FDSCo Meeting 2007-10-07 IRC log

[Paul W. Frields]

12:59 -!- stickster changed the topic of #fedora-meeting to: Fedora Documentation -- Oct 7, 1700-1800 UTC -=- http://fedoraproject.org/wiki/DocsProject
13:00  * stickster checks agenda on wiki to see if there's any
13:00  * couf present
13:00 < jmbuser> John Babich
13:00 < stickster> and moi
13:01 < stickster> OK, we can keep this short and sweet by sticking to the agenda for now
13:01 < jmbuser> +1
13:01 < stickster> Does anyone have an objection to me chairing in quaid's absence?
13:01 < stickster> I know he wouldn't. :-D
13:01 < couf> stickster: have at it
13:01 < jmbuser> go for it
13:02 < stickster> OK, for all intents and purposes, I'm using this as a "catch-up" meeting.
13:02 < stickster> I have missed many of the previous ones due to the scheduling conflict with $DAYJOB, for which I apologize
13:02 < stickster> Water under the bridge.
13:03 -!- stickster changed the topic of #fedora-meeting to: Fedora Documentation -- Oct 7, 1700-1800 UTC -=- http://fedoraproject.org/wiki/DocsProject -=- TOPIC: F8 Docs status
13:03 < couf> we all need some catching up
13:03 < stickster> OK, the "core" documentation is still the release notes and the installation guide, both of which I have kept up with fairly well.  We are on target for all the translation dates.
13:03 < stickster> We are mournfully absent any updates on other stuff AFAIK though
13:04 < stickster> Does anyone have other docs status? DUG? Anything?
13:04 < couf> translation quick start guide has been updated, that's all I know
13:04 < couf> john dropped
13:04 < couf> jmbuser: jinx
13:05 -!- Irssi: #fedora-meeting: Total of 81 nicks [1 ops, 0 halfops, 0 voices, 80 normal]
13:05  * jmbuser is back
13:06  * stickster just filled jmbuser in
13:06 < jmbuser> I am trying to boot up F8T3 to perhaps revive the DUG, but am having trouble
13:07 < jmbuser> Thia has nothing to do with the drops, BTW
13:07 < stickster> sure
13:07 < jmbuser> s / Thia / This /
13:07  * stickster recommends using some sort of virtualization for doing testing
13:08 < stickster> That way you decouple (for the most part) hardware discrepancies from having a test environment available
13:08 < jmbuser> stickster: good idea
13:08 < jmbuser> stickster: Does QEMU work well for that?
13:08 < stickster> I use VMWare (not a FOSS product) because my office pays for it. You also have KVM/qemu, Xen, and VirtualBox I believe
13:09 < stickster> Yes, it's a bit slow but less so for Linux on Linux IIRC
13:09 < stickster> I thought with modern KVM and Xen you can get near-hardware speed for a Linux guest running on a Linux host
13:09 < jmbuser> I'm using QEMU for OLPC image and it works fine with my 2GB RAM
13:09 < stickster> There you go then
13:09 < couf> anyway, major thank you to stickster for keeping up with our core docs
13:10 < jmbuser> +1
13:10 < stickster> Well, I would certainly not object to people jumping in and helping, I don't own them or anything (well, maybe in name only)
13:10 < jmbuser> stickster: I have a big chunk of time in a few days
13:10 < stickster> One way to help is to either watch wiki pages on Docs/Beats/* or subscribe to the relnotes content feed via email to see when changes are made
13:11 < stickster> Then you can simply port in changes to the XML in CVS.  Things are harder when we first capture the content, but if we stay on top of it, incremental changes are very very easy to do, and take only a few minutes a day at most.
13:11 < jmbuser> stickster: Perhaps we can have a private chat (or public) and you can walk me through a few things
13:12 < stickster> Absolutely... I will be around for about an hour after the meeting, then gone for several
13:12 < jmbuser> stickster: great
13:12 < couf> I'm wondering: stickster what do you find the best: editing text in xml, or wiki?
