[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]

Release Notes Meeting 2009-07-23 IRC log



15:02:17 <Sparks> #startmeeting Release Notes Meeting
15:02:22 <Sparks> Who is here?
15:02:23 * Sparks 
15:02:25 * zoglesby is here
15:02:27 * rudi is here
15:02:42 * bcotton is here
15:02:57 * jjmcd is taking a break from spending the state's money
15:03:23 * laubersm found you
15:04:54 <Sparks> Okay...  I'd like to discuss the following:
15:05:08 * laubersm goes to refill coffe mug before settling in for this hour of fun
15:05:29 <Sparks> 1. How to improve on the F12 RNs
15:05:43 <Sparks> 2. What products need to be generated.
15:05:55 <Sparks> 3. Using Publican for the RNs
15:06:30 <Sparks> 4. Training people to write beats.
15:06:44 <Sparks> laubersm: Hour?  We've got this channel all day.
15:07:08 <Sparks> Anything else we need to discuss?
15:07:14 * zoglesby get coffe IV
15:07:38 * laubersm found only dregs and was too lazy to make more - but returned to this fun anyways
15:08:02 <jjmcd> Probably we should discudd who the audience is before attempting to answer those questions
15:08:10 <Sparks> jjmcd: yes
15:08:12 <jjmcd> sheesh the typing
15:08:58 <Sparks> anything else?
15:09:17 <rudi> Packaging
15:09:32 <rudi> As in, when and what
15:09:34 <jjmcd> Influenced by 2 of course
15:09:40 <rudi> Yeah
15:10:29 <Sparks> Okay, let's get started...
15:10:41 <Sparks> #topic Audience
15:10:48 <jjmcd> I see 3
15:10:57 <jjmcd> People who just installed and want to know what's new
15:11:06 <jjmcd> People preparing to install
15:11:12 <jjmcd> People looking for a new feature or fixed bug
15:11:24 <Sparks> Wouldn't the second one be the IG?
15:11:35 <jjmcd> Not necessarily
15:11:51 <jjmcd> For example, database changes almost always require you to do something before install
15:11:52 <Sparks> errr... shouldn't it be included in the IG
15:11:57 <jjmcd> those won't be in the IG
15:12:00 <Sparks> True
15:12:11 <Sparks> So more of an upgrade guide
15:12:29 <zoglesby> I think there is one more
15:12:33 <jjmcd> Yeah, almost anything that needs a backup or something before upgrade needs to be captured
15:12:54 <zoglesby> People that hear about Fedora and are trying to see what it has to offer
15:13:17 <zoglesby> I often look at release notes when I hear about a new distro
15:13:33 <jjmcd> Yeah, probably looking for more detail than the announcement
15:13:37 <Sparks> Okay, so I see this as:
15:14:00 <Sparks> The BIG RNs goes to the third group.
15:14:10 <Sparks> A "Welcome to Fedora ##" for the last
15:14:44 <Sparks> The "pretty" RN (from Marketing) for the first
15:14:47 <jjmcd> I think welcome can also address the first group in an introductory way
15:14:51 <Sparks> yes
15:15:14 <jjmcd> good point.  Perhaps marketing should do the welcome to fedora xx
15:15:29 <Sparks> definitely
15:17:13 <Sparks> So we are already working on the "pretty" version with Marketing.
15:17:34 <Sparks> And we currently have the BIG RNs
15:17:35 <laubersm> I generally find most "users" - new or upgrading - think of wiki and IG and UG as resources more than the RN - they tend to think Release Notes are for the geeks
15:17:46 <Sparks> So we need to figure out the Upgrade guide
15:17:49 <laubersm> The RN can include links to those general resources - and should
15:18:01 <laubersm> UG = users guide
15:18:03 <Sparks> agreed
15:18:24 <laubersm> do we really need a separate upgrade guide?  can't it be in the IG?
15:18:27 <jjmcd> So the beat writers will need to be upgrade writers too, or at least help guide the upgrade writers
15:18:51 <jjmcd> laubersm, I doubt if the experienced user will look at the install guide
15:18:53 <Sparks> Well, I think the IG should have information on upgrading
15:19:15 <rudi> FWIW, the IG already details several different upgrade paths
15:19:47 <laubersm> jjmcd, I was thinking more of the reminders of good general practices - RN would still be for the "this time it is bar than needs to be reconfigured instead of foo"
15:19:58 <jjmcd> But does it include things like "save your qle database to sql before upgrading and reimport after the upgrade"?
