Log from yesterdays (20070726) EPEL SIG Meeting

Thorsten Leemhuis fedora at leemhuis.info
Thu Jul 26 05:29:41 UTC 2007


00:00:00 <       knurd> | Meeting ping dgilmore, Jeff_S, knurd, mmcgrath, nirik, stahnma, quaid and everyone interested in EPEL -- EPEL meeting in #fedora-meeting now!
00:00:00 <       knurd> | Hi everybody; who's around?
00:00:00            --- | knurd has changed the topic to: EPEL Sig meeting -- Meeting rules at http://www.fedoraproject.org/wiki/Extras/Schedule/MeetingGuidelines -- Init process
00:00:02              * | 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
00:00:15 <         f13> | I'm here by accident.
00:00:30              * | dgilmore is here
00:00:45 <      Jeff_S> | f13: that's what my parents said too :(
00:00:52 <    dgilmore> | f13: we will take you as one of our own :)
00:01:22 <       knurd> | mmcgrath ponged in #epel
00:01:26              * | Jeff_S here and making lame jokes
00:01:50 <       knurd> | well, let's start slowly
00:02:01            --- | knurd has changed the topic to: EPEL Meeting – broken deps – Jeff_S, All
00:02:28 <    mmcgrath> | Doah :)
00:02:31 <    mmcgrath> | pong
00:02:34 <       knurd> | the stuff mether posted to #epel really annoys me 
00:02:46 <       knurd> | why fuse one day before we want to annouce EPEL
00:02:51 <       knurd> | and then obviously broken?
00:02:56              * | marek is here too
00:03:04 <       marek> | hello guys, my first meeting
00:03:05 <       knurd> | (in case it's true what mether wrote, but I assume that for now)
00:03:08 <       knurd> | hi marek 
00:03:19 <       knurd> | anyway, how to move on
00:03:21 <         f13> | knurd: you mean epel-list?
00:03:26 <       knurd> | f13, yes
00:03:34 <    dgilmore> | knurd: thats Rhaul for you
00:03:39 <         f13> | er epel-devel-list
00:03:47 <       knurd> | move those to testing which I porposed on epel-devel-list?
00:04:03 <       knurd> | f13, there is only epel-devel-list (at least afaik)
00:04:04            --> | jbowes (James Bowes)  has joined #fedora-meeting
00:04:15 <       knurd> | dgilmore, ?
00:04:28 <    dgilmore> | knurd: it seems to be what he does best lately
00:04:32 <    dgilmore> | anyways
00:04:40 <    dgilmore> | knurd: there is only the one list
00:04:44 <       knurd> | ohh :-)
00:04:44 <      Jeff_S> | ok, well if he's correct then let's move them to testing (or remove them)
00:04:48            <-- | svahl has quit ("Ex-Chat")
00:04:51 <       knurd> | Jeff_S, +1
00:04:54 <       marek> | is there any consideration to postpone the announce of EPEL?
00:04:57 <       knurd> | what about the other broken deps?
00:05:09 <       knurd> | marek, we could do that, but it looks bad
00:05:22 <      mether> | dgilmore: i cant really guess that fuse was going to be pushed out before it was
00:05:25 <       knurd> | I#d prefer to get EPEL5 in shape now, and annouce EPEL and EPEL5
00:05:26 <    mmcgrath> | knurd: I think people have been responsive.
00:05:26 <      Jeff_S> | knurd: at first I didn't like the idea of putting them in testing/ but the idea is growing on me
00:05:31 <       marek> | better to release good product later, that bad product at time (imho)
00:05:48 <       knurd> | marek, yeah, but otherwise we might never release...
