16:01:49 #startmeeting EPEL 16:01:49 Meeting started Mon Aug 30 16:01:49 2010 UTC. The chair is smooge. Information about MeetBot at http://wiki.debian.org/MeetBot. 16:01:49 Useful Commands: #action #agreed #halp #info #idea #link #topic. 16:01:56 #meetingname epel 16:01:56 The meeting name has been set to 'epel' 16:02:23 #ping stahnma tremble nirik smooge Jeff_S 16:02:28 * nirik is somewhat around, but busy catching up on monday morning stuff... 16:02:33 same here 16:02:38 * tremble is actually about. 16:02:42 #topic who is here? 16:02:47 I'm here for now 16:02:48 :) 16:02:53 * sgallagh loiters 16:02:57 See above 16:03:06 #topic agenda https://www.redhat.com/archives/epel-devel-list/2010-August/msg00162.html 16:03:23 #topic RHEL Beta 6 Server/Workstation 16:03:38 this problem is nearly identical to what we had with rhel5 16:03:51 you type gaster than me 16:03:59 new kehboard 16:04:02 * Jeff_S here but trying to wrap some things up before I can join the meeting... 16:04:37 I don't think we ever really solved it for el5 either.... 16:04:37 * smooge watches Jeff_S wrap up the body and orders new carpet rolls 16:04:44 ;) 16:04:56 stahnma, actually they created a productivity channel for server 16:04:59 in 5 we are using the productivity channel. 16:05:09 dgilmore: any news on this in 6 ? 16:05:14 but they didn't have it until I want to say 5.0 or 5.1 16:05:36 nirik: i dont expect news until RHEL 6 is GA 16:05:53 so what do we do before then? ;( 16:06:00 we do have some broken deps due to this... 16:06:00 cry 16:06:15 Just a suggestion, but since we're doing it for a shed load of perl packages what's wrong with just rebuilding the RHEL packages ? 16:06:55 Although we would want a very defined process for doing so... 16:07:00 tremble: its really a bad idea. it takes alot of effort to make sure they stay in sync 16:07:05 tremble: At that point, why wouldn't we be working with CentOS? 16:07:11 and we have no way to truely enforce it 16:07:39 sadly, I think I have to go AFK. I'll leave this up to read scrollback. 16:07:40 tremble: things could be just fine when RHEL6 is ga 16:07:41 Fair, although if we can catch it it would be within the remit of a pp 16:07:41 * stahnma & 16:07:53 but we have no way of knowing what the final product will be 16:07:59 * tremble nods 16:08:14 sadly as this point we have to play a waiting game 16:08:31 trying to guess will only lead us to having to clean up messes 16:08:37 * tremble nods 16:08:40 yeah, there's a trac bug and a bugzilla bug where maintainers are getting upset. ;( 16:08:40 Fair 16:09:12 nirik: they have to have paitence 16:09:20 yeah. 16:09:21 ;( 16:09:52 ok I would say that we would need to pull the broken deps and then work on a plan B 16:10:45 smooge: It sounds to me like "Plan B" is close up shop until RHEL 6 is released 16:11:20 sgallagh : at least on the packages that have problems. 16:11:33 yes 16:11:38 no that is plan A on packages that have problems 16:11:55 plan C is CentOS. 16:12:22 https://bugzilla.redhat.com/show_bug.cgi?id=553142 and https://fedorahosted.org/rel-eng/ticket/3916 16:12:46 I think z00dax was intending on matching RHEL as closely as possible which would probably include the server/desktop split. 16:12:49 I don't think there is much else we can do aside from waiting at this point. 16:13:37 tremble: You're telling me that CentOS 6 won't have a single combined repo? 16:13:47 we don't know 16:14:16 z00dax, has said he wanted to be more upstream but it might be that its too ugly for the CentOS crowd 16:14:31 so its another waiting game 16:14:54 I mean they are currently on 2 DVD's by combining things together in EL5.5 16:16:23 Proposal: We have to wait for EL-6 final before we can fix these issues because things are still not stable upstream 16:16:38 +1 16:16:53 +1 16:16:58 dgilmore, nirik ? 