15:20:25 <jjmcd> Yes, I don't see the upgrade guide as including all the version specific details
15:20:32 <rudi> jjmcd -- no; and I think that's probably out of scope
15:20:33 <laubersm> jjmcd, if that is an everytime you upgrade the db than it should be in a general guide and not repeated with every RN
15:21:09 <Sparks> So beat writers talk about new stuff and also issues to upgrading
15:21:27 <jjmcd> Yes, but most of the time you don't need to, seems like every third or fourth time with MYSQL, Postgres, and some others there are special things to do
15:21:39 <jjmcd> But I agree, if its every time, then the upgrade guide
15:22:27 <jjmcd> I see beat writers as listing all changes, but providing prose for "important" changes and upgrade issues
15:23:15 <jjmcd> That way the change that you care about, but I don't see as important, is at least noted
15:23:48 <laubersm> remind me - who makes up the beat writers?  some are docs team members but aren't some from other teams that just help with RN?
15:24:08 <jjmcd> Mostly docs team members but not always
15:24:25 <jjmcd> We generally seem to do better with team members
15:24:43 <laubersm> I ask because maybe docs members - or people more familiar with the many guides - should be reviewing to make sure RN stays trim and important stuff gets rediurected to the guides
15:24:57 <laubersm> it seems to me that we just haven't had time for that in the past -
15:25:02 <laubersm> even though the intent has been good
15:25:24 * laubersm prefers reviewing to writing by the way
15:25:25 <jjmcd> I'm hoping to have more lists and less prose this time around
15:25:35 <jjmcd> Should give us more opportunity to review
15:26:21 <jjmcd> And less work for translators
15:26:37 <laubersm> and will the review period be as free form as last time or will it be a bullet on the docs team schedule (or in a wiki chart)?
15:27:09 <jjmcd> I hope to take poelcat's schedule and add in some docs and l10n bullets
15:27:19 <laubersm> cool.
15:27:35 <jjmcd> We never got the l10n notifications in there and we need that
15:27:40 <Sparks> Yeah, I think we need to make sure we get these beats done well ahead of time
15:28:44 <laubersm> I think if beat writers trade reviewing it will help the overall flow of all the RNs as well - and catch more duplication (or need for general docs tips)
15:28:57 <Sparks> Okay...  I don't want to drag this meeting out...  :)
15:29:00 <Sparks> laubersm: I agree.
15:29:14 <Sparks> So can we say that three "RNs" will make this work?
15:29:30 <jjmcd> Most of that has to do with communications with other teams.  Looking at the F11 beats, the only writer who wasn't a docs member at all was Chitlesh, although Jens I think is in the group, but not at the meetings etc
15:29:42 <jjmcd> what three?
15:29:54 <Sparks> 1) Big RNs
15:29:56 <zoglesby> I think we had so much duplicate stuff last time cause on the last day 3 of of were writing all the stuf that didnt get done
15:29:59 <laubersm> How are we naming these so they catch the correct audience?
15:30:00 <Sparks> 2) "Pretty RN"
15:30:08 <Sparks> 3) Upgrade "Guide"
15:30:23 <jjmcd> You don't see the "pretty RN" as chapter one of big rn?
15:30:44 <Sparks> The "Pretty RN" will be very graphical
15:30:50 <jjmcd> ahhh good
15:30:52 <Sparks> kind of like a "newsletter"
15:30:59 <Sparks> a flyer
15:31:11 <jjmcd> sounds like a bitch to translate
15:31:31 <jjmcd> but we did it with IG if I recall
15:31:46 <rudi> Not as long as people are careful to keep text out of the graphics
15:31:58 <rudi> And rely on captions, or callouts
15:32:18 <rudi> And yeah, the IG has about 70 images in each of about 35 localisations
15:32:35 <Sparks> So are we good with those three?
15:32:47 <jjmcd> question fro rudi
15:33:03 <jjmcd> did you do those with dot or something so that you could translate easily?
15:33:26 <rudi> Yeah, where needed.
15:33:41 <jjmcd> cool
15:33:57 <jjmcd> I like it Sparks
15:34:26 <Sparks> #idea Three "RN" guides: big RN, "pretty" RN, and Upgrade Guide
15:34:30 <rudi> jjmcd -- http://docs.fedoraproject.org/install-guide/f11/en-US/html/s1-guimode-textinterface-x86.html
15:35:31 <Sparks> Okay to move on?