00:06:06 <      Jeff_S> | mmcgrath: yep, it seems like some people are taking the initiative to get their deps into EPEL
00:06:09 <       quaid> | release early, release often :)
00:06:11 <       knurd> | Jeff_S, yeah, we should fix everything in testing soon as well
00:06:27 <    dgilmore> | knurd: so anything broken needs to get moved into testing
00:06:35 <       knurd> | dgilmore, well, I posted to the list
00:06:37 <    dgilmore> | which is manual  but needs doing
00:06:39 <       knurd> | some stuff is in progress
00:06:43 <      Jeff_S> | knurd: yes, I don't like having things broken in /testing/ but at least people can see what's there and build on that
00:06:44 <       knurd> | dgilmore, yeah, that's manual
00:06:52 <       knurd> | dgilmore, if you give me access I can handle it
00:07:22 <       knurd> | (in case you are busy with other stuff)
00:07:23 <     stahnma> | here now
00:07:30 <    dgilmore> | knurd: i can get you setup.  or i can do it
00:07:39              * | dgilmore is traveling toomoroow
00:07:54 <       knurd> | dgilmore, well, I'd think it might be a good idea if I or someone else gets access
00:08:08 <       knurd> | as you and mmcgrath seems to be a bottleneck afaics
00:08:12 <    dgilmore> | knurd: :) we will make it happen
00:08:19 <       knurd> | dgilmore, k, thx
00:08:31 <       knurd> | what about EPEL4? shall we try to fix that up in a similar way?
00:08:49 <      Jeff_S> | knurd: yes
00:09:21 <       knurd> | I'm busy tomorrow at 18:00 UTC, but I think it could try to fix everything up to then
00:09:55 <       knurd> | quaid, could you send the annoucement afterwards?
00:10:25 <       knurd> | dgilmore, can you give me access later today or early tomorrow?
00:10:34 <       knurd> | then I'll try to fix everything
00:10:42 <       knurd> | dgilmore, but you'd need to adjust the push scripts
00:10:47 <       quaid> | knurd: sure
00:10:53 <       knurd> | to make sure that all the new stuff from now own goes to testing
00:11:04 <       knurd> | otherwise we'll have broken deps soon again
00:11:05 <    dgilmore> | knurd: yes
00:11:11 <       knurd> | dgilmore, k, thx
00:11:17 <       knurd> | anything else regarding this topic?
00:11:33 <    mmcgrath> | knurd: one quick question.  What exactly are dgilmore and I the bottlenecks of?
00:11:39 <      Jeff_S> | well, the yum-related packages I posted earlier
00:11:51 <      Jeff_S> | I'd rather have those in the main repo than in testing/
00:11:54 <       knurd> | mmcgrath, not exactly bottlenecks, but you are busy and have lots of load already
00:11:59 <      Jeff_S> | but that's just my opinion :)
00:12:04 <    mmcgrath> | knurd: what hasn't been getting done?
00:12:22 <       knurd> | mmcgrath, it wasn#t meant as offense
00:12:37 <       knurd> | but I would have fixed the deps days agao if I would have had access
00:12:42              * | mmcgrath not offended, just confused.  We can't branch and build peopels packages for them.
00:12:56 <       knurd> | mmcgrath, sure, I meant the moving
00:13:05 <       knurd> | I'd preferred to have that done last weekend or so
00:13:06 <    mmcgrath> | ahh, well have fun :)
00:13:07 <    dgilmore> | knurd: we will need to change mock configs on the builders also
00:13:16 <       knurd> | dgilmore, upphs
00:13:21 <       knurd> | dgilmore, can you handle that?
00:13:32 <    dgilmore> | we will want whats in testing in the buildroot
00:13:36 <       knurd> | dgilmore, +1
00:13:39 <    dgilmore> | yes
00:13:41 <      Jeff_S> | yes
00:13:49 <     stahnma> | yes
00:14:19 <       knurd> | k, so I'll get acess and fix the broken deps and dgilmore will update the mock configs and the push scripts
00:14:20 <       quaid> | sorry, did we pick a time (yet) for official door opening/
00:14:29 <       quaid> | or just be ready to announce when it is ready?
