23:00:33 <jjmcd> #startmeeting Docs Project Meeting
23:00:33 <zodbot> Meeting started Wed Jun  8 23:00:33 2011 UTC.  The chair is jjmcd. Information about MeetBot at http://wiki.debian.org/MeetBot.
23:00:33 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic.
23:00:41 <jjmcd> #topic Roll Call
23:00:44 * jjmcd 
23:00:49 * Sparks .
23:01:06 * jhradilek is here
23:01:18 <fnadge> here
23:01:37 * nb 
23:02:42 <jjmcd> #topic Last week's action items
23:02:52 <jjmcd> They are so old they are no longer relevant
23:02:59 <jjmcd> #topic Redirects
23:03:07 <jjmcd> This is more of an FYI
23:03:30 <jjmcd> As you may recall, the shift to having publican do the website broke a lot of links
23:03:51 <jjmcd> Back on F14 we decided to make a redirect for release notes that would need to be updated each release
23:04:06 <jjmcd> Well, Nick and I weren't real quick on it this release
23:04:33 <jjmcd> But meanwhile, we got a bug, 667706, that wanted a whole raft of redirects
23:04:54 <Sparks> .bug 667706
23:04:56 <zodbot> Sparks: Bug 667706 All of the "Old Unsupported Releases" release notes links are wrong - https://bugzilla.redhat.com/show_bug.cgi?id=667706
23:05:02 <jjmcd> We went ahead and did them, and it is probably worth mentioning that it might make sense to use those redirected URLs
23:05:22 <jjmcd> That way if we change things again we aren't hunting down more external links
23:05:48 <jjmcd> There is a release notes redirect for every release
23:05:55 <jjmcd> 2 for FC 1 and 2
23:06:28 <jjmcd> (which had separate notes for i386 and x86_64)
23:06:57 <jjmcd> Thanks for the link Sparks
23:07:04 <jjmcd> Any questions about this?
23:07:55 <jjmcd> So basically, instead of changing a redirect each release we need to change a redirect AND add one
23:08:48 <jjmcd> #info Bug 667706
23:09:06 <jjmcd> no questions, moving on ...
23:09:15 <jjmcd> #topic Build system status
23:09:39 <jjmcd> nb, you and rudi were working furiously on a box to publish from and build drafts.  what ever happened to that?
23:10:32 <jjmcd> nb?
23:10:59 <nb> i don't really know, rudi is/was working on it
23:11:07 <Sparks> Was this the project to hook into Bodhi?
23:11:17 <jjmcd> pity, he doesn't seem to be here
23:11:37 <jjmcd> Sparks, no. to do the builds of drafts for L10N and publish to docs.fp.o
23:11:51 <Sparks> Oh
23:12:01 <Sparks> another project that has gone over my head...  :(
23:12:17 <jjmcd> In 14, some docs didn't get drafts to L10N because folks didn't have 24/7 machines they could schedule
23:12:51 <jjmcd> On 15, we REALLY dropped the ball, especially on guides, but RNs were nothing great, either
23:13:25 <jjmcd> You, me and Zach were pretty much all time-challenged
23:13:26 <Sparks> Why can't the new version of Transifex hook into our repos, again?
23:13:43 <jjmcd> Don't get me started ...
23:14:03 <jjmcd> Release Notes are still en-US only
23:14:10 <Sparks> I really don't remember Docs being brought into that discussion when it was made.
23:14:26 <jjmcd> Nope, it was just made
23:14:26 <Sparks> I hate to rehash things but I'm really not happy with the changes that were made.
23:14:41 <jsmith> Sparks: I did everything I could to get docs' attention
23:15:00 <Sparks> jsmith: You have my attention.
23:15:04 <Sparks> :)
23:15:13 <jjmcd> Well, it did come at a bad time.  As I just mentioned, all the experienced folks had other priorities
23:15:21 <jsmith> I agree that the timing was awful
23:15:27 <Sparks> I guess the damage is done...  how can we go about fixing the situation now?