16:17:23 yes, +1 to wait until we get more info... 16:17:29 +1 16:18:08 Approved: : We have to wait for EL-6 final before we can fix these issues because things are still not stable upstream 16:18:12 next topic? 16:19:03 Ok I sent an email to various people with Scientific Linux and CentOS about what issues adding a cost will do. I forgot to cc dag on it.. so will do so and find out 16:19:34 I know what one answer is: "use a reptag :)" 16:20:22 I don't have any objections to using a cost. I think it would be easier on those perl packages that we have allowed in that are only in some branches. 16:21:02 do we have any kind of list of those? 16:21:24 Err I can probably generate one and send it to the list. 16:21:29 it might be nice to spam maintainers and tell them about this... so the process is to just keep exactly the same version/spec as rhel has. 16:22:07 Which would probably be something we specifically state that PPs can do after X days? 16:23:39 PPs? 16:23:48 ProvenPackagers 16:23:53 * smooge votes for repotags... oh never mind 16:24:13 tremble: sure... if we have a list I could work on it, or other folks... 16:24:20 but hopefully maintainers would be responsive. 16:24:47 * dgilmore gets ready to visit smooge 16:25:11 nirik: indeed but I know I occasionally drop off the grid and I'm one of the people who'd need to stay on top of it... 16:25:46 yeah, I wish we had some proactive way to note when updates come along... 16:26:27 There's a repo in 5 that updates a couple of days early... 16:26:44 (Through RHN) 16:27:19 might be cool (although I have no idea how much work) to run repodiff and mail the list anytime the builders update. 16:28:09 nirik I was thinking of just running the script I was using to generate the comparisions. 16:28:30 and see if any are out of sync from our repo? 16:28:37 * tremble nods 16:29:24 that would be cool. ;) 16:30:45 Needs a fair bit of polishing, but shouldn't be impossible... Could *possibly* get it to fire off BZ bugs but I'll make no guarantees on that on. 16:30:58 so, everyone is ok with cost adding? what value? 16:31:11 it has to be over 1000 16:31:18 What's the default cost? 1000? 16:32:38 1000 is default I think. 16:32:55 I guess we should wait to hear back from all the other repos... really all we care is that we are higher than rhel... 16:33:03 but it would be good to play nice with others. 16:33:08 * tremble nods 16:33:50 Would say wants to be at least 1100 to allow people to put their own scores above and below easily 16:33:54 well my main worry would be if CentOS had a default of 1100 or something and we set to 1005 16:34:13 I think 1100 would be good 16:35:14 Proposal: EPEL is looking to set a yum.conf cost to 1100 for repositories that support it. [Tabled til next meeting for time for feedback] 16:35:30 +1 16:35:41 sounds reasonable. 16:36:33 ok I think its time for the next topic? 16:36:47 #topic Fedora Electronics Lab 16:37:04 I know there is some email on this.. not sure what was the resolution about it 16:37:50 anyone? 16:38:02 I have no idea what this topic is. ;) What was the email about? 16:38:05 * smooge has a horrible sneez fit going on 16:38:45 Oh, right, someone wanting to install electronics lab software and having some issues... 16:38:59 not sure what all we can do here. I think they did get help on the list. 16:39:02 https://www.redhat.com/archives/epel-devel-list/2010-August/msg00121.html 16:39:24 it looks like either non-closed deps or missing groups in comps 16:40:03 there was something last month also 16:40:21 since it had come up twice I wanted to make sure it was covered. 16:40:29 I don't think everything is available in EPEL that is in fedora... 16:40:39 some things just can't work with el5... 16:40:41 ugh I think I just sneezed out my vrains 16:40:55 #chair nirik 16:40:55 Current chairs: nirik smooge 16:41:06 in case those were nmy brrains 16:41:07 nasty. 