00:14:50 <       knurd> | quaid, I'd say we should aim 15:00 UTC
00:15:08 <       knurd> | quaid, I'll give you a heads up when I have the deps fixed
00:15:15 <       knurd> | does that sound like a plan?
00:15:59 <       quaid> | ok
00:16:38 <       knurd> | mmcgrath, are you around tomorrow in case I need help and in case dgilmore is traveling?
00:17:06 <    mmcgrath> | knurd: yep
00:17:11 <       knurd> | great
00:17:17              * | knurd moves on then
00:17:23            --- | knurd has changed the topic to: EPEL Meeting – when to run the spam o magic script – mmcgrath
00:17:33 <       knurd> | mmcgrath, could the script run when we push new pacakges?
00:17:42 <       knurd> | that's what Extras did/does
00:17:56 <       knurd> | that sounds a bit better then just once a week 
00:18:03 <       knurd> | at least if it isn#t to hard to set up
00:00:19              * | knurd moves on for now, we can get back if needed
00:00:23            --- | knurd has changed the topic to: EPEL Meeting – branch for EPEL if Fedora maintainer does not
00:00:30            --- | knurd has changed the topic to: EPEL Meeting – branch for EPEL if Fedora maintainer does not react – knurd
00:00:40 <    mmcgrath> | knurd: the spam-o-matic script?  Probably not.  It takes over an hour to run right now.
00:00:52            <-- | giallu has quit (Read error: 110 (Connection timed out))
00:20:03 <       knurd> | mmcgrath, could it just be triggered to run maybe?
00:20:07            <-- | JSchmitt has quit ("Konversation terminated!")
00:20:11 <       knurd> | then nobody would have to wait for it
00:20:14 <    mmcgrath> | http://git.fedoraproject.org/hosted/fedora-infrastructure.git/scripts/?p=hosted/fedora-infrastructure.git;a=tree;f=scripts/spam-o-epel
00:20:20 <     stahnma> | any idea why it takes over an hour?  Is that normal (I just don't know)
00:20:22 <      Jeff_S> | mmcgrath: how many repos is it looking at?  repoview takes a few minutes at the most for me
00:20:33 <    mmcgrath> | stahnma: not really, I'd have to take a look.
00:20:47 <    mmcgrath> | Perhaps we should re-think what we're doing.
00:21:14 <    mmcgrath> | we're looking to do more of a "Sorry, we couldn't push this file because of these deps:"
00:21:23 <    mmcgrath> | and so they'd stay in needsign.
00:21:27 <    mmcgrath> | dgilmore: does that make sense?
00:21:34 <    mmcgrath> | s/push this file/push this package/
00:21:43 <       knurd> | mmcgrath, won't we get that with bodhi soon?
00:21:48 <    dgilmore> | mmcgrath: yeah
00:22:08 <    mmcgrath> | knurd: 'soon' is very misleading.
00:22:10 <    dgilmore> | knurd: its not known we we will be able to move that way
00:22:18 <       knurd> | k, was just wondering
00:22:28 <    dgilmore> | knurd: it will probably be many months
00:22:33 <    mmcgrath> | knurd: I have a suspicion it won't be in there until someone steps up and submits the patches.
00:22:47 <       knurd> | k
00:23:14 <       knurd> | in that case it might be worth to enhance the current push scripts
00:23:33 <     stahnma> | brb
00:23:37 <    mmcgrath> | stahnma: well it only takes an hour for epel5, epel4 is significantly faster.
00:23:39 <       knurd> | k, so back to the current topic
00:23:45 <       knurd> | branch for EPEL if Fedora maintainer does not react
00:23:50            <-- | Kevin_Kofler  has left #fedora-meeting ( "Bye!")
00:23:54 <       knurd> | is https://www.redhat.com/archives/epel-devel-list/2007-July/msg00104.html fine for everyone?
