16:07:00 <abadger1999> #startmeeting fpc
16:07:00 <zodbot> Meeting started Thu Jul 31 16:07:00 2014 UTC.  The chair is abadger1999. Information about MeetBot at http://wiki.debian.org/MeetBot.
16:07:00 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic.
16:07:02 <abadger1999> #meetingname fpc
16:07:02 <zodbot> The meeting name has been set to 'fpc'
16:07:08 <abadger1999> #topic Roll Call
16:07:10 <abadger1999> Who's around?
16:08:11 <abadger1999> spot, geppetto, RemiFedora, SmootherFrOgZ, tibbs|h, Around for FPC?
16:08:18 * RemiFedora here
16:08:26 <abadger1999> #chair RemiFedora
16:08:26 <zodbot> Current chairs: RemiFedora abadger1999
16:09:32 <abadger1999> I'll wait a few to see if more people show up.
16:09:38 <abadger1999> RemiFedora: Question -- will you be at flock?
16:09:45 <RemiFedora> yes!
16:09:48 <abadger1999> excellent.
16:10:02 * geppetto is here
16:10:07 <abadger1999> #chair
16:10:07 <zodbot> Current chairs: RemiFedora abadger1999
16:10:10 <abadger1999> #chair geppetto
16:10:10 <zodbot> Current chairs: RemiFedora abadger1999 geppetto
16:10:31 <geppetto> abadger1999: https://lists.fedoraproject.org/pipermail/packaging/2014-July/010231.html … it's late, but available
16:11:47 <abadger1999> Alright there's only three of us, it looks like.
16:12:01 <abadger1999> geppetto: you're also going to be at flock, correct?
16:12:59 <abadger1999> #chair tibbs|w
16:12:59 <zodbot> Current chairs: RemiFedora abadger1999 geppetto tibbs|w
16:13:08 <tibbs|w> Sorry I'm late.
16:13:25 <abadger1999> Hey tibbs|w, there aren't many of us yet so we may go open floor and cancel if there's nothing people want to talk about.
16:13:40 <abadger1999> I'm just finding out who all is going to flock right now.
16:13:46 <tibbs|w> Works for me.  Not the best morning so far.
16:13:56 <abadger1999> RemiFedora, geppetto: Are either of you going to be in brno before flock?
16:14:05 <RemiFedora> only flock for me
16:14:14 <geppetto> no flock, no brno
16:14:26 <abadger1999> geppetto: okay.
16:14:43 <geppetto> abadger1999: were you are europython?
16:14:49 <geppetto> s/are/at/
16:14:50 <abadger1999> geppetto: Nope I wasn't.
16:14:58 <abadger1999> I think that Brno sent bkabrda there though.
16:15:12 <geppetto> yeh, valentina went too … maybe others
16:15:16 <abadger1999> k
16:15:34 <abadger1999> Hopefully some of them will get to go to pycon NA next year.
16:15:57 <abadger1999> a lot of work happens there... europython is only a shadow compared to that.
16:16:01 <geppetto> just wondered how it compared, if you spoke etc. … Valentina seems really interested in PyPy now … so there's that :-o
16:16:08 <abadger1999> hehe :-)
16:16:28 <abadger1999> That's cool.  dmalcolm was doing some work on pypy upstream and was very intrigued by the code.
16:16:46 <abadger1999> But right now we don't have much upstream presence.
16:17:02 <abadger1999> pypy's potential is pretty large.
16:17:34 <abadger1999> the big problem is that it's constantly chasing the CPython reference implementation.  So you get a slightly older version of the language.
16:18:33 <abadger1999> The other problem for us as packagers is that the language it implements isn't 100% compatible.  So we'd have to decide whether to share the same packages/code, create separate files on disk for them. or something else.
16:18:48 * geppetto nods
16:18:53 <abadger1999> #topic Open Floor
16:19:04 <abadger1999> Okay, is there anything anyone wants to talk about?
16:19:49 <abadger1999> #topic flock travels
16:20:22 <tibbs|w> BTW, that pypy/cpython problem sounds exactly like the ruby/jruby problem, and that wasn't solved particularly well.
