00:00:12 < knurd> | Meeting ping dgilmore, Jeff_S, knurd, mmcgrath, nirik,
stahnma, quaid and everyone interested in EPEL -- EPEL meeting in #fedora-meeting now!
00:00:12 < knurd> | Hi everybody; who's around?
00:00:12 --- | knurd has changed the topic to: EPEL Sig meeting -- Meeting
rules at http://www.fedoraproject.org/wiki/Extras/Schedule/MeetingGuidelines
00:00:13 * | knurd likes to remind people that the schedule and the topic
list for todays meeting can be found on http://fedoraproject.org/wiki/EPEL/Schedule
00:00:28 * | Jeff_S here
00:00:30 * | mmcgrath here
00:00:43 * | nirik is here.
00:01:14 --> | daMaestro (Jon) has joined #fedora-meeting
00:01:31 < knurd> | lala
00:01:42 * | daMaestro sings along
00:01:48 < daMaestro> | epel++
00:01:50 < daMaestro> | ;-)
00:02:01 < knurd> | well, let's start slowly
00:02:04 < knurd> | daMaestro, :-)
00:02:06 --- | knurd has changed the topic to: EPEL Meeting – broken deps –
00:02:12 --> | mra (matt anderson) has joined #fedora-meeting
00:02:19 < knurd> | Jeff_S, any progress?
00:02:25 < knurd> | did you poke people?
00:02:43 < Jeff_S> | no, I thought mmcgrath did enough of that for me :)
00:02:53 < mmcgrath> | :: sigh ::
00:02:55 < Jeff_S> | actually, I know a couple packages got fixed
00:02:56 < knurd> | hehe :)
00:02:57 < daMaestro> | i got poked
00:03:05 < daMaestro> | but i have no idea how to fix it..
00:03:09 < knurd> | daMaestro, eeryone got accidentally
00:03:12 < Jeff_S> | but I ran it again today and there is still a big list
00:03:21 < daMaestro> | ahh, ok; good
00:03:37 < knurd> | mmcgrath, is the script working now?
00:03:41 < Jeff_S> | spot says he is planning to import a bunch of his packages
00:03:50 < Jeff_S> | so that should help
00:03:54 < mmcgrath> | knurd: Here's the problem with the script... it
00:04:02 --- | knurd has changed the topic to: EPEL Meeting – broken
deps/spam-o-magic – Jeff_S/mmcgrath
00:04:05 * | spot is working on it
00:04:07 < knurd> | mmcgrath, what?
00:04:13 < mmcgrath> | So I'm working to rewrite that part of it to read from
00:04:19 < knurd> | how does that script run for FE-6?
00:04:19 < Jeff_S> | but there are some packages which seem to be looking at
stuff that is not available in EL4
00:04:19 < spot> | i've just been sidetracked by the need to get
00:04:35 * | mmcgrath has no idea.
00:04:42 < knurd> | mmcgrath, ask mschwendt?
00:04:44 < mmcgrath> | spot: Are you working on the script?
00:04:46 < knurd> | mmcgrath, he might know
00:04:56 < knurd> | mmcgrath, and remmeber, owners.epel.list ;-)
00:04:56 < nirik> | I think it has a local copy of owners.list it uses?
00:05:02 < Jeff_S> |
see for example postgresql-pgpoolAdmin which requires a higher version of PHP than is in
00:05:02 < mmcgrath> | :)
00:05:41 < quaid> | oi!
00:05:42 < knurd> | well, the plan is to annouce in 8 days iirc
00:05:45 < quaid> | sorry I'm late
00:05:47 < mmcgrath> | spot: ping?
00:05:55 < knurd> | do we get everything fixed until then?
00:05:59 < knurd> | I doubt that a bit
00:06:15 < nirik> | I think we should try and if we need to slip the
announcement we can do that...
00:06:23 < Jeff_S> | I doubt it too. The spam script would help, IMO.
00:06:25 < knurd> | mmcgrath, spot afaik only wanted to build his packages,
which should solve some of the deps
00:06:35 < mmcgrath> | knurd: ahh. thanks.
00:06:48 < nirik> | mmcgrath: did the spam-o-matic make a valid run after the
invalid one? Or it hasn't made a valid run yet?
00:06:52 < knurd> | mmcgrath, send mschwendt a mail about the script
00:06:58 < knurd> | he really should know the answer
00:07:05 < knurd> | mmcgrath, or shall I do?
00:07:13 < mmcgrath> | I'll ping him
00:07:18 < knurd> | mmcgrath, thx
00:07:34 < knurd> | while we were on the shedule already
00:07:36 --- | knurd has changed the topic to: EPEL Meeting – EPEL annoucement
00:07:45 < knurd> | so, broken deps not resolved yet
00:07:50 < knurd> | shall we move it one week?
