cockpit_weekly_meeting_160321
LOGS
14:03:53 <andreasn> #startmeeting Cockpit weekly meeting 160321
14:03:53 <zodbot> Meeting started Mon Mar 21 14:03:53 2016 UTC.  The chair is andreasn. Information about MeetBot at http://wiki.debian.org/MeetBot.
14:03:53 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic.
14:03:53 <zodbot> The meeting name has been set to 'cockpit_weekly_meeting_160321'
14:04:06 <dperpeet> .hello dperpeet
14:04:07 <zodbot> dperpeet: dperpeet 'None' <dperpeet@redhat.com>
14:04:13 <andreasn> .hello andreasn
14:04:14 <zodbot> andreasn: andreasn 'Andreas Nilsson' <anilsson@redhat.com>
14:04:27 <stefw> .hello stefw
14:04:27 <zodbot> stefw: stefw 'Stef Walter' <stefw@redhat.com>
14:04:53 <andreasn> #topic agenda
14:05:35 <dperpeet> * known issues
14:05:44 <dperpeet> * selinux troubleshooting
14:06:00 <mulhern> .hello mulhern
14:06:01 <zodbot> mulhern: mulhern 'None' <amulhern@redhat.com>
14:06:17 <stefw> * GSoC
14:06:21 <mulhern> yup, that's me.
14:06:32 <stefw> * Debian builds
14:06:43 <dperpeet> mulhern, yay, I'm not the only 'None' person here.
14:06:50 <mulhern> :)
14:07:59 <andreasn> ok, lets start
14:08:04 <andreasn> #topic Known Issues
14:08:14 <dperpeet> This is related to our integration tests
14:08:29 <dperpeet> as you know, some issues happen fairly often and thus clobber the github issue threads
14:08:40 <dperpeet> that made us learn about the comment limit =)
14:08:56 <dperpeet> I've done some work to reduce the verbosity: https://github.com/cockpit-project/cockpit/pull/4049
14:09:10 <dperpeet> Posts look like this: https://github.com/cockpit-project/cockpit/issues/4001#issuecomment-199257031
14:09:28 <dperpeet> the idea is to have one comment per context (e.g. fedora-atomic or fedora-24)
14:09:41 <dperpeet> this will get updated as the error occurs
14:09:59 <dperpeet> and it'll keep the first and last 10 occurrences of the error
14:10:07 <dperpeet> with links to the logs, if present
14:10:32 <dperpeet> for each traceback, there'll be a different "section"
14:10:34 <stefw> looks much nicer
14:10:40 <dperpeet> thanks
14:10:54 <dperpeet> as an added bonus, I think comment edits don't generate notification e-mails
14:11:15 <dperpeet> if you're waiting for an error to happen, this might be a bad thing, so let me know if this concerns you
14:11:34 <dperpeet> we could add a mechanism to ping users
14:11:51 <dperpeet> but in my opinion our problem was/is rather too much notification traffic
14:12:03 <dperpeet> end of topic from my side
14:12:36 <andreasn> #topic SELinux Troubleshooting
14:12:49 <dperpeet> There's still the work in progress here: https://github.com/cockpit-project/cockpit/pull/3962
14:13:08 <dperpeet> updated test images have been merged today https://github.com/cockpit-project/cockpit/pull/4046
14:13:34 <dperpeet> and last weekend I figured out (thanks petervo) how to make the integration tests work
14:13:56 <dperpeet> we need(ed) a bind shim in cockpit.js https://github.com/cockpit-project/cockpit/pull/3962/files#diff-34cbdb779d20a4a75591fcba50bd25d4
14:14:10 <dperpeet> so phantomjs works with the bind syntax I used
14:14:43 <dperpeet> I've opened a pull request with the setroubleshoot project to fix some message capitalization that I talked about with andreasn: https://github.com/fedora-selinux/setroubleshoot/pull/11
14:15:07 <dperpeet> and we'll probably get another dbus api update soon
14:15:23 <andreasn> I just subscribed to that pull
14:15:30 <dperpeet> so we get alert severity, timestamps and the ability to dismiss alerts properly
14:15:41 <andreasn> we also spoke about some layout fixes on irc on Friday
14:15:58 <dperpeet> I'm in dialog with plautrba for a workaround that lets us dismiss alerts on current systems
14:16:08 <andreasn> cool
14:16:39 <dperpeet> yes, I've been iterating on the layout and design with andreasn
14:17:06 <dperpeet> this will probably have to change in a followup as we play around with it some more
14:17:11 <dperpeet> end of topic from my side
14:17:21 <andreasn> nice
14:17:34 <dperpeet> oh, plautrba was happy with the current state of how we implemented it
14:17:40 <andreasn> #topic GSoC
14:17:47 <andreasn> nice to hear!