16:20:48 <abadger1999> I'm headed to Brno for meetings this weekend and then flock.  I will be back on aug 11th
16:21:16 <geppetto> I kind of hoped PyPy would catch up when CPython had the feature moratorium for a year or so
16:21:17 <abadger1999> tibbs|w: yeah -- when the ruby/jruby proposal was made I hoped it would work well and we could reuse it for pypy cpython.
16:21:23 <abadger1999> but it didn't :-(
16:22:13 <abadger1999> so I don't know whether we should do something different or try to refine the ruby/jruby idea to be workable.
16:22:24 <abadger1999> Yeah.
16:22:43 <abadger1999> And the PSF has put money towards pypy development to catch up with CPython.
16:22:55 * geppetto nods
16:23:42 <geppetto> I mean … it'll still not have RAII, which I think will be bad, but if they can make all the stuff run it'd be a huge help.
16:27:05 <abadger1999> spot will also be at flock and RemiFedora.
16:27:14 <abadger1999> Not sure about the others that aren't here.
16:28:30 * geppetto nods
16:28:34 <geppetto> So flock is next week?
16:29:07 <abadger1999> So I figure we'll all be someplace that has good wifi or we'll all be off the net... won't know until we get there.
16:29:09 <abadger1999> Yeah.
16:29:26 <abadger1999> http://flocktofedora.org/schedule/  6th->9th
16:29:31 <geppetto> ok, well in case there is no internet … or you are doing way more fun things
16:29:37 <abadger1999> :-)
16:29:39 <geppetto> I'm not going to be around the week after
16:29:40 <abadger1999> that's true too :-)
16:29:51 <abadger1999> Okay.
16:30:28 <abadger1999> #info flock next week.  Might not have quorum if the flock attendees can't make the meeting
16:30:40 <abadger1999> #info geppetto won't be around the week after flock
16:31:00 <abadger1999> #topic Toshio Leaving
16:31:10 <geppetto> ?
16:31:19 <abadger1999> So I'm going to be leaving Red Hat.
16:31:26 <geppetto> Gah!
16:31:29 <RemiFedora> Ugh
16:31:41 <langdon> :(
16:32:09 <abadger1999> And I think I should give up my seat on FPC as well -- I've been here a long time and I'm not as enthusiastic as I was when I started... not sure how much time I'll have once I start a new job.
16:32:09 <tibbs|w> That place seems to have crazy turnover.
16:32:46 <abadger1999> So we'll have to put out a call for volunteers to fill my seat.
16:33:05 <abadger1999> Usually spot handles that and I'll see if he's willing to do so this time.
16:33:24 <RemiFedora> perhaps he already have some name...
16:33:33 <RemiFedora> from last call
16:33:39 <abadger1999> There are a few people that would be good candidates but I don't know if any of them would be willing to serve yet.
16:35:32 <abadger1999> So if you have ideas of good candidates, let them know that the call for volunteers will be going out soon.
16:35:49 <abadger1999> #topic How to handle the chair
16:36:04 <abadger1999> So last week I proposed that we do a rotating chair.
16:36:11 <abadger1999> like fesco does.
16:36:36 <abadger1999> Do people feel good about that?
16:37:06 <tibbs|w> Hmm.
16:37:19 <abadger1999> A rotating chair would be more work than fesco's rotating chair
16:37:19 <tibbs|w> I'm afraid I'd suck at it, is all.
16:37:38 <abadger1999> as we copy things into the wiki a lot more than fesco.
16:37:43 <abadger1999> so there's more post-meeting work.
16:38:14 <abadger1999> but unless some one person can do the work most weeks, a rotating chair might spread the load better.
16:38:31 <geppetto> yeh, it's rock and hard place time
16:38:44 <abadger1999> Another alternative is rotating chair for running hte meeting.  But writeups get assigned when they're voted on.
16:39:22 <abadger1999> with the caveat that everyone would have to pitch in... no fair letting a few people volunteer to writeup everything.
