14:59:45 #startmeeting 14:59:45 Meeting started Wed Dec 5 14:59:45 2012 UTC. The chair is mattdm. Information about MeetBot at http://wiki.debian.org/MeetBot. 14:59:45 Useful Commands: #action #agreed #halp #info #idea #link #topic. 14:59:56 #meetingname cloud-sig 14:59:56 The meeting name has been set to 'cloud-sig' 15:00:04 #topic Cloud SIG meeting 15:00:15 #meetingtopic Cloud SIG meeting 15:00:33 Good morning! Anyone around? 15:00:59 * tdawson waves his hand. 15:01:02 Good morning 15:01:11 * mattdm waves back 15:02:09 I think we're going to need to have a mailing list discussion about how to make these meetings more useful to people. 15:02:25 Or whether we want to conduct them at all in the absence of pressing concerns 15:02:49 True 15:03:00 It doesn't do too much if I'm the only other person that comes. 15:04:42 Yep. I think I'll aim instead for more mailing list traffic about wht's going on with various things. 15:04:51 And by "aim", I mean "generate" :) 15:05:21 In the meantime, how's the openshift world? :) 15:08:30 tdawson: yeah looks like it's just us today 15:08:40 got anything of interest? 15:08:49 We're trying to get it working on F17 15:09:04 F17 has the correct version of rails, and we know it works. 15:09:23 I'm hoping I'll have all the packages at least in F17 testing by the end of the week. 15:09:43 We're still working on the rails 3.2 problems for F18 and F19. 15:09:46 cool. 15:10:03 A good example of how moving too fast in some areas prevents us from moving quickly in others. *sigh* 15:10:44 With the last OpenShift release, alot of our Fedora changes (changes so the spec file's match Fedora's) are in the OpenShift release, so these updates are much easier than they were before. 15:11:00 mattdm: Yep ... sometimes bleeding edge bits you. 15:11:28 do the problems need changes in openshift or in rails and other packages? (or both?) 15:11:49 It's needing changes in our openshift code. 15:12:01 The new code should work on both older and newer versions of rails. 15:12:13 * mattdm nods 15:12:45 There was also some httpd 2.4 issues, but those were fairly minor and fixable via spec file %if statements. 15:12:49 So, with the software collections thing -- what's the general plan for that and Fedora? 15:13:28 Well ... once we get it working and stable somewhere, we'll try to keep it fairly stable for that release. 15:13:53 Meaning, hopefully when we get this last release into F17, we'll hopefully keep that stable and not update it much. 15:14:16 But we'll try to get each release into rawhide. 15:14:18 Will there be separate infrastructure and repos for maintaining the scl cartridges? 15:15:12 Well, that will be for RHEL and EPEL. We aren't totally sure what to do for that. Right now those packages will be in their own openshift repo. 15:15:51 I would really like to see the issues with software collections resolved so we can use them in fedora. 15:16:09 True ... if that happens, then we can put these into EPEL. 15:16:40 As a general sysadmin using scl ... it "works" ... but it isn't elegant or pretty. 15:17:00 OK, maybe I should put works in quotes, it really does work. 15:17:23 Do you have ideas on making it more elegant and pretty? 15:17:28 But it does occasionally require jumping through a few more hoops that you expect. 15:18:11 I wonder if an integration layer with the existing environment-modules package would help. 15:18:22 Ya know ... I really don't. I understand the premise behind it, and I don't know how to make it better. 15:18:33 That's a thing from the HPC community which aims to solve the same problem and which has been around for a while. 15:18:58 (And which is in Fedora and RHEL for mpi support) 15:19:04 Well, from my work at Fermilab, we had UPD/UPS that did something simliar. 15:19:17 mattdm: What is that? 15:19:36 http://fedoraproject.org/wiki/PackagingDrafts/EnvironmentModules 15:22:31 I'm going to file an RFE request against scl and see where that goes, and maybe raise a general "so, how _can_ we make this work" thread on fedora devel. 15:22:51 anything else? if not, I think we can declare this small meeting over. 15:24:59 Nope, that's it for me. 15:25:06 okay then :) 15:25:11 #endmeeting