00:23:59            <-- | LetoTo  has left #fedora-meeting ( )
00:24:16 <       knurd> | gets a +1 from me
00:24:19            --> | LetoTo (Paul Wouters)  has joined #fedora-meeting
00:24:37 <       marek> | +1 from my side
00:24:48            <-- | LetoTo has quit (Client Quit)
00:24:53 <      Jeff_S> | +1
00:24:55              * | knurd will wait until we get 4 +1 from steering committee members
00:25:05 <    dgilmore> | knurd: as long as someone  and no
00:25:05 <    dgilmore> | more than one month has passed, then the EPEL maintainer of the package
00:25:05 <    dgilmore> | must hand primary per release maintainership back to the Fedora
00:25:08 <    dgilmore> | maintainer (and become comaintainer, if interested).
00:25:11 <    dgilmore> | gahh
00:25:15            --> | LetoTo (Paul Wouters)  has joined #fedora-meeting
00:25:24            <-- | LetoTo has quit (Remote closed the connection)
00:25:24 <    dgilmore> | that doesnt rub overly well with me
00:25:33 <       knurd> | then why didn#t you post to the list
00:25:39 <       knurd> | dgilmore, could you do that please
00:25:47 <       knurd> | then we'll get to this topic in next weeks meeting
00:26:03 <    dgilmore> | i think we simplify it and say if the maintainer wants to work on EPEL then great both people become co maintainers
00:26:04 <     stahnma> | I thought that was changed on list, to something like, become comaintainers or something
00:26:18 <     stahnma> | dgilmore: +1
00:26:27              * | Jeff_S happy either way
00:26:30 <    dgilmore> | otherwise im good with it
00:26:33 <       knurd> | dgilmore, can you work out the wording?
00:26:34 <     stahnma> | me too
00:26:43 <       knurd> | dgilmore, then we'll ACk it next week
00:27:07 <    dgilmore> | knurd: sure
00:27:12 <       knurd> | dgilmore, tia
00:27:19            --- | knurd has changed the topic to: EPEL Meeting – EPEL announcement -- everything ready? – quaid, all
00:27:23 <       knurd> | we discussed this already
00:27:28 <       knurd> | did we miss anything?
00:27:48 <     stahnma> | do we want some blogs/articles on it?
00:27:55 <       knurd> | stahnma, good idea
00:28:05 <     stahnma> | I can write up something on opensource.apress.com
00:28:20 <     stahnma> | and maybe try to get something into newsforge, if I have time
00:28:31 <       knurd> | stahnma, would be great
00:28:41 <       knurd> | I'll try to blog about it as well
00:28:58              * | knurd moves on
00:29:02            --- | knurd has changed the topic to: EPEL Meeting – new meeting time? – all (see also
00:29:03 <       knurd> | http://fedoraproject.org/wiki/EPEL/NewMeetingTime )
00:29:29 <       knurd> | dgilmore, your times are still missing
00:30:11 <       knurd> | stahnma, yours as well afaics
00:30:14 <    dgilmore> | knurd: i have not had time to look at it.
00:30:21 <       quaid> | huh, weird
00:30:25 <       quaid> | how did I get up there?
00:30:30 <       quaid> | must be the old numbers :)
00:30:42 <       knurd> | quaid, check the history; no idea...
00:30:43              * | quaid thinks they are still good
00:30:44 <      Jeff_S> | quaid: I think you edited the page I created and I moved your times over
00:30:47              * | stahnma doesn't really care for any of the times....I'm at my dayjob during the entire open schedule
00:30:50 <       quaid> | ok, yeah
00:30:52 <      Jeff_S> | but I don't remember :)
00:31:27            --- | knurd has changed the topic to: EPEL Meeting -- Free discussion around EPEL
00:31:30 <       knurd> | anything else?
00:31:35              * | mmcgrath has nothing.
00:31:54 <       marek> | I want to ask, what do you think about security response team for EPEL?