00:07:58 < knurd> | annoucements isn't prepares either afaik
00:08:00 < knurd> | quaid, ?
00:08:21 < quaid> | eh, I put it off because it's not hard to write
00:08:36 < knurd> | hehe
00:08:36 < quaid> | it's not a blocking factor, fwiw
00:08:40 < Jeff_S> | mmcgrath: also if you need help w/ changing the script for
epel, I can help
00:09:05 < mmcgrath> | Jeff_S: thanks.
00:09:23 < knurd> | so, move annoucement back by one week as broken deps
likely won't get fixed in the next 8 days?
00:09:35 < quaid> | sounds wise
00:09:48 < quaid> | this is one of those "release early, but only when
it's not broken" situations
00:09:54 < Jeff_S> | knurd: +1 -- we need to get this sorted out first
00:09:55 < knurd> | agreed
00:09:57 < nirik> | sure.
00:10:11 < knurd> | 20070726 would be the new target then
00:10:39 < Jeff_S> | and maybe we need a big warning on the wiki about pushing
stuff to EPEL before deps are there?
00:10:52 * | dgilmore is here
00:10:55 < Jeff_S> | people seem to be assuming that the deps are there just
because they are in fedora (extras)
00:10:55 < knurd> | Jeff_S, maybe add something to the faq
00:11:23 < knurd> | that might be the best afaics
00:11:38 < knurd> | hi dgilmore
00:11:46 < nirik> | wish we could block anything with broken deps from
00:11:58 < knurd> | so, nobody yelled, so 20070726 is the new target then
00:12:09 < knurd> | nirik, bodhi in hte long term :-)
00:12:12 * | knurd moves on
00:12:19 --- | knurd has changed the topic to: EPEL Meeting – push new packages
to testing after EPEL annoucement – dgilmore
00:12:24 < knurd> | dgilmore, I added that
00:12:26 < nirik> | true, although bodhi doesn't currently do repoclosure
00:12:32 < knurd> | just wanted to make sure we still are in agreement
00:12:43 < knurd> | all new packages after annoucement go to the testing/
00:12:59 < knurd> | and we hand move stuff over if peole ask for it?
00:13:26 < knurd> | any maybe do a full testing/X > X move some months
later (with RHEL 5.1 maybe)
00:13:36 < knurd> | do we agree on that?
00:13:46 * | nirik thinks that sounds good.
00:13:52 < Jeff_S> | +1
00:14:16 < knurd> | dgilmore, is that possible from the infrastrucutre/pushers
00:14:23 < dgilmore> | knurd: i said last week I wont spend the time to manually
move bits from testing to stable
00:14:33 < dgilmore> | its manual and labour intensive
00:14:44 < knurd> | dgilmore, so what do you suggest as solution
00:14:46 < dgilmore> | i can push everything to testing pretty easily
00:14:55 <-- | XulChris has quit (Read error: 113 (No route to host))
00:14:58 < knurd> | sure, that's the plan
00:15:01 < nirik> | the only reason we would need anything manual is for
00:15:06 < knurd> | but what about security fixes?
00:15:06 < dgilmore> | we have no easy way to move things to stable
00:15:24 < knurd> | well, we need a solution
00:15:27 < dgilmore> | how do we know whats security
00:15:39 < knurd> | someone tells us (via the wiki or a e-mail alias)
00:15:42 < dgilmore> | we can easily put things in one location
00:15:56 < dgilmore> | but to move things we have no way to do that
00:16:05 < nirik> | yeah, without a bodhi we need a manual process. ;(
00:16:19 < knurd> | dgilmore, we use the push scripts for another repo
00:16:30 < knurd> | there a simple move and createrepo works
00:16:39 < knurd> | sure, it's a ugly as hell
00:16:42 < nirik> | dgilmore: wouldn't it just be a matter of 'mv
testing/foobar.rpm ../5/; createrepo '?
00:16:42 < knurd> | but possible
00:16:51 < nirik> | yeah... ;(
00:16:52 < knurd> | nirik, exactly that is what I mean
00:17:16 < dgilmore> | nirik: sure but you need to do that mv for each package
for each arch and SRPMS
00:17:39 < knurd> | dgilmore, sure, but a simple script can do that
00:17:40 < nirik> | yeah. Not fun at all... but beats having no security
00:17:48 < knurd> | and it should not happen that often
00:17:52 < knurd> | nirik, +1
00:18:09 < dgilmore> | what about new packages?