13:13 < stickster> I confess to not loving the wiki.
13:13  * couf thinking about a divide & conqueor idea
13:13 < stickster> For some things it's good, others not so much.
13:13 < stickster> It's good for contributions, not good for upkeep.
13:13 < couf> indeed
13:14 < stickster> Our Plone timeline seems to be stretching out again... Our Plone helper, Jon, decided at some point that he needed Plone 3 for some reason, which dragged in a bunch of other development baggage which is slowing him down AFAICT
13:14 < couf> 1/2 of us try to keep xml and wiki synced, you could edit in xml (as I'm not by far a good editor)?
13:14 < stickster> I don't know why he needed it, but I respect his decision so I guess we won't have Plone for some time.
13:15 < stickster> couf: Actually, the editing is not so hard -- I usually do that on the Wiki where lots of people see the changes
13:15 < jmbuser> +1
13:15 < stickster> Then I take the diff of changes and apply that to XML in CVS
13:15 < stickster> It's not really automated, that was what Plone is supposed to solve for us.
13:15 < couf>  or vice versa :-)
13:16  * couf just thinking something to take some load off your back
13:16 < stickster> My load's not superheavy, but it would help if we had other guides really moving forward.
13:16 < stickster> We seriously need a DUG and an AdminGudie
13:17 < stickster> *Guide
13:17 < stickster> Whcih brings us to the next topic.
13:17 < jmbuser> I'm sticking my neck out for reviving the DUG
13:17 < stickster> That would be super
13:18 < couf> that leaves AG for me :-)
13:18 < stickster> That brings us to...
13:18 -!- stickster changed the topic of #fedora-meeting to: Fedora Documentation -- Oct 7, 1700-1800 UTC -=- http://fedoraproject.org/wiki/DocsProject -=- TOPIC: Activating New Contributors
13:19  * glezos jumps in
13:19 < stickster> Please do!
13:19 < jmbuser> glezos: Greetings
13:19 < stickster> Something I've noticed over the last months is that we've developed many disparate task lists and no way to track any of them effectively.
13:20 < stickster> So many tracking options leads to confusion.  I personally believe simplicity makes it easier for a newcomer to master basic project skills.
13:20 < stickster> That's why I prefer to use Bugzilla, since it's what everyone else in the project uses.
13:21 < stickster> If Fedora as a whole were to switch to something else, I'd do that.
13:21 < stickster> This is just my US $0.02 -- what do you guys think about this issue?
13:22 < couf> I tend to agree, for stuff in CVS
13:22 < jmbuser> Is it usable for tracking doc progress?
13:22 < jmbuser> I'm willing to give a try
13:22 < stickster> Well, that depends on what docs you're talking about.
13:22 < couf> wiki drafting is somewhat strange to keep in Bugzilla, seems to be double work
13:23 < stickster> If a doc is being drafted on the wiki, there's no point in using any bug tracking -- people should just scrawl on the doc on the wiki without permission
13:23 < stickster> The point of the community wiki is for people to free and just mess around.
13:23 < stickster> However, BZ is good for not just CVS doc issue tracking, but also project work tracking too.
13:24 < couf> ah, important distinction +1
13:25 < stickster> So if there are tasks to be done, they can be tracked there as well, against our "project-tracking" component
13:25 < stickster> There's also a generic "docs-requests" component for "I'd like a XXX tutorial"
13:26  * stickster feels like he's lecturing too much for a meeting, would like other people to express their concerns, love, hate, etc... :-)
13:27  * jmbuser is absorbing words of wisdom :-)
13:27 < couf> the thing is, we're going to have to change a lot of wiki-pages again, which will bring confusion, imo
13:28 < couf> otoh we've doing that for say the last 10 months, so way not :)
13:28 < jmbuser> change is good - you go first
13:30  * jmbuser injects wisdom from a Dilbert t-shirt
13:30 < stickster> jmbuser: That's a pretty wise and cutting remark
13:30 < couf> so what would we define as project-tracking stuff?