15:35:33 <rudi> Do we have any specific plan to keep down the bigness of the big RN?
15:35:45 <jjmcd> tables
15:36:29 <rudi> Hmmm -- tables might get L10N hating us for whole new reasons :)
15:36:36 <zoglesby> I think that one of the big issues for 11 was the amount of info trans had to deal with if we add 2 more we need to cut some down
15:36:39 <jjmcd> Nothing in the tables to translate
15:36:47 <rudi> Oh OK.
15:37:05 <rudi> Just that elements in lists and tables are very challenging
15:37:09 <jjmcd> Jus package name, old version, new version link
15:37:29 <jjmcd> Specifically intended to not need translation
15:37:38 <zoglesby> lol
15:37:57 <rudi> Without any description of what actually changed?
15:37:59 <zoglesby> I think we can automate that to,
15:38:14 <jjmcd> zoglesby, a lot of the prose was just fluff.  Already automated ;-)
15:38:27 <jjmcd> rudi, only "important" changes
15:38:33 <rudi> OK
15:38:49 <jjmcd> https://fedoraproject.org/wiki/User:Jjmcd/Drafts/Fedora_12_tables
15:39:06 <jjmcd> Not organized by beats, and needs a lot of editing
15:39:35 <rudi> sure, but yeah, I get the picture now
15:39:46 <zoglesby> why do we have meetings again? Why not force jjmcd to 6S everything for us
15:40:24 <Sparks> +1
15:40:34 <jjmcd> We still need ppl who can reed and spel gud
15:40:40 <laubersm> zodbot, because last time he built and rpm and left town and it had to be redone before he returned
15:40:53 <laubersm> arg...  that was for zoglesby
15:40:55 <zoglesby> this is more of a fedora change log then release notes
15:41:09 * laubersm relies on tab complete too much without looking
15:41:40 <jjmcd> Well, we do need the change log.  In my view, we still need prose for "important" changes, but this way someone looking for a feature or bugfix can see if it is there
15:41:47 <zoglesby> laubersm: ah yes that was a good day
15:42:09 <jjmcd> ANd that list is long because it is against rawhide.  As we close in on release it will get shorter
15:42:17 <rudi> zoglesby; true to some extent, but then, there was a lot of stuff in previous RNs that was basically "Package X has been upgraded from version 1.7 to 1.9"
15:42:19 <zoglesby> jjmcd: very true, it takes care of one group
15:42:34 <jjmcd> Yes, and the prose takes care of the others
15:42:49 <zoglesby> ok
15:42:58 <jjmcd> rudi, and that prose only provided useless work for translators
15:43:06 <rudi> Yep.
15:43:12 <Sparks> jjmcd: Are you reviewig bug reports for stuff to go into the RN?
15:43:22 <jjmcd> Yes, hit a bunch last week
15:43:33 <laubersm> How about a para on how to use a tool to find that list and not include it at all :)
15:43:51 <jjmcd> One "content" bug I still needs more research
15:44:00 * laubersm thinks such lists reinvent the wheel and will always miss something
15:44:19 <jjmcd> And the remaining RN bugs either need a lot of work or need to wait for something else
15:44:21 <zoglesby> put a link to the fedora community site
15:44:29 <jjmcd> laubersm, probably not a bad idea
15:45:02 <Sparks> jjmcd: Are people actually flagging items to be included in the RNs?
15:45:02 <zoglesby> but there went your 0 trans :P
15:45:13 <jjmcd> Sparks, generally no
15:45:25 <Sparks> jjmcd: Maybe we should encourage that and then query that.
15:45:40 <jjmcd> I wonder how good an idea that is actually
15:45:56 <jjmcd> For the person working on a package, any change is "important"
15:46:47 <jjmcd> Although when you have someone like Chitlesh or Jens who sees the bigger picture, it is great when they can work on the prose
15:46:57 <laubersm> jjmcd, will the list at least be at the end - ie "other updates include:"  so I don't have to scroll through it to get to the prose?
15:47:12 <jjmcd> yes, that was my view - but by beat
15:47:22 <jjmcd> So you get the database prose, then the list
15:47:33 <laubersm> ok
15:47:57 <jjmcd> Problem is, the yum groups are all screwed up, so organizing the list by beat is going to be a bit of a job
15:48:12 <jjmcd> That drafts page is by yum group
15:50:42 <jjmcd> But as you said, perhaps we make a page that gets updated nightly for the list, or make the tool easy enough for people to use
15:51:11 <jjmcd> That way you can see what changed from F11 to NOW rather than to the initial release of 12
15:51:46 <zoglesby> +1
15:52:54 <jjmcd> is there a php interface to sqlite?
