| *** dims has joined #openstack-relmgr-office | 00:43 | |
| *** AK has quit IRC | 01:15 | |
| *** dims has quit IRC | 02:32 | |
| *** dims has joined #openstack-relmgr-office | 02:33 | |
| *** AzherKhan has joined #openstack-relmgr-office | 02:37 | |
| *** dims has quit IRC | 02:41 | |
| *** AzherKhan has quit IRC | 05:34 | |
| *** AzherKhan has joined #openstack-relmgr-office | 06:04 | |
| ttx | dhellmann: about https://review.openstack.org/#/c/173972/ -- which dep does it raise the minimum version of ? | 06:42 |
|---|---|---|
| openstackgerrit | Merged openstack-infra/release-tools: Add release_many.sh https://review.openstack.org/189858 | 08:16 |
| openstackgerrit | Merged openstack-infra/release-tools: let the caller specify the project owner https://review.openstack.org/191809 | 08:17 |
| openstackgerrit | Merged openstack-infra/release-tools: use announce list instead of dev list https://review.openstack.org/191815 | 08:17 |
| openstackgerrit | Merged openstack-infra/release-tools: Add list_repos_by_tag.py https://review.openstack.org/191227 | 08:17 |
| openstackgerrit | Thierry Carrez proposed openstack-infra/release-tools: Add adjust_blueprints.py script https://review.openstack.org/192686 | 08:20 |
| *** dims has joined #openstack-relmgr-office | 08:44 | |
| *** dims has quit IRC | 08:48 | |
| *** isviridov_away is now known as isviridov | 08:56 | |
| *** dims has joined #openstack-relmgr-office | 10:00 | |
| dims | ttx: does this mox3 into governance need more rollcall votes? https://review.openstack.org/#/c/190344/ | 10:26 |
| ttx | dims: no I can approve it now | 10:33 |
| ttx | done | 10:33 |
| dims | thanks ttx | 10:34 |
| *** AK has joined #openstack-relmgr-office | 11:05 | |
| *** AzherKhan has quit IRC | 11:07 | |
| *** AK has quit IRC | 11:11 | |
| *** isviridov is now known as isviridov_away | 12:25 | |
| *** dims has quit IRC | 12:40 | |
| *** dims has joined #openstack-relmgr-office | 12:41 | |
| dhellmann | ttx: http://lists.openstack.org/cgi-bin/mailman/admin/openstack-announce doesn't have a "send me a password" link. Is that on another page? | 13:15 |
| mestery | Good morning ttx and dhellmann! I see you've figured out the versioning changes, how do I fix some of the networking-foo repos (e.g. like this one) https://review.openstack.org/#/c/192419/ | 13:15 |
| mestery | I have to tag them with a pre version, right? | 13:16 |
| dhellmann | ttx: regarding 173972, the older version of the patch raised mins but this one doesn't. Still, to land it we'll have to do something special because it doesn't match the global requirements. I proposed that we turn off the requirements check job for that repo, since I think we did that for another project's stable branch | 13:16 |
| ttx | dhellmann: hmm, I'm a bit lost, can't remember how I got mine | 13:16 |
| dhellmann | ttx: afaik, there's only one password per list | 13:16 |
| ttx | dhellmann: ok, let me retrieve that from password cache | 13:17 |
| dhellmann | mestery: ah, I thought those repos were libs, so I didn't update them. What are they? | 13:17 |
| ttx | mailman is the only thing I ever store passwords in my browser for | 13:17 |
| mestery | dhellmann: Yes, they are libs, but some of them (odl and ofagent for sure) have already released to pypi as 2015.x versions | 13:18 |
| dhellmann | oh, that's going to make the requirements management interesting | 13:18 |
| mestery | Yeah :( | 13:18 |
| mestery | Well, we released those to test out the release process and match the old versioning during Kilo :) | 13:19 |
| dhellmann | mestery: one way we can handle it is releasing the new version then capping requirements < 2000 | 13:19 |
| dhellmann | yeah | 13:19 |
| dhellmann | lifeless may have another idea for us -- I don't know if pip supports epochs | 13:19 |
| mestery | So we'd release the new 1.0.0 version for both? | 13:19 |
| dhellmann | yes | 13:19 |
| mestery | OK, so as you see, those patches are failing now | 13:20 |
| dhellmann | mestery: yeah, that part's easy to fix with a tag, which I can do later today | 13:21 |
| dhellmann | today is my commute day, so after breakfast I'm going to disappear for a while | 13:21 |
| dhellmann | openstack atl meets tonight | 13:21 |
| mestery | cool! | 13:22 |
| mestery | No worries, I'll get patches up for the other repos that have this problem and make them the same topic | 13:22 |
| mestery | Thanks dhellmann! | 13:22 |
| dhellmann | mestery: I see a bunch of networking-* repos, should I do them all? | 13:22 |
| dhellmann | oh, heh, scrollback -- sounds like we're on the same page | 13:23 |
| mestery | dhellmann: I think only a few have release to pypi, so only those are the ones to worry about (I think), right? | 13:23 |
| mestery | yeah | 13:23 |
| mestery | :) | 13:23 |
| mestery | dhellmann: https://review.openstack.org/#/q/topic:semver-releases+owner:%22Kyle+Mestery%22,n,z | 13:33 |
| mestery | dhellmann: Those are the ones, thanks for hte help! | 13:33 |
| mestery | Some of them had versions of 0.1, which I left alone, is that ok? | 13:33 |
| mestery | All the ones there I moved to 1.0.0 | 13:33 |
| dhellmann | mestery: are those libs going to be released independently, or on milestones? | 13:39 |
| mestery | dhellmann: Indepdently | 13:39 |
| mestery | Similar to other clients and libs | 13:39 |