13:30 < EvilBob> Hi All
13:30 < stickster> Hi EvilBob!
13:30 < jmbuser> Hi
13:30 < stickster> Project tracking, hmmm.... things like the work currently shown on http://fedoraproject.org/wiki/DocsProject/Tasks?
13:31 < stickster> Some of that needs to be culled.
13:31 < stickster> We've been vigilant in the past about hoarding all these tasks and not forgetting any of them.
13:31 < stickster> However, almost none of them are being done.
13:31 < couf> true that
13:32 < jmbuser> +1
13:32 < stickster> Here's a suggestion I'm going to throw on the table for you guys to chew up and spit out: Take all tasks not being done and throw them away.
13:32 < stickster> We eliminate all these pages and start fresh in some way that everyone agrees on.
13:32 < jmbuser> Oh goody, a purge :-)
13:32 < stickster> OK, tear it apart ye scurvy dogs!
13:33 < jmbuser> Should we wait for the meeting to be over?
13:33 < stickster> Well, I mean, let's tear apart that idea. 
13:33 < stickster> Assume that idea sucks and that I'm a moron for suggesting it.
13:33 < stickster> (Probability studies show you'd be right, I'm thinking.)
13:33 < jmbuser> oh
13:34 < stickster> So what makes that idea suck?
13:34 < jmbuser> I second the motion for a clean slate
13:34 < couf> the fact that we "know" what tasks we need to do
13:34 < couf> and we might forget a tn
13:34 < couf> s/tn/ton
13:35 < stickster> jmbuser: Nope, you don't get off that easy :-)  You have to rip me a new one!
13:35 < jmbuser> ok, prioritize
13:35 < jmbuser> 1. Release Notes
13:36 < couf> damm, had an answer ready for him :-)
13:37 < couf> the real problem here is getting people to understand what the task actually is about
13:38 < stickster> couf: So we're not communicating well in terms of explaining how to do something?
13:39 < couf> stickster: yeah, my 0.02cts
13:39 < stickster> Yup, I'll pay you US $1 for it too
13:39 < stickster> Oops jmbuser appears to have dropped again
13:40 -!- Irssi: #fedora-meeting: Total of 83 nicks [1 ops, 0 halfops, 0 voices, 82 normal]
13:40 < couf> the place where we put the info isn't important, but more the *how-to* behind it
13:40 < stickster> couf: So, I'm going to argue with you about this -- not because you're wrong, but to see if there's more we're not exploring
13:41 < couf> sure
13:41 < stickster> Is a task worth doing in a community project, if no one wants to do it?
13:42 < couf> who knows? at any time someone could jump in and get it done, does that mean we should erase the task?
13:43 < stickster> Right, but if they need to know a bunch of other information to perform that task, and that's not documented either, how serious are we about wanting that task done?
13:43 < stickster> Especially since we are the docs project :-)
13:44 < couf> indeed, that means we're defining "this-needs-to-be-done-once" type of tasks but we don't really know how to start it ourselves?
13:45 < stickster> Right
13:45  * jmbuser won't be doing the minutes
13:45 < stickster> haha!
13:45 < couf> in other words, is everyone in docs on the same page?
13:45 < stickster> That's a very good question. The list is generally where we try and argue these things out, but it's not happening of late.
13:46 < stickster> We should maybe take this philosophical argument there where people can chew on it.
13:46 < jmbuser> +1
13:46 < stickster> I might be tempted to argue that "this needs to be done" without anyone knowing how, or by whom, is effectively clogging up our task list.
13:47 < couf> that's what I'm thinking too
13:47 < stickster> If we want to encourage people to join and help, there are two things we should pursue: 1. easier task finding with explanations, and 2. fewer tasks that mean more.