00:18:18 < nirik> | we can ponder on it and see if we can think of a better
way to do it? I can't off hand tho.
00:18:20 < dgilmore> | they go to testing?
00:18:21 < knurd> | we could do monthly "make testing become official
repo" until EPEL is bigger
00:18:24 < knurd> | dgilmore, yes
00:18:27 < nirik> | new packages and any other udpates go to testing.
00:18:28 < dgilmore> | when do they go stable?
00:18:43 < nirik> | when rhel does a minor bump.
00:18:46 < knurd> | dgilmore, the old plan was to do that with each RHEL minor
00:18:54 < nirik> | 5.0 -> 5.1... then we move testing over
00:18:58 < knurd> | but for the start phase it might be better to do it more
00:00:07 < dgilmore> | people will want new packages immediatly
00:00:21 < knurd> | every month or ever two months maybe?
00:00:43 < knurd> | dgilmore, and when we do what people want we have lots of
00:00:46 < nirik> | couldn't people who want a new package enable
00:00:47 < knurd> | just look at EPEL now
00:00:52 < knurd> | nirik, +1
00:20:52 < knurd> | EPEL is for EL, not for Fedora -- thus we need to be a bit
00:20:56 < knurd> | espeially with broken deps
00:21:05 < knurd> | we don't want to scare people away
00:21:13 < knurd> | other opinions?
00:21:21 < dgilmore> | we need new push scripts then
00:21:36 < dgilmore> | we cant easily do what you want with what we have
00:21:56 < knurd> | dgilmore, I'm willing to do that
00:22:08 < knurd> | e.g. move the securty updates from testing to stable
00:22:18 < knurd> | just need access to the repo then
00:22:29 < dgilmore> | knurd: that is nota huge problem
00:22:30 < knurd> | I suppose nirik would be willing to help as well with it
00:22:39 < nirik> | sure.
00:22:42 < dgilmore> | what is is checking for broken deps before doing pushes
00:22:46 < knurd> | if we share the burden then it should work until we get
bodhi for epel
00:23:10 < nirik> | dgilmore: now? nothing. :( repoclosure is really slow too
00:23:17 < knurd> | security updates should normally not introduce broken
00:23:42 < dgilmore> | nirik: exactly and each time you find something broken
you have to start over
00:23:42 < nirik> | for security I would be willing to manually run
repoclosure before any push...
00:23:47 < knurd> | that should be a rare prolem
00:23:57 < knurd> | nirik, +1
00:24:06 < knurd> | nirik, or just look at old package <-> new package
00:24:13 < knurd> | that's easier
00:24:14 < nirik> | I think wwoods was working on a way to get repoclosure or
something like it going faster. Not sure where that is tho
00:24:20 <-- | lancelan has quit (Read error: 104 (Connection reset by
00:24:49 --> | lancelan (Lance Davis) has joined #fedora-meeting
00:25:00 < knurd> | so, deal?
00:25:06 < knurd> | all new packages go to testing
00:25:13 < knurd> | security updates hand-moved over
00:25:23 < knurd> | nirik and I will help with that
00:25:32 < knurd> | and we set up a wiki page or a alias up for that?
00:25:34 < nirik> | so, we need script for moving security updates and mass
moving testing on minor release? The other current scripts would work ok for testing
00:25:52 < knurd> | script for moving security updates > +1
00:26:06 < knurd> | mass moving testing > + 0.5
00:26:16 < Jeff_S> | well, the mass move is quite simple
00:26:19 < knurd> | that should not happen that often and likely can be done
00:26:24 < knurd> | Jeff_S, agreed
00:26:36 < knurd> | check that all deps are sane in stable + testing
00:26:40 < nirik> | well, one is a more general case of the other. ;) But
yeah, not a big deal now.
00:26:44 < knurd> | move everything over from testing, done
00:26:47 < Jeff_S> | but yes, a script for moving security updates would be
quite helpful I think
00:27:25 < knurd> | settled then?
00:27:55 * | knurd will take silence as agreement soon
00:27:55 < Jeff_S> | sure :)
00:28:03 < knurd> | so let's move on
00:28:12 --- | knurd has changed the topic to: EPEL Meeting – new meeting time?
– all (see also http://fedoraproject.org/wiki/EPEL/NewMeetingTime
00:28:18 * | knurd checks that page
00:28:25 < wwoods> | notting and skvidal and lmacken and I discussed a tool for
doing quick single-package depchecking but there's no implementation yet
00:28:37 < wwoods> | no one has committed to implementing it
00:28:47 < knurd> | dgilmore, what would be your preferred time?
