20:29:59 #startmeeting requirements 20:30:00 Meeting started Wed Sep 18 20:29:59 2019 UTC and is due to finish in 60 minutes. The chair is prometheanfire. Information about MeetBot at http://wiki.debian.org/MeetBot. 20:30:01 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 20:30:03 The meeting name has been set to 'requirements' 20:30:12 #topic rollcall 20:30:13 tonyb, prometheanfire, dirk, smcginnis ping 20:30:15 o/ 20:31:42 o/ 20:32:12 #topic Any controversies in the Queue? 20:32:49 Only controversial thing I've seen is how many rechecks we need to do to actually get an update through. 20:32:55 I'd say no, we are in freeze, so, eh... 20:33:23 prometheanfire: Speaking of freeze, is it OK if I approve this - https://review.opendev.org/#/c/680497/ 20:34:04 I'd say yes 20:34:29 Looks useful for keystone at least. 20:34:34 would have been better for a new requests-mock release to have been done, but meh 20:34:42 Yeah 20:35:19 just checked, ya, still no release 20:36:06 Actually, looks like 1.7.0 had it? 20:36:11 https://github.com/jamielennox/requests-mock/releases/tag/1.7.0 20:36:38 it didn't have the fix I don't think 20:36:38 Oh wait, the issue is with 1.7.0. 20:36:43 Nevermind, ignore me. :] 20:36:44 yep 20:36:46 :D 20:37:01 #topic branching 20:37:07 gonna work on https://etherpad.openstack.org/p/requirements-cycle-process 20:37:57 will involve updating tox for all repos, branching, and updating that public link that tox uses 20:38:07 so expect that over the next weekish 20:38:20 ++ 20:38:33 #topic open floor 20:38:39 Wait, isn't the tox update automatic now? 20:38:57 smcginnis: not sure, it's something to verify 20:39:44 Looks like it - https://review.opendev.org/#/c/682312/1/tox.ini 20:40:19 right, but they will still need to update to whatever we named U 20:40:45 That's on the new stable/train branch, so master still points to master constraints. 20:40:55 U is officially Ussuri. 20:41:13 ya, but once the projects all branch they need to update the link to usuri 20:41:20 after we branch as well 20:41:45 I thought they stay pointed to master, then when the stable branch is created it moves to the cycle name. 20:41:55 so we can branch now if we wanted 20:41:58 Guess I'm missing some part of the machinery. 20:42:24 it could go that way, but that causes binding / pileup at release time 20:43:04 So we would skip those "Update TOX/UPPERCONSTRAINTS_FILE for stable/*" patches that are proposed now? 20:43:15 And just move it to happen right away at freeze? 20:43:36 Which is actually right around the same time, so I'm still not sure I'm following how this reduces patches. 20:43:45 over the next six months after we branch we have the projects change master to point to ussuri 20:44:11 Ah, so just spread that out over the course of the cycle instead of all at once? 20:44:11 then we can branch whenever and not have to wait on a random project 20:44:29 doesn't reduce patches, just spreads it out, ya 20:44:56 I think with the static link default to master for any unknown names, we're at least protected to be able to branch requirements at any point. 20:45:24 Random project would then just be testing against master, which should be fairly stable for a little while at least. 20:46:20 Oh man, I shouldn't have looked at https://review.opendev.org/#/q/topic:create-stein+status:open 20:46:26 :D 20:46:35 ya, this whole thing needs to be plotted out 20:46:43 I'm gonna be working at that over the next few days 20:47:24 OK, cool. Just let me know if you need someone to take a look at anything. 20:48:01 yarp 20:48:12 anything else? 20:48:18 nada 20:48:31 sorry I couldn't make last weeks... (power outage), first long one since I moved in 20:48:48 Funny, I just had one yesterday. First time in at least over a year. 20:49:06 Infrequent enough to get me to buy a UPS. :) 20:49:19 Or rather, not buy a UPS. 20:50:00 mine last 15-20 minutes and that was not long enough, would have needed a generator 20:50:04 #endmeeting