23:15:31 <jsmith> Personally, I tried to get folks to hold off until after F15
23:15:38 <jjmcd> But still, as best I can tell glezos is the only one who knows how it works
23:15:45 <jsmith> In short, it's up to the packagers to push and pull translations now
23:15:53 <jsmith> They don't get pushed or pulled automagically
23:16:03 <Sparks> jjmcd: You coming to SELF?
23:16:07 <jsmith> https://fedoraproject.org/wiki/FAQ_on_migration_to_Transifex.net
23:16:10 <jjmcd> Sparks, nope
23:16:28 <jsmith> Ah, there's rudi :-)
23:16:30 <jjmcd> Yes, and tons of docs on tx, too.  None especially relevant
23:16:31 <rudi> Morning all!
23:16:32 <Sparks> jsmith: I'd really like to talk about this at SELF if I could...
23:16:33 <jjmcd> HEY rudi
23:16:39 <jsmith> Sparks: Absolutely!
23:16:39 <rudi> Sorry I'm late
23:16:40 <jjmcd> we were asking about the build system
23:16:50 <jsmith> Sparks: I'd be happy to make time to talk about it :-)
23:16:52 <rudi> In principle, it's there
23:16:58 <jjmcd> It suddenly seems to have fallen off the radar
23:17:10 <Sparks> jsmith: Perhaps we can get jjmcd on the phone and tie up some other talent and try to figure a way forward.
23:17:22 <rudi> There's a practical bug I haven't been able to solve so far
23:17:33 <jjmcd> If we can make the build system play, the tx things could become a moot point
23:17:36 <Sparks> It just seems that we went from an outdated system to several systems stuck together with duct tape.
23:17:56 <Sparks> So what does the build system do, exactly?
23:18:04 <Sparks> Perhaps I'm not understanding what's happening.
23:18:33 <jjmcd> The idea was to build draft docs on the crontab on a box in the infrastructure
23:18:33 <jsmith> Well, I think we're mixing two different issues here
23:18:55 * jsmith can speak to the Transifex move and possible future scenarios, but knows nothing about a build system
23:19:07 <jjmcd> jsmith, yes, the build system was the topic that led us to the tx questions
23:19:08 <rudi> Primarily; but moreover, to build and publish docs in an automated environment
23:19:26 <rudi> ie; without manually doing all the web.git stuff
23:19:30 <jjmcd> Yes, and if part of that automation is pulling translations, then tx issues disappear
23:19:34 <rudi> Yes
23:19:41 <rudi> I have that working already
23:19:47 <jjmcd> coolness
23:19:50 <rudi> Pulling translations happens automatically
23:20:04 <Sparks> So what needs to happen now to make everything function?
23:20:12 <rudi> Although getting around Transifex's desire to throttle us was a challenge
23:20:43 <rudi> Sparks -- I have a UTF-8 problem with the sqlite database
23:21:20 <rudi> Meaning that tables of contents for languages written in anything but basic Latin characters are getting scrambled
23:21:29 <rudi> Hang on and I'll point you to the result.
23:21:41 <Sparks> rudi: But will I understand the result?
23:21:49 <jjmcd> I seem to recall some magic perl incantations for that
23:22:18 <rudi> Sparks -- you'll be able to see the kind of problem; it's readily apparent even if you don't read the language/script in question
23:23:46 <Sparks> rudi: Okay
23:24:59 <jjmcd> rudi, when I was trying to do a menu I seem to recall running into something similar, not sure if it was the same, and it seems like there was different perl and C behavior with sqlite.  I don't recall the workaround, or even whether I found it
23:25:05 <jjmcd> But I think I may have
23:25:08 <rudi> Uh.... magic has happened!
23:25:22 <rudi> It's working now....
23:25:29 <jjmcd> heh - gotta love it
23:25:42 <rudi> I wonder if nb got in there and fixed it?
23:25:49 <nb> rudi, what is wrong?