13:47 < stickster> sorry, 2. fewer, more meaningful tasks..
13:48 < stickster> So maybe what we need is selective culling
13:48 < stickster> Find the things that are nebulous and unlikely to pick up steam, and drop them for now.
13:48 < couf> right
13:48 < stickster> OK, this makes good fodder for the list, anyone object to taking it up there/
13:48 < jmbuser> Did anyone catch what I was saying before or did I get cuf off by my ISP?
13:49 < jmbuser> s / cuf / cut /
13:49 < stickster> jmbuser: We saw 1. Release Notes
13:49 < jmbuser> 2. IG
13:49 < jmbuser> 3. Admin Guide
13:49 < jmbuser> 4. DUG
13:49 < jmbuser> and pretty much stay focused on them for now.
13:49 < jmbuser> Comments?
13:50 < couf> hmm, we've been trying that approach for some time now, is it working?
13:50  * couf bites the idea
13:50 < stickster> That depends on the value of "working"
13:51 < jmbuser> Release Notes are pretty amazing
13:51 < stickster> Community projects succeed when people work on what they enjoy, and are effective at it
13:51 < couf> working as in "getting results"
13:51 < couf> ah right, good point
13:51 < stickster> Think about how a meeting runs.  If everyone spends the whole meeting tossing out a thousand different ideas without ownership or direction, you end up with a crappy meeting
13:52 < stickster> If we concentrate on doing what we're doing well, and work on that to make it so good that people want to help, then we can branch out
13:52 < jmbuser> +1
13:52 < stickster> I wonder if our zeal made us start tossing around too many ideas to ever get done, and then we try and do them all, and when we fail, we hit ourselves over the head and think we're doing a terrible job
13:53 < stickster> We need to start encouraging a culture of success in our project. 
13:53 < stickster> If that means cutting back on our goals, then we should do that.
13:53 < couf> definitly
13:53 < jmbuser> agreed
13:54 < stickster> Interestingly, quaid and I both got mad at Rahul when he suggested this exactly -- not because what he said was wrong, but in the way he said it.
13:54 < couf> rephrazing can have a miraculous result
13:54 < couf> and there goes jmbuser
13:55 -!- Irssi: #fedora-meeting: Total of 83 nicks [1 ops, 0 halfops, 0 voices, 82 normal]
13:56 < stickster> Well, I would encourage people to work on our DUG and AdminG then
13:56 < stickster> Don't fib to yourselves that it's not hard work, it's very hard work to just write these docs
13:56 < stickster> And if it's half-done, we're going to publish it anyway, and let people file bugs.
13:56 < jmbuser> I can attest to that
13:57 < stickster> And when the bugs start rolling in, that gives us the ideas for what to do next.
13:57 < stickster> That's helped me a lot with the IG
13:57 < jmbuser> +1 on half-baked, serve it anyway
13:57 < jmbuser> maybe a better cook will jump in
13:57 < stickster> We'll check, of course, that what's in there is technically correct -- but if there are topics left out, so be it
13:58 < couf> right, I think we can agree on this, seems fair enough
13:58 < jmbuser> +1
13:58  * jmbuser orders dinner
13:58 < stickster> We had training on the agenda, but I think there's a group after us in this room
13:58 < stickster> couf: Can we keep baking this idea until next time?
13:58 < couf> Yeah sure
13:59 < couf> we've got half an hour to go
13:59 < jmbuser> recipe sounds delicious
13:59 < stickster> Oh!
13:59 < couf> jmbuser: you're welcome to cook :)
14:00 < stickster> Well, we can continue if you guys want
14:00 < jmbuser> I'm digging in
14:00 < jmbuser> stickster: re-reading Fogle's Producing OSS - very sound - recommend it
14:01 < stickster> Yup, I read and reviewed it for RHM last year
14:01 < stickster> lemme see if I can find that link....
14:01 < jmbuser> His comments on value of docs is motivating to me
14:02 < couf> small interjection: 17UTC: good time for meeting?