00:32:13 <      Jeff_S> | any thoughts about those yum-related packages I posted?  It'd be nice to get those taken care of for EPEL-4
00:32:15 <    dgilmore> | marek: its needed you volunteering?
00:32:30 <       knurd> | marek, I'm all for having one, but such issues should be discussed on the list IMHO
00:32:36 <       quaid> | yep
00:32:37 <       marek> | dgilmore: not yet, just an idea. I want to know what do you think about it
00:32:44 <       knurd> | marek, that's for the initial discussion a lot easier than here in the meeting
00:32:48 <       quaid> | we should be coordinate with the main Fedora SRT
00:32:48            <-- | k0k has quit (No route to host)
00:32:49 <    dgilmore> | Jeff_S: we cant do yum for epel
00:32:50 <       marek> | ok, I'll bring it to the list
00:32:58 <      Jeff_S> | dgilmore: why not?
00:33:01 <       knurd> | Jeff_S, I like the plan, but didn't look at the packages itself
00:33:06 <      Jeff_S> | I thought we decided it was needed?
00:33:16 <    dgilmore> | Jeff_S: because RHEL is not available in a way yum could use
00:33:40 <     stahnma> | Jeff_S: I like what you did.  It fixes the gap for people trying to use yum-utils or createrepo on RHEL 4
00:33:41 <    dgilmore> | Jeff_S: so if you went to yum install a package from epel and needed RHEL deps your screwed
00:33:52 <     stahnma> | ah
00:33:55 <       knurd> | marek, many thx
00:34:10 <       knurd> | dgilmore, yum is mainly needed for mock afaik
00:34:10 <       quaid> | dgilmore: but ... wouldn't you just get a list of deps and could go up2date them?
00:34:21 <       knurd> | or am I missing something here?
00:34:22 <    dgilmore> | quaid: you could  but thats ugly
00:34:32 <    mmcgrath> | cobbler has yum deps (which was how we started talking about it)
00:34:32 <      Jeff_S> | dgilmore: it is not there to become the system update tool for RHEL (although people could do that if they really wanted) -- but this will let people build things which rely on yum, yum-utils, etc.
00:34:39 <    dgilmore> | up2date the package from epel direct
00:34:46 <       quaid> | k
00:34:53 <    mmcgrath> | hink in that case, yum will just have to be another tool on RHEL4, and not one they can update their system with.
00:34:57 <      Jeff_S> | it's not providing any repo files
00:34:59              * | quaid didn't realize up2date in RHEL 4 supported yum repos
00:35:10 <     stahnma> | it does....poorly
00:35:20              * | stahnma loves up2date (ducks)
00:35:28 <       marek> | yes, but it isn't very good :/
00:36:01 <    dgilmore> | Jeff_S: i feel if its there people will expect it to work as an update tool
00:36:01 <       knurd> | Jeff_S, why do we need yum? for mock? for something else? dgilmore's point is valid
00:36:29 <      Jeff_S> | currently it was needed to have cobbler (which needs yum-utils, which needs yum)
00:36:44 <      Jeff_S> | I think there were some other requests, but I don't recall off the top of my head
00:37:06 <       knurd> | Jeff_S, i#d say wwe go without yum for now and look closer at the issues at hand later
00:37:23 <       knurd> | even if that means to move mock into testing/4/ for now
00:37:29 <    dgilmore> | Jeff_S: we need to be super careful here  we need to use the a EVR lower version than CentOS
00:37:40 <     stahnma> | I think Jeff_S did that
00:37:41 <    mmcgrath> | perhaps we should document it very well in the yum config files.
00:37:51 <    mmcgrath> | which, by default, would point to nothing (not even epel)
00:37:54 <      Jeff_S> | dgilmore: yes, I did that
00:37:58 <    dgilmore> | an d we need to make sure people understand they can not use itto keep there RHEL box up2date
00:38:04 <      Jeff_S> | mmcgrath: yep, there is only yum.conf but it doesn't point to any repos
00:38:51 <    mmcgrath> | I think thats reasonable.