23:26:00 <rudi> Absolutely nothing, I'm happy to report! :)
23:26:03 * nb has not touched anything
23:26:10 * nb is not actaully even sure what pt we are using
23:26:17 <rudi> Weird...
23:26:50 <rudi> Anyway, up to just before launch (the last time I looked), this page (for example): http://publictest01.fedoraproject.org/toc.html
23:26:57 <rudi> showed all the non-Latin characters munched
23:27:01 <jjmcd> #info Build system seems to be mysteriously working
23:28:01 <rudi> Is anyone seeing that now? Because it all looks fine to me...
23:28:20 <jsmith> Looks fine to me (other than the fact that the Spanish page shows all English text)
23:28:34 <jjmcd> Hard for me to tell whether assamese is really bengali
23:28:35 <jhradilek> Looks OK to me as well.
23:28:42 <jjmcd> But it looks reasonable
23:29:06 <jjmcd> oh wait, no it isn't
23:29:08 <jhradilek> Wait a minute.
23:29:09 * jsmith sees some Portuguese titles in the Portuguese translation
23:29:12 <jhradilek> Poznámky k vydaniu
23:29:15 <jjmcd> The smaller chars at the top are wrong
23:29:15 <jhradilek> This is wrong.
23:29:32 <rudi> OK; so problem persists :(
23:29:43 <jjmcd> kannada has bunches of squares with numbers
23:30:13 <rudi> Yeah OK; spoke too soon
23:30:20 <rudi> System is *not* working
23:30:48 <rudi> Sorry -- I remember the corruption affecting a lot more stuff and got excited too quickly by the top half of the page
23:30:51 <jjmcd> So there is some sort of character thing going on.  I assume that the sqlite is created with perl?
23:31:09 <rudi> Yeah; and the characters are correct in the DB
23:31:41 <rudi> They're getting munched while rewriting the table-of-content pages
23:32:23 <jjmcd> So basically a perl script reads the db and outputs some html/javascript incorrectly?
23:32:50 <rudi> Yeah
23:33:10 <rudi> The perl script in the %post section of the document package's spec file
23:33:35 <jjmcd> Are the incorrect characters being written by js?
23:34:09 <rudi> Not AFAIK; I don't know that we have js writing anything to those HTML files
23:34:46 <jjmcd> OK, interesting.  So it looks like the same script is writing some chars correctly (the bg ones) and others not
23:36:22 <rudi> Also; this same script works on the Red Hat machines that generate the docs.redhat.com website
23:36:31 <jjmcd> heh
23:36:55 <Sparks> You *think* it works
23:37:20 <rudi> Well, we don't get this corrupted output
23:37:42 <jjmcd> So what should we do next?
23:38:18 <rudi> I'll get more eyes on it
23:38:39 <rudi> It's bound to be something trivial (once it's found)
23:38:42 <jjmcd> This is a script in the rpm when I make an rpm from publican?
23:38:54 <jjmcd> rudi, those are always the hardest to find
23:38:56 <rudi> Yes; in the %post part of the spec file
23:39:03 <jjmcd> I'll take a peek too
23:39:08 <jjmcd> OK, lets go on
23:39:14 <rudi> Ta; any and all help appreciated!
23:39:27 <jjmcd> #topic Publican/pdf status
23:39:36 <jjmcd> ANy update on the F15 PDF ?
23:40:09 <rudi> Only negative results I'm afaid
23:40:20 <Sparks> bug?
23:40:23 <rudi> I can't get FOP working on F15
23:40:37 <rudi> Hang on; I'll grab the number
23:40:56 <jjmcd> Interesting ... I guess I better keep my F14 VM up then
23:41:26 <rudi> I've isolated where the problem is occurring, but the code is extremely convoluted, and I got to the point yesterday afternoon where I'm going to have to appeal upstream for help
23:42:00 <jjmcd> #info rudi is going to upstream on FOP problem in F15
23:42:17 <rudi> (help from any Java packaging experten also welcomed and appreciated)
23:43:16 * jjmcd has heard unkind comments about java packaging but lacks sufficient expertise to validate
23:43:19 <rudi> We would dearly love to drop FOP; but there's no other open-source FO processor out there that' sin the same league...
