15:00:21 <ttx> #startmeeting releaseteam 15:00:22 <openstack> Meeting started Fri Jun 30 15:00:21 2017 UTC and is due to finish in 60 minutes. The chair is ttx. Information about MeetBot at http://wiki.debian.org/MeetBot. 15:00:24 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 15:00:26 <openstack> The meeting name has been set to 'releaseteam' 15:00:31 <ttx> Ping list: dhellmann, dims, fungi, tonyb, stevemar, lbragstad 15:00:38 <lbragstad> o/ 15:00:43 <fungi> mmmeetings 15:00:53 <stevemar> o/ 15:01:05 <ttx> Agenda under R-9 at: 15:01:10 <ttx> #link https://etherpad.openstack.org/p/pike-relmgt-tracking 15:01:21 <ttx> Feel free to add stuff if you have anything to discuss 15:01:27 <ttx> since it's pretty empty 15:01:49 <ttx> #topic dependencies on cycle-with-milestones stuff 15:02:00 <ttx> Looks like we removed the last of the preblamtic packages 15:02:09 <ttx> problematic 15:02:24 <ttx> with tripleo now depending on mistral-lib rather than mistral 15:02:44 <ttx> I don't think we have anything more, so prometheanfire shall be happy 15:02:51 <ttx> i'll drop a note on #openstack-requirements 15:02:52 <dhellmann> o/ 15:02:59 <ttx> Anything else to do on that topic ? 15:03:29 <dhellmann> nope, sounds good 15:03:51 <ttx> ok, note dropped 15:04:00 <ttx> #topic Look at stuck reviews 15:04:22 <ttx> Since this is a calm week agenda-wise I thought we could have a look at stale reviews 15:04:51 <dhellmann> sfinucan has been doing some work on reno to add the setuptools commands we need. I've been tied up with the doc migration work, so if someone else has time to look at that it would be good. I expect to get to them next week. 15:04:57 <ttx> and decide what to do with them, if anything 15:05:17 <ttx> so yeah, let's start with reno 15:05:44 <ttx> dhellmann: that one still is W-1 ? 15:05:59 <dhellmann> oh, that part of my window was covered so I didn't notice that 15:06:19 <ttx> https://review.openstack.org/#/c/448066/ was refreshed, yes 15:06:22 <dhellmann> the prelude section renaming patch should be good to review, too 15:06:28 <ttx> (Support repodir config files) 15:06:58 <ttx> right, so the stuck ones would be: 15:07:13 <ttx> https://review.openstack.org/#/c/433247/ ( 15:07:15 <ttx> Adds Source Links to Reports) 15:07:35 <ttx> https://review.openstack.org/#/c/350827/ (optionally build the cache file when building an sdist) 15:08:05 <dhellmann> I think 827 can probably be abandoned in favor of 247 15:08:19 <dhellmann> sorry, in favor of 066 15:08:33 <ttx> yeah, was just thinking about that 15:08:45 <ttx> I'll let you do it since it's yours 15:08:56 <dhellmann> done 15:09:31 <ttx> for 247, https://review.openstack.org/437241 was merged instead 15:10:12 <ttx> hm, maybe I misread 15:10:23 <dhellmann> yeah, I wasn't sure if we wanted to update that to include the full url or not 15:10:36 <dhellmann> but it does essentially the same thing by showing which file included which note 15:10:54 <ttx> I can leave a note asking if 247 is enough for him 15:10:55 <dhellmann> https://review.openstack.org/433247 is 4 months old, I think we could abandon it 15:10:58 <dhellmann> sure 15:11:04 <ttx> err 241 15:11:35 <ttx> ok done 15:11:42 <dhellmann> there are a few patches for the release-tools repo as well 15:11:56 <ttx> the rest of the reno stuff is fairly active 15:12:06 <ttx> yes,, moving up to release-tools 15:12:38 <ttx> https://review.openstack.org/#/c/464949/ Fixing TypeError: unorderable types: int() > str() 15:13:32 <dhellmann> I can't tell if that's a real bug fix or what. 15:13:39 <ttx> dhellmann: you seem to understand what's going on , would you have enough information to fix it if the submitter doesn't come back ? 15:14:01 <fungi> i'm guessing that's to accommodate prerelease version strings, or maybe not break when encountering things like eol tags? 15:14:05 <ttx> I'm not sure the fix fixes it 15:14:08 <dhellmann> the patch looks fine, it's just that I can't figure out what code path actually triggered it 15:14:17 <ttx> Looks more lie a band-aid for a case that should not happen 15:14:21 <dhellmann> right 15:14:42 <dhellmann> and actually, using str is not a good idea there either 15:14:53 <dhellmann> so the patch itself isn't fine, which is why I asked for more detail 15:15:08 <dhellmann> they need to be the same type, but I think they need to be ints 15:15:47 <dhellmann> the only tool in that repo that uses semver.py is sanity_check_version.py and we don't use that any more 15:15:51 <ttx> ok, left a new ping on it 15:16:00 <fungi> yeah, having an actual reproducer first is pretty necessary there 15:16:18 <ttx> https://review.openstack.org/#/c/464967/ Using fixtures instead of deprecated mockpatch module 15:16:20 <dhellmann> right 15:17:09 <ttx> I have