14:02 < stickster> http://www.redhat.com/magazine/015jan06/features/review_fogel/
14:02 < jmbuser> Matt Asay also is stressing importance of good docs to the success of a project
14:02 < stickster> Works for me, I know this was touchy for quaid but he said he would aquiesce if it worked for everyone else
14:02 < jmbuser> OK for me
14:03 < stickster> Interesting thing -- you can find some decent docs for some parts of Fedora project work, like packaging and such
14:03 < jmbuser> I'll read your review, Stickster
14:03 < couf> all right, continue?
14:04 < stickster> Let's talk about training then
14:04 < jmbuser> proceed
14:04 -!- stickster changed the topic of #fedora-meeting to: Fedora Documentation -- Oct 7, 1700-1800 UTC -=- http://fedoraproject.org/wiki/DocsProject -=- TOPIC: Training SIG
14:04  * couf digs up link
14:04 < couf> http://fedoraproject.org/wiki/BartCouvreur/Training
14:04 < stickster> Yup
14:04 < couf> as well as: http://couf.livejournal.com/1424.html
14:05 < jmbuser> oh-oh, I can't Moodle - blasted ISP blocking it
14:05 < stickster> There are a couple things we need to be aware of in looking at the training issue -- 
14:05 < stickster> 1. Training is as difficult to maintain as Docs
14:05 < stickster> 2. Plus, it needs infrastructure to back it
14:06 < stickster> These are NOT arguments against Training.
14:06 < stickster> I dig it, especially since I have done that sort of work "IRL" before
14:06 < couf> right, major blocker seems to be #1
14:06  * jmbuser did public seminars a while back
14:07 < stickster> I think these courses are a heck of a lot better than things like screencasts, which generally are done because someone found a tool for doing screencasts
14:07 < stickster> Screencasts are a maintenance nightmare.
14:07 < stickster> On the other hand, they are VERY easy to make, meaning if you restrict the topic to a two-minute window, who cares about maintenance?
14:07 < couf> true
14:08 < stickster> Courses on the other hand, are much better for growth and modularity
14:08 < stickster> Plus they can be translated!
14:08 < stickster> (I would think)
14:08 < couf> think / hope
14:08 < couf> :)
14:08  * stickster hasn't looked to see if moodle allows PO for example
14:09 < couf> the big plus for training is that we don't need to write up everything, we could distribute it to the project as whole
14:09 < stickster> It would be supercool if you could write a course in en_US, and moodle could create PO for translators to work on
14:09 < stickster> Yup
14:09 < stickster> couf: +10
14:09 < stickster> I'm only speaking as someone interested in writing courseware :-)
14:10 < couf> hehe
14:10 < stickster> So the localization is important from a project perspective; interested locales could put their own training in the system
14:10 < stickster> But anyone could write courses, and then the good ones could bubble up to the top?
14:11 < couf> yeah
14:11 < jmbuser> Do any upstream project do similar courses which can serve as a model?
14:11 < vpv> I did Moodle stuff about a year ago and I think it didn't have the possibility of translating courses with PO then. Things may have changed, though.
14:13 < stickster> Well, that's a point for improvement, at least
14:13 < littlecharly> Hi
14:14 < stickster> jmbuser: I think Bart had posted a couple links
14:14 < couf> I'm afraid most LMS systems lack that functionality
14:14 < stickster> Yeah, that's a shame.
14:14 < jmbuser> ok, thanks
14:14 < couf> jmbuser: I know one, but it's not really upstream for us
14:15 < couf> anyway, we should move from here, the fedora-fr meeting is upon us
14:15 < stickster> Yeah, understood.
14:15 < jmbuser> ok
14:15 < stickster> Let's wrap for today, OK?
14:15 < jmbuser> +1
14:15 < couf> all right
14:15 < stickster> Thanks everyone for coming to the Docs Project meeting!