23:43:36 <rudi> https://bugzilla.redhat.com/show_bug.cgi?id=689930
23:43:50 <jjmcd> #link https://bugzilla.redhat.com/show_bug.cgi?id=689930
23:44:05 <jsmith> rudi: Have you asked the Java SIG in Fedora if they might be able to help?
23:44:26 <rudi> That was one of the groups I was going to contact
23:44:49 <jsmith> OK, please do
23:44:58 <jsmith> I'll try to reach out to some Java folks I know as well
23:45:15 <rudi> It was only last thing yesterday that I was finally in a position to ask intelligent questions about the problem! ;)
23:45:49 <jsmith> Well, that still puts you ahead of me
23:45:59 * jsmith knows very little about Java packaging
23:46:09 <jjmcd> OK, shall we move on then? It's quarter till
23:46:46 <jjmcd> #topic Release Notes
23:46:54 <jjmcd> Just a quick heads-up
23:47:19 <jjmcd> I've drained the bug queues on Release Notes, pushed a new RPM late last week
23:47:31 <jjmcd> ALso updated docs.fp.o
23:47:36 <jjmcd> Questions?
23:48:30 <jjmcd> #topic Guide Status
23:48:48 <jjmcd> jhradilek, have you pretty much been the guy for the User Guide?
23:49:07 <jhradilek> jjmcd: Deployment Guide.
23:49:26 <jhradilek> jjmcd: I volunteered as a QA for the User Guide though. :)
23:49:29 <jjmcd> ahhh - I figured you since you are the only name there that's still around ;-)
23:50:02 <jhradilek> jjmcd: Do we know if it is still maintained?
23:50:15 <jjmcd> Someone, I don't recall who, came on channel and offered some contribution to the UG.  Even provided xml.
23:50:23 <jjmcd> jhradilek, at this point it appears not
23:50:38 <jjmcd> Unless someone can wake laubersm
23:51:11 <jjmcd> OK, not sure what the next move is on that then
23:51:16 <jjmcd> ANything else on Guides?
23:51:20 <jhradilek> :-) I quickly read the guide, and it looks like it shares some topics with the Deployment Guide (network configuration, printer configuration, etc.)
23:51:49 <jjmcd> I think the big barrier is GNOME 3
23:51:57 <rudi> IMHO, the scope of the two books is quite different though
23:52:56 <jhradilek> That is true.
23:53:03 <rudi> UG audience is a basic desktop user who might be coming straight from a Windows background and never touched Linux before; emphasis is on the GUI and automatic, default configurations
23:53:25 <rudi> DG audience is admins and "power users"
23:53:50 <jjmcd> yep - and UG is lots and lots of screen captures
23:53:51 <rudi> I actually think the UG could/should be further simplified
23:55:08 * nb has not seen laubersm lately
23:56:31 <jjmcd> OK, I guess we need to poke Zach to recruit some new writers.  I saw some new folks on the list lately ...
23:56:55 <jhradilek> I see some of them here. :)
23:57:13 <fnadge> :-)
23:57:13 <jjmcd> Apparently holding their tongues (fingers)
23:57:17 <mprpic> I'd be happy to help with a chapter or two, but I'd rather focus on the RMG :)
23:57:27 <jjmcd> ahhh, a possible victim or two
23:57:50 <Sparks> I think dgrift has been working on the SELinux Guide...
23:58:00 * jjmcd still doesn't have a GNOME3 box that isn't a VM
23:58:09 <dgrift> yessir
23:58:16 <jjmcd> good deal
23:58:36 <jjmcd> So we have the SELinux Guide and the RMG being worked on at least
23:58:42 <dgrift> i\\you can expect an email on the docs email list soon asking for help
23:59:08 <jjmcd> dgrift, lots of folks willing to help you get going
23:59:18 <jjmcd> Although the channel has been pretty quiet lately
23:59:29 <dgrift> tat would be awesome, i am determined
23:59:32 <dgrift> that
23:59:34 <jjmcd> Good deal
23:59:35 <Sparks> dgrift: I've been perusing what you've got already...  looks good!