no idea of the oslotest mockpatch situation, so can't arbitrate between Rajat and Dai Dang 15:17:24 <dhellmann> I think the suggestion there is just to move the import statement 15:17:46 <dhellmann> and oslotest is definitely deprecated 15:18:18 <dhellmann> I don't know how strongly we feel about having the import statements organized 15:18:32 <dhellmann> other than that, I would be +2 on this 15:18:40 <dhellmann> if we don't care, I can be +2 anyway 15:19:12 <ttx> yeah, I don't care at all 15:19:21 <ttx> especially on that specific file 15:19:45 <ttx> i suspect Rajat just dropped patches and disappeared 15:19:48 <dhellmann> ok, I left a +2 15:19:52 <dhellmann> I think so, too 15:20:00 <ttx> https://review.openstack.org/#/c/464954/ Add 'rm -f .testrepository/times.dbm' command in testenv 15:20:07 <dhellmann> I'm opposed to that 15:20:16 <dhellmann> it's not the right solution to that problem 15:21:21 <ttx> we should shame lifeless into fixing those 15:21:54 <dhellmann> or just stop using python 2 15:22:25 <ttx> ok just -1ed it in preparation for abandonment 15:22:32 <fungi> i think the commit message has it backwards anyway 15:22:48 <fungi> generally running py35 and then py27 works 15:22:54 <dhellmann> true 15:22:56 <fungi> it's the other way around where it breaks 15:23:03 <ttx> https://review.openstack.org/#/c/464601/ Remove extral spaces when 15:23:28 <fungi> and the workaround most projects have adopted is to add the py35 job first in the global envlist at the top 15:23:31 <dhellmann> we have 2 different tools that edit yaml files, and they do it in different ways so they produce different results 15:23:43 <dhellmann> this one tries to make the "simpler" tool work the same way as the interactive tool 15:24:12 <dhellmann> but we still have a mix of data files, which means we'd have a bunch of inconsistent files that would have to be touched to make them work with the tool after this change 15:24:25 <ttx> ok, so this one is rightly wip 15:24:34 <dhellmann> if we're going to do something like this, we should do it at the release boundary, when we start with blank files 15:24:36 <ttx> I'll add it to the plan backlog 15:24:54 <dhellmann> then we would only have to worry about the few stable releases 15:25:18 <dhellmann> I don't know if we want to try to standardize the whitespace in the files to deal with that 15:25:39 <ttx> ok, I added it to the relmgt plan backlog so that we don't lose track of it 15:25:43 <dhellmann> k 15:25:54 <ttx> sound slike a bug that we should fix sooner or later 15:26:03 <dhellmann> yeah, it's an annoyance 15:26:18 <fungi> if you're going to get picky about consistent yaml formatting, it's probably time to look into adding a yamllint run to the style checks 15:26:33 <dhellmann> I don't know if anyone other than harlowja uses that interactive tool for preparing releases 15:26:43 <dhellmann> fungi : we do have yamllint 15:27:02 <fungi> oh, right on 15:27:05 <ttx> all other releases requetss are relatively fresh, so nothing is really silently rotting 15:27:09 <dhellmann> that's the thing that breaks when you mix indentation styles :-) 15:27:30 <fungi> got it, so that's what he means by "doc test" 15:27:54 <dhellmann> fungi : yes, I think so 15:28:48 <ttx> #topic Open discussion 15:28:58 <ttx> Alright, anything else ? 15:29:12 <ttx> I assume most of you in the US will be off on Tuesday ? 15:29:12 <dhellmann> asettle has agreed to a basic schedule for the migration work 15:29:27 <dhellmann> we're going to try to get teams to update by M3 so that it is all done before branching at RC1 15:29:34 <dhellmann> that avoids backporting documentation patches 15:30:10 <dhellmann> we're going to proceed next week with removing the migrated documentation from openstack-manuals, and instruct teams to check out the before-migration tag to get to the files they need to move if they haven't already started their work 15:30:20 <dhellmann> yes, I will be off Tuesday 15:31:07 <dhellmann> can we include a reminder in the release announcement emails to encourage folks to pick up the migration work? 15:31:35 <ttx> hmm, I did not have one scheduled for next week, and already sent this week's 15:31:50 <dhellmann> ok. we'll just rely on another email from asettle then 15:32:00 <ttx> I can add it to the Jul 14 one 15:32:05 <dhellmann> ok 15:32:09 <asettle> dhellmann: hey yo what? 15:32:12 <asettle> Sorry, I'm in meetings all afternoon 15:32:36 <dhellmann> asettle : np, I was asking ttx to include a reminder in the release countdown email and letting him know about the schedule we worked out with the m3 deadline 15:33:07 <ttx> OK, anything else to cover in meeting ? 15:33:19 <dhellmann> nothing from me 15:33:33 <ttx> alright then 15:33:37 <ttx> #endmeeting