00:28:57 < Jeff_S> | wwoods: sounds interesting
00:29:22 < Jeff_S> | looks like everyone who edited the page likes the current
00:29:33 < dgilmore> | knurd: i havent had time to even think about it
00:30:03 < knurd> | well, let's try to get this settled until next week
00:30:16 < dgilmore> | the times that work best for me are probably no good for
00:30:40 < knurd> | dgilmore, sorry, timezone problem...
00:31:02 < knurd> | maybe I should just leave, that would make things easier
00:31:17 < knurd> | dgilmore, ping me when you got some minutes, maybe we can
work something out
00:31:23 * | knurd moves on for now
00:31:37 --- | knurd has changed the topic to: EPEL Meeting – Communication
plan for enterprise customers/ISVs/IHVs – stahnma, quaid
00:31:39 < knurd> | any news?
00:31:55 < knurd> | quaid, or just revisit after annoucement?
00:32:35 < knurd> | lala
00:32:48 * | knurd takes that as revisit next week
00:32:52 --- | knurd has changed the topic to: EPEL Meeting – ExcludeArch
TrackerBugs for EPEL – notting/knurd
00:32:55 < knurd> | opinions?
00:33:05 < knurd> | + 0.75 for EPEL-Excludearch trackers from me
00:33:21 < knurd> | can't hurt and is not htat much work
00:33:35 < nirik> | yeah, I think we can have epel versions... as long as the
arch folks know to look at them.
00:33:44 < dgilmore> | -1 i think we should reuse what exists
00:33:56 < knurd> | dgilmore, that's to confusing
00:33:57 < quaid> | sorry, was writing :)
00:34:08 < Jeff_S> | dgilmore: but won't we have differences between
00:34:12 < nirik> | I don't much care. I think we should track them tho.
00:34:12 < knurd> | dgilmore, as people that might want to fix Fedora stuff
would run into EPEl stuf
00:34:31 <-- | tibbs has quit (Remote closed the connection)
00:34:33 --> | tibbs_ (Jason L Tibbitts III) has joined #fedora-meeting
00:34:56 < dgilmore> | having to keep track of things in two places will mean
sometimes things get missed
00:35:01 --> | frozty_sa (frozt01100101) has joined #fedora-meeting
00:35:05 < knurd> | dgilmore, we can block the old Fedora bugs with the EPEL
00:35:09 < dgilmore> | we dont have different trackers for FC-6 and F-7
00:35:32 < dgilmore> | EL-4 and EL-5 are just different versions
00:35:51 < nirik> | what are the current blockers called?
00:36:12 < knurd> | FE-ExcludeArch-x86_64 iirc
00:36:37 < nirik> | were their different ones for fe and fc?
00:36:44 < knurd> | nirik, nope
00:36:53 < knurd> | not that I'm aware off
00:37:20 <-- | k0k has quit (Read error: 104 (Connection reset by peer))
00:37:33 < knurd> | I'm fine going without them
00:37:39 < knurd> | as long as it's no general problem
00:38:00 < nirik> | we can always revisit... so for now we reuse the existing
00:38:18 < knurd> | nirik, I'd say just ignore the issue for EPEL
00:38:34 < knurd> | devel counts, and that's fedora
00:38:42 --> | _blah_ (rob) has joined #fedora-meeting
00:38:57 < knurd> | I'll send something to the list
00:39:03 < knurd> | then we can decide next week
00:39:12 --- | knurd has changed the topic to: EPEL Meeting – RHX and EPEL –
00:39:26 < knurd> | quaid, can wait for some more weeks iirc?
00:39:29 < quaid> | nothing new to report yet
00:39:34 < quaid> | yep
00:39:41 < quaid> | it's going to take the time it takes
00:39:47 --- | knurd has changed the topic to: EPEL Meeting -- Free discussion
00:39:49 < knurd> | quaid, thx
00:39:53 < knurd> | anything else?
00:40:15 * | knurd will close the meeting in 60
00:40:17 < knurd> | seems not
00:40:29 * | quaid no tengo mas
00:40:41 < dgilmore> | i noticed that dag made some derogatory comments about
epel on the centos mailing list
00:40:51 < quaid> | oh, good
00:40:58 < quaid> | nothing quite like keeping the flames warm
00:41:01 < dgilmore> | saying that we dont want to work with him
00:41:12 < quaid> | in fact
00:42:01 --- | tibbs_ is now known as tibbs
00:42:01 < quaid> | I thin that's in our guidelines
00:42:02 < quaid> | it says somewheren "EPEL refuses to work with the
00:42:02 < quaid> | and then a list
00:42:02 < dgilmore> | quaid: :)
00:42:02 * | quaid hopes his sarcasm is evident in the irclog
00:42:03 < nirik> | ok.