16:39:47 <RemiFedora> I agree this is a good solution to share the work (rotating chair), but I a bit afraid of writeup (my english is so poor), so the second way could be better
16:41:04 <abadger1999> RemiFedora: you'd probably still have as many writeups to do... just not all in the same week :-)
16:41:34 <RemiFedora> and checking my writeups should be simple ;)
16:42:56 <abadger1999> Okay -- I'll try to write down something in the wiki about what a rotating chair would be responsible for and what people who writeup a policy are responsible for.
16:43:25 <abadger1999> and we can try it out after flock.
16:44:08 <abadger1999> #topic Open Floor
16:44:20 <abadger1999> anything else people would like to talk about?
16:44:43 <abadger1999> we still don't have quorum but four is enough to discuss issues if there's something you want brought up.
16:45:23 <RemiFedora> do we have a wiki page with all bundle exception ?
16:45:29 <abadger1999> RemiFedora: most yes.
16:45:52 <abadger1999> RemiFedora: https://fedoraproject.org/wiki/Packaging:No_Bundled_Libraries#Packages_granted_exceptions
16:45:54 <RemiFedora> I was searching for fastlz... haven't found
16:46:12 <abadger1999> There are two tables -- one for permanent exceptions and one for temporary exceptions.
16:47:06 <abadger1999> RemiFedora: yeah -- I don't remember fastlz ever coming up before.
16:47:31 <RemiFedora> talking with some other packagers, I think "tracking" bundled library is probably the most important thing
16:47:49 <abadger1999> If someone wants a quick writeup -- the firefox/icecat ticket is only waiting on the list of virtual provides be entered into that table
16:48:17 <abadger1999> https://fedorahosted.org/fpc/ticket/391
16:48:48 <abadger1999> RemiFedora: yeah -- spot had brought up at one time the idea of radically changing the bundled library policy.
16:49:17 <abadger1999> RemiFedora: Going to a model where we make sure that libraries are tracked.  But pretty much always allowing them to be bundled.
16:49:32 <RemiFedora> yes, that is also my idea
16:49:49 <RemiFedora> less strict on bundling, more strict on tracking
16:50:06 <abadger1999> RemiFedora: I think if bressers, sparks or someone else from the security team were willing to sign off on that, it could be a good change.
16:50:27 <abadger1999> I think it got put to a straw poll at one point in FPC and didn't pass.
16:50:54 <abadger1999> But positive input from the security team would likely sway votes.
16:51:19 <RemiFedora> and we have a "fedora security team" :)
16:51:55 <abadger1999> <nod> and I think sparks is in charge of that.
16:52:23 <abadger1999> anyhow -- that's my suggestion on how to move that forward if you're interested.
16:52:57 <RemiFedora> So, I have to subscribe this ML and raise the discussion there
16:53:53 <abadger1999> RemiFedora: yeah ... but make sure you get input from bressers/sparks/other people we
16:53:59 <abadger1999> 've known for years.
16:54:18 <RemiFedora> yes
16:54:27 <abadger1999> the security team is newly formed so they won't have as much weight as the specific people that have been around for a long time.
16:54:36 <abadger1999> *fedora security team
16:55:16 <jsmith> I'd be happy to bring it up in the next Security Team meeting
16:55:25 <abadger1999> jsmith: aweome.
16:55:33 <RemiFedora> jsmith, yes please
16:56:10 <abadger1999> RemiFedora: Would you be able to write up the proposal/show up to that meeting/etc?
16:57:04 <RemiFedora> jsmith, when is next meeting ?
16:57:13 <abadger1999> FPC experience shows that having someone who understands the change being asked for and the pros and cons of it if the change is controversial is extremely helpful :-)
16:57:34 <jsmith> RemiFedora: Next Wednesday, assuming we don't skip for Flock
16:58:15 <RemiFedora> abadger1999, I won't be much available in august, 1 week in Flock, 2 weeks holidays....
16:58:24 <RemiFedora> so, this probably have to wait for september
16:58:29 <jsmith> RemiFedora: https://fedoraproject.org/wiki/Security_Team_meetings
16:58:33 <jsmith> RemiFedora: OK
16:58:41 <abadger1999> RemiFedora: k.  I'll be out of FPC by then.