23:59:45 <Sparks> #link http://fedorapeople.org/gitweb?p=domg472/public_git/SELinux_Administration_Guide.git;a=summary
00:00:22 <jjmcd> OK, lets leave a minute or two for general stuff
00:00:34 <jjmcd> #topic Open floor discussion
00:00:41 <jjmcd> Anyone have anything else?
00:01:13 <dgrift> well just that i am looking for help to make my idea work
00:01:24 <dgrift> but i guess its been deiscussed
00:01:24 <Sparks> I'd like to hear from some of the new people...  Anything that you need?  Anything that isn't explained well on the wiki or... ?
00:02:22 <mprpic> Sparks: I've recently went through the whole process of creating and publishing a new guide and didn't come across any troubles ;)
00:02:36 * jjmcd has been trying to generate a list of relevant versus outdated wiki pages on process
00:02:40 <mprpic> good wiki info there then ;)
00:02:51 <jjmcd> mprpic, good deal.  That is encouraging
00:03:08 <mprpic> I updated one wiki page that noted an invalid link to a git repo
00:03:22 <jjmcd> excellent
00:03:40 <Sparks> Awesome
00:04:19 <jjmcd> Anything else?  We've run over our hour but I don't think anyone else is scheduled here
00:04:50 * gholms raises hand
00:05:04 <jjmcd> go ahead
00:05:17 <gholms> Are the packaged release notes going to be updated to reflect the recently-updated memory requirements for installation?
00:05:24 <jjmcd> done
00:05:25 <gholms> IIRC that was fixed online a week or so ago.
00:05:42 <jjmcd> The rpm was actually built before the online was updated
00:05:51 <jjmcd> But it takes some time to make it to the repos
00:05:54 <gholms> Great. Thanks!
00:05:57 <jjmcd> I believe it is there now, tho
00:06:31 <ciupicri> gholms, +1 for that memory thing; I wasted about an hour trying to install F15 on a virtual machine with 512 MB of RAM
00:07:24 <dgrift> by the way not sure if it was mentioned in the release notes
00:07:52 <dgrift> but there was this issue with fedora installs where it didnt install policycoreutils-restorecond
00:08:10 <dgrift> and it was causing issues for all those ubuntu migrators
00:08:33 <dgrift> because it lets $home mislabelled causing chromium to not work
00:08:43 <jjmcd> I'm pretty sure that wasn't in the release notes.  And I was unaware of the ubuntu migrators.
00:09:22 <dgrift> anyways chrome failed to run the tweeter plugins because of this
00:09:27 <dgrift> so we got a lot of reports
00:09:40 <dgrift> fix is to install polictcoreutils-restorecon
00:09:53 <dgrift> relogin to session and restorecon -R -v ~
00:09:53 <jjmcd> ahhh, interesting. Sounds like something for known issues
00:10:39 <dgrift> i believe so , yes
00:10:56 <jjmcd> #action jjmcd to check known issues page for Ubuntu migration / Chrome / policycoreutils issue
00:10:57 <dgrift> although the polictcoreutils-restorecon is being added to comps i believe
00:11:08 <dgrift> but its a bit late i guess
00:11:13 <jjmcd> yeah
00:11:19 <dgrift> see the bugzillas about it for details
00:11:36 <dgrift> or /query me
00:11:45 <jjmcd> Good deal, thanks
00:12:11 <jjmcd> ANything else?  I hate to keep everyone so late.  But I think we covered a lot of territory
00:12:49 <jjmcd> OK, thanks everyone.  Good meeting.  Always good when everyone contributes.
00:13:12 <jjmcd> #endmeeting