00:42:03 <-- | Jeff_S has quit (Read error: 104 (Connection reset by peer))
00:42:03 * | mmcgrath twiddles thumbs :)
00:42:05 < quaid> | it's hard because these guys all had good points
00:42:15 < quaid> | but they put way too much heat into it
00:42:29 < quaid> | and burned the bridge before it could be built
00:42:45 < quaid> | anyway, water under the ... oh, wait, nevermind
00:42:49 < knurd> | we maybe should some carefully choose words into the wiki
00:42:55 < knurd> | why we didn#t go for repotags
00:43:03 < dgilmore> |
00:43:04 < quaid> | a good FAQ entry
00:43:08 < knurd> | and that we suggest that our contributors cooperate
00:43:26 --> | Jeff_S (Jeff Sheltren) has joined #fedora-meeting
00:43:27 < quaid> | we may be answering this for RHEL's support people
when they have customers confused in the future :)
00:43:32 < dgilmore> | i didint know if we should have a statement on the wiki
somewhere saying we are willing to accept anybody
00:44:15 < dgilmore> | knurd: repotags were not part of it
00:44:35 < knurd> | well, can't hurt to put some carefully choose words
00:45:00 < knurd> | any yes, saying somewhere "we are open to annybody to
join us" can't hurt either
00:45:24 < knurd> | I can work some sentences out and will send them out
00:45:24 < dgilmore> | i think skvidal would need to do some serious voodoo magic
in yum to try and deal with multiple repos providing the same packages in a sane and
00:45:29 < dgilmore> | not sure what his voodoo skills are like
00:45:49 < dgilmore> | that is all :)
00:45:53 --> | orc_emac (orc_emac) has joined #fedora-meeting
00:45:59 < knurd> | k, thx for you hints dgilmore
00:46:02 < knurd> | anything else?
00:46:02 < nirik> | priorities or something taging gpg keys to repo would be
the only options I can think of.
00:46:39 < daMaestro> | !
00:46:49 < dgilmore> | knurd: feel free to close up shop
00:47:00 * | knurd will close the meeting in 30
00:47:02 < daMaestro> | I have a question
00:47:11 < knurd> | nirik, a solution might arise over time (or not)
00:47:13 < orc_emac> | knurd: at the end of the day, epel is NOT open to anyone
who wants to add patent restricted material; will not sign onto the CLA; or is unwilling
to build with non-free binaries
00:47:25 < orc_emac> | that is all OK
00:47:26 < knurd> | orc_emac, okay, you have a point
00:47:35 * | knurd will close the meeting in 20
00:47:39 < orc_emac> | there cannot be total union of all independent packagerts
00:47:39 < mmcgrath> | daMaestro: yes?
00:47:46 < knurd> | knurd, I can add that to the wiki, can't hurt
00:47:50 < daMaestro> | revisor needs the pykickstart from f7; are we
"allowed" to build revisor-pykickstart from the bits from f7?
00:47:57 * | knurd will close the meeting in 10
00:47:57 < daMaestro> | or do we have to branch?
00:48:06 < daMaestro> | basically, we'd like to see revisor in epel
00:48:16 < quaid> | fwiw, I'm going to work with the RH GSS (support team)
to make sure their standard system checking script knows hunt for non-RHN-provided
packages (such as EPEL); I'm sure they have a method, but that's one that would
help the support team
00:48:21 < daMaestro> | but we have a blocker requirement of the pykickstart from
00:48:21 < nirik> | orc_emac: very true. Can't be everything for everyone.
00:48:26 < mmcgrath> | daMaestro: hit us up in #epel after the meeting :)
00:48:30 < daMaestro> | ok
00:48:30 < quaid> | there ya go :)
00:48:34 < knurd> | quaid, good to know :)
00:48:37 * | knurd will close the meeting in 10
00:48:47 < knurd> | -- MARK -- Meeting end
00:48:48 --- | knurd has changed the topic to:
-- Meetings often get
logged -- see schedule in the wiki for next meeting
00:48:51 <-- | _blah_ has left #fedora-meeting ( )
00:48:51 < knurd> | thx everyone
00:48:54 < mmcgrath> | knurd: thanks
00:49:00 <-- | daMaestro has left #fedora-meeting ( "Leaving")
00:49:00 --- | You're now known as knurd_afk
00:49:03 < knurd_afk> | have fun!
00:49:08 < knurd_afk> | bye
00:49:15 <-- | orc_emac has left #fedora-meeting ( )
00:49:28 <-- | Jeff_S has left #fedora-meeting ( )