16:58:55 * RemiFedora is now subscribed to the ML
16:59:19 <abadger1999> Alright, anything else?
17:02:08 <RemiFedora> quick feedback on 449 ?
17:02:21 * spot is around now
17:02:27 <RemiFedora> (as I will impact my work as "upstream" for the near future
17:02:30 <abadger1999> #chair spot
17:02:30 <zodbot> Current chairs: RemiFedora abadger1999 geppetto spot tibbs|w
17:03:04 <abadger1999> #topic #449 Bundle lib exception, json-c on php-pecl-jsonc
17:03:08 <abadger1999> https://fedorahosted.org/fpc/ticket/449
17:04:26 <spot> not sure i understand what is being requested there
17:04:49 <RemiFedora> use bundled libjson-c instead of system one
17:05:02 <RemiFedora> until switch to forked version
17:05:09 <abadger1999> Why are tests failing?
17:05:15 <abadger1999> Is it because the output is different?
17:05:48 <RemiFedora> more permissive, some input which was not acceptable by previous version are now parsed
17:06:07 <RemiFedora> json-c is far less strict than it should be...
17:06:51 <abadger1999> Is the difference wrong in real-world use or could the tests be relaxed?
17:06:58 <RemiFedora> and, in all case, I will have to fork version 0.11
17:07:50 <RemiFedora> I don't want to change test coming from ext/json (non-free ext.), as this will say more difference (and user are already crying about the few ones)
17:08:34 <spot> Can you just package json-c11 ?
17:09:05 <RemiFedora> only for f21 ? and only for 1 consumer ?
17:09:10 <spot> Sure.
17:09:36 * spot doesn't think it is the _best_ plan, per se, but it is worth noting
17:09:39 <RemiFedora> sorry but doesn't make sense to me
17:10:13 * abadger1999 agrees with spot.
17:10:17 <spot> fwiw, i don't have an issue with a temporary bundling exception here.
17:10:39 <RemiFedora> must remind that in all case, in a few months, it will bundle a fork
17:10:43 <abadger1999> yeah -- I also agree with that.
17:10:48 <abadger1999> err
17:10:53 <abadger1999> agree with spot's second sentiment.
17:11:20 <abadger1999> RemiFedora: why bundle a fork rather than create a forked library?
17:11:59 <RemiFedora> because it will not be a usable library. Using PHP internal API.
17:12:09 <abadger1999> k
17:12:45 <abadger1999> yeah, I think temporary exception and eventual fork would work for me with that information.
17:12:50 <RemiFedora> my goal is to drown 80% of json-c (hask table, memory management, ...) and only use the parser + php tools (hash table, memory, ...)
17:13:06 <RemiFedora> /drown/drop/
17:13:17 <geppetto> drown worked too :)
17:14:43 <geppetto> I'm somewhat confused about what you are doing … as spot said it seems slightly better to ship the c11 version than bundle it … but I think I'm +1 on a tmp. exception, no matter which way you want to do it.
17:15:44 <abadger1999> Proposal: temporary bundling exception for php-pecl-jsonc to include libjson-c  in F21.  Library code will have been forked and integrated into php-pecl-jsonc using php library routines in f22 and beyond.
17:16:42 * abadger1999 waits for RemiFedora to vote first in case he misphrased that
17:16:48 <RemiFedora> +1
17:16:51 <abadger1999> +1
17:17:41 <spot> +1
17:17:58 <abadger1999> geppetto, tibbs|w: proposal to vote on if you're still around
17:18:15 <tibbs|w> +1
17:18:24 <geppetto> +1
17:19:40 <RemiFedora> thanks
17:19:47 <abadger1999> #info php-pecl-jsonc granted a temporary exception to bundle libjson-c  for F21.  Will fork the code for f22 and beyond.
17:19:52 <abadger1999> #topic Open Floor
17:21:27 <abadger1999> Okay, if nothing else, I'll close in 1minute.
17:22:35 <abadger1999> #endmeeting