14:17:56 <dperpeet> the student registration deadline is fast approaching for GSoC
14:18:17 <dperpeet> as expected, activity is picking up and we're seeing more potential applicants in irc
14:18:40 <dperpeet> thanks for your interest, students!
14:19:06 <dperpeet> the timeline is here https://developers.google.com/open-source/gsoc/timeline
14:19:38 <dperpeet> and as a reminder, the Cockpit topics are here: https://fedoraproject.org/wiki/Summer_coding_ideas_for_2016#Idea_list_for_Cockpit
14:19:51 <dperpeet> as part of the Fedora org in GSoC 2016
14:20:17 <andreasn> #link https://fedoraproject.org/wiki/Summer_coding_ideas_for_2016#Idea_list_for_Cockpit
14:20:33 <dperpeet> petervo, anything to add?
14:20:50 <petervo> nope
14:20:53 <dperpeet> maybe to make it clear: 25 March 19:00 UTC Student application deadline.
14:21:08 <dperpeet> after that the internal discussion will begin
14:21:37 <dperpeet> end of topic from my side
14:21:56 <andreasn> #topic Debian builds
14:22:13 <abhishekg5> dperpeet has already given feedback on my shared proposal. petervo, would you like to add anything ?
14:22:41 <stefw> larsu, anything of note on the debian package builds?
14:22:53 <larsu> yep, I have the release script generate a binary repository for unstable
14:22:54 <dperpeet> abhishekg5, as feedback is proposal specific, any mentor who wants to comment can do so in the document itself
14:23:07 <larsu> and copy that to fedorapeople.org (which obviously I can't test yet)
14:23:23 <larsu> putting on the finishing touches right now
14:23:38 <stefw> cool. by the way it's easy to get a fedora account
14:23:40 <dperpeet> larsu, why can't you test copying to fedorapeople?
14:23:48 <stefw> https://admin.fedoraproject.org/accounts/user/new
14:23:55 <abhishekg5> dperpeet, okay cool
14:24:01 <andreasn> larsu: do you have a url for that? I have a friend who's a debian guy and he's really interested
14:24:14 <andreasn> larsu: he declined my offer for him to maintain it though :)
14:24:30 <larsu> dperpeet: need to make an account :)
14:24:34 <andreasn> really interested in trying cockpit that is
14:24:51 <larsu> dperpeet: also, I'd want it under the cockpit-project user
14:24:51 <dperpeet> larsu, let me know if you need help with that after the meeting
14:25:09 <larsu> andreasn: soon. I'll post to cockpit-devel, but can let you know separately, too
14:25:15 <larsu> dperpeet: thanks
14:25:16 <dperpeet> I've fared well trying new stuff under my own fedora user first :)
14:25:17 <andreasn> larsu: thanks!
14:25:24 <dperpeet> that way you can clean up your own home nicely
14:25:46 <larsu> oh, I do have an account. Does that come with fedorapeople space?
14:25:58 <dperpeet> it should
14:26:09 <larsu> cool :)
14:27:07 <dperpeet> I'm looking forward to getting that repo up and running automatically
14:27:24 <dperpeet> thanks for wading through those steps!
14:27:46 <larsu> ya! biggest problem was actually a weird test failure inside the chroot
14:30:03 <andreasn> anything else on Debian?
14:30:21 <larsu> don't think so
14:30:23 <andreasn> #topic Open Floor
14:31:09 <sgallagh> Mostly just an FYI: Cockpit wasn't accessible by default on F24 alpha builds until the other day, but this was my fault not yours. It's fixed.
14:31:43 <stefw> good to know
14:31:47 <andreasn> is it working well in general?
14:32:05 <sgallagh> andreasn: A few hiccups I reported last week
14:32:09 <andreasn> what's the easiest way to get my hands on the install image?
14:32:13 <sgallagh> Particularly around some NetworkManager stuff
14:32:30 <sgallagh> andreasn: https://kojipkgs.fedoraproject.org/compose/24/Fedora-24-20160316.3/compose/
14:32:38 <andreasn> thanks!
14:32:39 <sgallagh> Specifically Server: https://kojipkgs.fedoraproject.org/compose/24/Fedora-24-20160316.3/compose/Server/x86_64/iso/
14:32:58 <sgallagh> (This one is fixed and launches Cockpit properly)
14:33:14 <andreasn> nice
14:33:37 <andreasn> that's all I guess
14:33:40 <dperpeet> wait
14:33:43 <dperpeet> I noticed one thing today
14:33:54 <dperpeet> I couldn't get fedora-testing images to build
14:34:06 <dperpeet> but that might be related to the fact that we're basing them off virt-builder fedora-23 images
14:34:15 <dperpeet> I didn't dive deeped, just fyi in case anyone else runs into this
14:34:26 <dperpeet> *dive deeper
14:34:56 <dperpeet> thanks
14:35:30 <andreasn> thanks for the meeting, everyone!
14:35:33 <andreasn> #endmeeting