15:53:10 * Sparks notes nine minutes left
15:53:45 <jjmcd> Still have most agenda items left
15:53:51 <Sparks> Yes
15:54:08 <jjmcd> yes sqlite or yes agenda?
15:54:13 <Sparks> yes agenda
15:55:08 <jjmcd> What is the issue with Publican?  I thought we had that answered.
15:55:19 <Sparks> Well...  If you can do what you need to do then I'm fine.
15:55:36 <Sparks> I can't get the SRPMs to generate in koji.  They fail
15:55:37 * jjmcd has no issue with Publican
15:55:55 <rudi> The Beta for Publican 1.0 should be announced soon now
15:56:27 <jjmcd> I've been manually building the SRPM, if Publican can do it right, great, but making the RPM isn't a biggie
15:57:10 <rudi> I can't help with the SRPMs, but until 1.0 appears, we now have hackish fixes to most of the glitches in 0.44 up on the wiki
15:57:13 <Sparks> Well, hopefully the "new" version of Publican will fix the problems.
15:57:38 <jjmcd> If we still include N docs in the RN RPM it doesn't matter anyway
15:57:49 <zoglesby> rudi: the fedora wiki or publican wiki?
15:57:55 <rudi> Fedora wiki
15:58:15 <laubersm> Even if there are glitches with the building the SRPM, I have no issues with publican - all items should be created with publican
15:58:24 <rudi> https://fedoraproject.org/wiki/Publican#Publican_on_Fedora_Tips_and_Tweaks
15:58:25 * laubersm did not like the old toolchain adventure
15:58:41 * jjmcd agrees with laubersm
15:59:15 <jjmcd> I'm afraid we need at least another meeting to answer the "what products" question though
15:59:29 <Sparks> jjmcd: That's cool.  Is this time good with everyone?
15:59:46 * Sparks notes that it is awfully late/early for rudi
15:59:48 * rudi isn't crazy about it, but will turn up with enough warning :)
15:59:58 <jjmcd> This time on Thursday is suboptimal but doable
16:00:05 <rudi> 2AM here in the middle of an Australian winter
16:00:18 <zoglesby> rudi: thats me every docs meeting!
16:00:44 <zoglesby> well not the winter part
16:00:49 <rudi> :D
16:01:00 * jjmcd will probably be out of town the next few Wednesdays, tho, so Thursday is better than Wednesday
16:01:03 <zoglesby> and I only wish I was in Australian
16:01:34 <laubersm> I will have a harder time next week - afternoon would be better when students are more likely to be doing labs instead of listening to me talk....
16:01:38 <zoglesby> Sparks: any time is good with me, if I have to I will use my phone and not talk so much
16:02:00 * laubersm is EDT when refering to afternoon
16:02:14 * Sparks is EDT, too
16:02:46 <jjmcd> If you move to 2000Z you are starting to get into a semi-sensible morning time for rudi
16:03:17 * Sparks notes 2000z is drive time for us in EDT
16:03:38 <rudi> Don't worry about me; I'm happy to work around you guys
16:03:52 <Sparks> rudi: Time and a half at 1A?
16:04:06 <rudi> I wish! :)
16:04:08 <zoglesby> lol
16:04:17 <Sparks> I miss shift-differential
16:04:49 <jjmcd> Not me ... back when I got it, I was making $1.82 an hour
16:05:17 <Sparks> Okay, anything else we need to discuss today?
16:05:19 <zoglesby> I missing getting paid for the amount of time I work not working 80 hour weeks and getting the same as everyone else working 40
16:06:21 <rudi> Not discuss, but a thought to leave with people that we should perhaps try to get a package into rawhide rooner rather than later
16:06:41 <Sparks> I agree
16:06:41 <zoglesby> indeed
16:07:00 <laubersm> +1
16:07:14 <rudi> ECS will be able to help out with some of the heavy lifting content-wise, but not so much with the packaging
16:07:40 <rudi> (and hopefully there will be less heavy lifting this time round anyway)
16:08:04 <Sparks> Okay, anything else?
16:08:40 <Sparks> Okay, while we didn't get through everything we did have a good discussion.
16:08:44 <Sparks> Good ideas
16:10:04 <Sparks> Okay, thanks for coming!
16:10:07 <Sparks> #endmeeting


[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]