00:38:55 <      Jeff_S> | ok, maybe we need to discuss this more on the list?
00:38:58            --> | rdieter (Rex Dieter)  has joined #fedora-meeting
00:39:00 <       knurd> | Jeff_S, +1
00:39:16 <     stahnma> | +1
00:39:23 <      Jeff_S> | the thread was 'packages that conflict with centos?' -- and everyone that responded was positive about including them...
00:40:19 <       knurd> | Jeff_S, that afaics seems to become more and more a problem
00:40:24 <       knurd> | we discuss stuff on the list
00:40:37 <       knurd> | and then suddenly someone raises problems in hte meetings
00:40:45 <       knurd> | and the whole discussion restarts
00:40:52 <    dgilmore> | Jeff_S: we need tomake sure it does not touch CentOS setups in bad ways
00:40:52 <       knurd> | that#s a bit annoying
00:41:08 <    dgilmore> | Jeff_S: epel-release has an epel repo file
00:41:19 <    mmcgrath> | knurd: I much prefer that to people not showing up to the meeting and commenting on the notes after the fact.
00:41:44 <       quaid> | we have to recognize that sometimes people only pay attention when a meeting is happening
00:41:49 <     stahnma> | specifically for Yum-based users in epel-release, this could be difficult
00:41:52 <       quaid> | because of busyness .. "Wow, it's been a week?!?"
00:41:54 <      Jeff_S> | dgilmore: agreed...  They all have a lower release number than the centos versions, so a centos-4 machine should never install them in the first place
00:41:56 <       knurd> | mmcgrath, maybe, but raising stuff directly when it gets discussed it teven better
00:42:16 <       quaid> | knurd: so maybe if someone comes in late, we can quickly decide if the conversation needs to continue ...
00:42:40 <       quaid> | knurd: OT -- what do you think about the idea that nothing can leave a list for a vote in a meeting until there is a consensus to vote?
00:42:45 <     stahnma> | if they have epel-release isntalled and then install yum, will yum on RHEL try to be an update tool?
00:42:46 <       quaid> | is that too much bureaucracy? :)
00:43:21 <       knurd> | quaid, liekly :-/
00:43:24 <      Jeff_S> | stahnma: if someone tries to use it as such, yes
00:43:41 <       knurd> | quaid, but the idea has some appeal nevertheless
00:43:49 <       quaid> | knurd: then we are always going to have that situation ...
00:43:55 <       quaid> | maybe a guideline rather than a rule
00:44:04 <       knurd> | quaid, the problem at hand would not be solved
00:44:07 <       quaid> | but it could be good across _all_ projects/sigs
00:44:08 <    dgilmore> | Jeff_S: my main concern is that people dont try use it on RHEL  it just wont work there
00:44:09 <       quaid> | true
00:44:13              * | quaid did note it was OT:)
00:44:24 <    dgilmore> | Jeff_S: we could probably not put a yum.conf in at all
00:44:24 <       knurd> | someone that did not take part in the discussion could step up in the meeting and say "hell, no"
00:44:39 <     stahnma> | dgilmore: perhaps we should test that
00:44:43 <     stahnma> | that might work
00:45:37 <       knurd> | anyway, any more stuff to discuss?
00:45:42 <       knurd> | or shall I close the meeting?
00:45:47              * | dgilmore needs to go to the colo 
00:45:52 <    dgilmore> | please clos e
00:46:00 <     stahnma> | thanks all
00:46:03              * | knurd will close the meeting in 30
00:46:08            <-- | stahnma  has left #fedora-meeting ( "Time for something else....")
00:46:26              * | knurd will close the meeting in 15
00:46:41 <       knurd> | -- MARK -- Meeting end




More information about the epel-devel-list mailing list