16:42:26 ok I see Chitlesh had some questions about EL-6 things.. I probably answered too nruskly 16:42:34 bruskly that is 16:43:05 ok since none of us know about it.. I will ask if he needs this on next weeks meeting 16:43:12 sounds good. 16:43:25 * nirik goes to get another cup of coffee... back in a few. 16:44:56 #topic How do we deal with Emerging Technologies versus Stable tools? 16:45:16 Ok this is meant again as a stimulation of debate which I will be emailing the list in a short while. 16:46:46 I think the basis of EPEL has changed somewhat from its early days. Early on we were aiming for a stable set of technologies that EL didn't ship that people wanted. 16:47:36 However it seems to me that our most passionate customers and packagers are more about pushing emerging technologies into EL stream 16:47:48 things that may not have clean upgrade cycles etc 16:48:23 examples being? 16:48:32 mediawiki is near and dear to my heart. 16:48:42 most of the web technologies fall into this 16:49:13 couchdb.. git (less so). cobbler all have had incompatible upgrades at times 16:49:14 right, so we have had this discussion before. ;) 16:49:36 I suspect a lot of the pressure will ease with rhel6. 16:49:44 since it will have new versions again. 16:49:50 well until 8 months after EL6 16:49:59 yeah, of course. 16:50:46 under our currently setup, the solution would be newer packages for those new versions... mediawiki99 or whatever. 16:50:57 I guess I am bringing up old stuff again.. sorry it seemed so new when I first thought of it as "Emerging Technologies for Enterprise Linux" versus "EPEL" 16:51:44 well, we did talk about another, more volitile repo... 16:51:49 but never really went anywhere. 16:52:35 one sec I have a cat trying to get into the house.. and I don't have a cat 16:53:09 ok I guess it wanted a petting or something 16:53:58 ok since I am rehashing old stuff.. next topic 16:53:59 ...now you have a cat. ;) 16:54:16 and call in cleaning gurus to fix our wiki. 16:54:33 oh that didnt go well 16:54:42 sure, I just don't know who would be such a cleaning crew. ;) 16:54:53 I think lots of our wiki could be revamped. ;( 16:54:54 I summon rbergerso 16:55:27 actually I think the Fedora wiki should be canned but I say that out of a deep love and respect for mediawiki 16:55:57 * smooge thinks he has the package to a point where it passes rpmlint 16:57:04 anyway.. I figure we should call in a guru who can help us clean out the ruffage and figure out what we want it to be. 16:57:16 thats a great idea if we can get one to help us. ;) 16:57:18 who is going to be at Fudcon NA 16:57:59 I am hoping to... not sure on details at this point. 16:58:33 * tremble was contemplating going to Fudcon Zurich... 16:59:30 I am hoping for 2011 Fudcon EU.. 17:00:14 smooge: im not planning on being at FUDCon NA 17:00:23 FudCon Lichtenstein 17:00:39 oh dgilmore I didn't know that 17:00:52 dgilmore: bummer. ;( 17:01:44 dgilmore: scheduling conflict? 17:01:48 * dgilmore is planning to be at LCA 17:02:03 if lca falls though ill be at FUDCon 17:02:14 LCA? 17:02:15 but if all works out ill be in Australia 17:02:21 Linux.conf.au 17:02:28 oooooooooh 17:02:49 ah, ok. 17:02:51 that would be nice to go to also. maybe I should look at that for 2012 17:03:33 ok well I was thinking we have an EPEL FAD during the con and clean up stuff there. 17:04:16 we talked about doing a FAD sometime... 17:05:34 perhaps we could put one together for later this year somewhere... 17:06:43 yeah.. I am trying to remember when the fiscal year begins/ends to see what would work better. I think the fiscal year is in March so this year makes sense still 17:08:36 ok will put that at the next meeting. 17:08:51 going to close in 30.. 17:09:16 yeah. since you and I are west, somewhere out here might make sense... 17:10:13 ok will put down for FAD-ABQ 17:10:23 #closemeeting 17:10:31 #endmeeting ?