg-x11-drv-nouveau -- ?,-._|\ ? ?Dennis Gilmore, RHCE /Aussie\ ? Proud Australian \_.--._/ ? | Aurora | Fedora | ? ? ? v ? ? From blizzard at redhat.com Tue Jan 30 21:21:55 2007 From: blizzard at redhat.com (Christopher Blizzard) Date: Tue, 30 Jan 2007 16:21:55 -0500 Subject: Fedora 7 will ship Nouveau? In-Reply-To: <20070130202745.GC4793@redhat.com> References: <45BF7F44.30303@redhat.com> <45BF8503.9000906@redhat.com> <20070130202745.GC4793@redhat.com> Message-ID: <45BFB6F3.3070604@redhat.com> Dave Jones wrote: > On Tue, Jan 30, 2007 at 12:48:51PM -0500, Christopher Blizzard wrote: > > Warren Togami wrote: > > > http://www.phoronix.com/scan.php?page=article&item=634&num=1 > > > "and the Fedora Project announcing its intent for including Nouveau in > > > the upcoming Fedora 7 release." > > > > > > I've seen this mentioned in many news articles without references. > > > > > > Huh? > > > > > > Is Nouveau really ready for inclusion? My understanding is that it is > > > nowhere near ready. > > > > > > > Dear god no. It's not ready by _any_ stretch. > > Devicescape is also by no means 'ready', yet if we want half decent > wireless, it's 'teh future'. nouveau shouldn't be any worse than the nv > driver we already ship in most cases (My laptop triggers one edge case > where the panel size is misdetected - we patch the nv driver in fedora, > but nouveau lacks the same change [or better change] right now). > > Getting this into F7 is something we should definitly attempt IMO. > > Dave > I got the impression from the preso in .au that it really wasn't working at all. As in, completely unconsumable? --Chris From davej at redhat.com Tue Jan 30 21:27:57 2007 From: davej at redhat.com (Dave Jones) Date: Tue, 30 Jan 2007 16:27:57 -0500 Subject: Fedora 7 will ship Nouveau? In-Reply-To: <45BFB6F3.3070604@redhat.com> References: <45BF7F44.30303@redhat.com> <45BF8503.9000906@redhat.com> <20070130202745.GC4793@redhat.com> <45BFB6F3.3070604@redhat.com> Message-ID: <20070130212757.GA16847@redhat.com> On Tue, Jan 30, 2007 at 04:21:55PM -0500, Christopher Blizzard wrote: > I got the impression from the preso in .au that it really wasn't working > at all. As in, completely unconsumable? depends largely on what card you have, and what you expect from it. For eg, nv works and nouveau doesn't on one system I tried (Quadro FX 370M). (works as far as gdm, but dies horribly with DMA hangs shortly after login). On others, it's no better/worse than nv. Dave -- http://www.codemonkey.org.uk From wtogami at redhat.com Wed Jan 31 03:52:39 2007 From: wtogami at redhat.com (Warren Togami) Date: Tue, 30 Jan 2007 22:52:39 -0500 Subject: Fedora 7 will ship Nouveau? In-Reply-To: <200701301453.09378.dennis@ausil.us> References: <45BF7F44.30303@redhat.com> <200701301542.29310.jkeating@redhat.com> <200701301453.09378.dennis@ausil.us> Message-ID: <45C01287.8070001@redhat.com> Dennis Gilmore wrote: > On Tuesday 30 January 2007 14:47, Jason L Tibbitts III wrote: >>>>>>> "JK" == Jesse Keating writes: >> JK> IIRC we can just make it fall out of the xorg-x11-drivers-nv >> JK> package as just another file. >> >> Yes, you could, but then if it's going to see frequent revisions (and >> I believe we all hope it will) then the poor nv users will be getting >> zero-content updates all the time. > right now the way x is packaged xorg-x11-drivers would need to require it so > all users will get frequent updates with no content for them. same as people > who dont have other cards that have been seeing frequent updates. it makes > zero difference if its in xorg-x11-drv-nv or xorg-x11-drv-nouveau > It makes sense as a separate package xorg-x11-drv-nouveau that the user may choose to install from the repo and enable manually. Is this good ajax? Warren Togami wtogami at redhat.com From tibbs at math.uh.edu Wed Jan 31 15:44:37 2007 From: tibbs at math.uh.edu (Jason L Tibbitts III) Date: 31 Jan 2007 09:44:37 -0600 Subject: Fedora 7 will ship Nouveau? In-Reply-To: <45C01287.8070001@redhat.com> References: <45BF7F44.30303@redhat.com> <200701301542.29310.jkeating@redhat.com> <200701301453.09378.dennis@ausil.us> <45C01287.8070001@redhat.com> Message-ID: >>>>> "WT" == Warren Togami writes: WT> It makes sense as a separate package xorg-x11-drv-nouveau that the WT> user may choose to install from the repo and enable manually. On the other hand, if they're in the same package then you can do the sort of switchover that i810/intel is doing, with cards being whitelisted over to the new driver. So I suppose it's completely situation-dependent. I don't personally think that nouveau is to the point where that is a reasonable thing to do now, but judging from the current rate of progress, it's not out of the question to think that it will be feasible within F7's lifetime. - J< From matt at domsch.com Wed Jan 31 17:43:43 2007 From: matt at domsch.com (Matt Domsch) Date: Wed, 31 Jan 2007 11:43:43 -0600 Subject: Fedora 7 will ship Nouveau? In-Reply-To: References: <45BF7F44.30303@redhat.com> <200701301542.29310.jkeating@redhat.com> <200701301453.09378.dennis@ausil.us> <45C01287.8070001@redhat.com> Message-ID: <20070131174343.GA2763@domsch.com> On Wed, Jan 31, 2007 at 09:44:37AM -0600, Jason L Tibbitts III wrote: > >>>>> "WT" == Warren Togami writes: > > WT> It makes sense as a separate package xorg-x11-drv-nouveau that the > WT> user may choose to install from the repo and enable manually. > > On the other hand, if they're in the same package then you can do the > sort of switchover that i810/intel is doing, with cards being > whitelisted over to the new driver. So I suppose it's completely > situation-dependent. I don't personally think that nouveau is to the > point where that is a reasonable thing to do now, but judging from the > current rate of progress, it's not out of the question to think that > it will be feasible within F7's lifetime. then Provides/Obsoletes the -nouveau package by the -nv package at that point. -Matt