| dhellmann | ok, in that case I think what we want to do is remove the version setting from the cfg file and just use tags | 13:39 |
| *** Kiall has joined #openstack-relmgr-office | 13:44 | |
| mestery | dhellmann: You mean for the ones with versions of 0.1? | 13:44 |
| mestery | or all of them? | 13:44 |
| Kiall | ttx: Q - If projects e.g. ironic now do 2x major releases within a dev cycle, what happens with stable branches? e.g. Dev cycle == Liberty, but Ironic releases v2.0.0, 2.0.1, and 3.0.0 during the cycle.. stable/liberty doesn't really fit anymore? Q arises due to how we track upstream code internally.. :) | 13:50 |
| ttx | Kiall: we would cut stable branch from the last release in the dev cycle | 13:53 |
| ttx | basaiclly there is still a common dev cycle | 13:53 |
| ttx | they may do intermediary releases but those wouldn't get a stable branch | 13:54 |
| ttx | "major" or not major | 13:54 |
| Kiall | ttx: Okay, makes sense. But- some major releases may end up never seeing a stable branch then? e.g. 2.0.0 comes out, master lands new features for 3.0.0, but Liberty hasn't ended - and a 2.0.1 is needed for $reason? | 13:55 |
| ttx | users of intermediary releases are on their own wrt stable backports | 13:57 |
| ttx | it's the tradeoff there. You get bleeding edge but you're a bit on your own | 13:57 |
| ttx | If you don't like that, stick with end-of-cycle releases | 13:57 |
| ttx | you should always be able to jump from last stable to current stable | 13:58 |
| Kiall | Okay - Perfect :) Nah, It's just that we use the dev cycle names for some things internally, and things could have been in need of rework :) | 13:58 |
| ttx | sice that's the only thing grebade tests | 13:58 |
| ttx | grenade* | 13:58 |
| ttx | If you stay on 6-month things do not really change basically | 13:58 |
| ttx | you just get interdiary tags and ignore them | 13:59 |
| ttx | intermediary | 13:59 |
| ttx | if you deploy those tags though, yes, you need extra care | 13:59 |
| Kiall | Perfect! Thanks, nice and clear. | 14:00 |
| *** openstackgerrit has quit IRC | 16:22 | |
| *** openstackgerrit has joined #openstack-relmgr-office | 16:23 | |
| *** samueldmq has quit IRC | 16:37 | |
| dhellmann | mestery: eventually all of them, but we can do that as we tag the next releases | 17:15 |
| mestery | dhellmann: OK, cool. | 17:19 |
| *** openstackgerrit has quit IRC | 17:50 | |
| *** openstackgerrit has joined #openstack-relmgr-office | 17:51 | |
| pabelanger | ttx, dhellmann any low hanging fruit you need help with? | 18:06 |
| dhellmann | pabelanger: we're working on the project guide today in #openstack-sprint | 18:22 |
| pabelanger | dhellmann, thanks, checking it out now | 18:26 |
| lifeless | morning | 18:51 |
| lifeless | pbr doesn't support epochs yet | 18:51 |
| lifeless | IIRC PEP-440 does specify them | 18:51 |
| lifeless | reading backlog now | 18:51 |
| lifeless | you can delete things from pypi, you just can't reuse | 18:52 |
| lifeless | so you could rewind - but there will be a set of folk who don't see the rewind (because they are using the -f form of wheel cache for instance) | 18:53 |
| lifeless | to get them to see it is either manual, or set a max bound lower than the old versions | 18:53 |
| lifeless | e.g. odl <2015 | 18:53 |
| dhellmann | lifeless: yeah, I think we'll probably want to go with a cap, I was thinking <2000 for things using liberty releases | 19:34 |
| dhellmann | hmm, mestery, I started tagging those networking-* repos but thinking about it again I think we want to just remove the version number there and go with tags | 20:14 |
| dhellmann | which may mean waiting until those teams are actually ready for a release | 20:14 |
| mestery | dhellmann: Ack, that makes sense! So, do nothing for now, but before a release circle back and take care of it? | 20:15 |
| dhellmann | mestery: yeah. and I guess you have a free 1.0.0a0 release of networking-odl :-/ | 20:19 |
| dhellmann | mestery: I'll update the patches to remove the setting, and then WIP them for now | 20:19 |
| mestery | dhellmann: Thanks Doug! Appreciate all your help and guidance here. | 20:19 |
| dhellmann | mestery: I'm going to remove your signed-off-by since I'm substantially changing the commits, ok? | 20:22 |
| mestery | dhellmann: Ack, no worries there. | 20:22 |
| dhellmann | mestery: ugh, I don't have wip permission on that repo? | 20:23 |
| dhellmann | mestery: https://review.openstack.org/#/c/192419/ | 20:23 |
| dhellmann | mestery: maybe it'd be better for you to do these, since I'm not core or the original author | 20:24 |
| mestery | dhellmann: OK, I'll do them all :) | 20:24 |
| * mestery looks | 20:24 | |
| dhellmann | mestery: sorry, thought this would be simple | 20:24 |
| mestery | dhellmann: Give me a few minutes, finishing something I was in the middle of right now | 20:24 |
| mestery | no worries | 20:24 |
| mestery | We'll get it all done ;) | 20:24 |
| dhellmann | mestery: sounds good, no rush | 20:24 |
| mestery | dhellmann: I think I got them all, even the ones which had versions like 0.1. I marked them all WIP for now. | 20:35 |
| dhellmann | mestery: looking | 20:41 |
| dhellmann | mestery: LGTM | 20:41 |
| mestery | sweet! | 20:42 |
| mestery | dhellmann: Some of them pass due to the projects having an existing 0.1 version, can we just merge those ones which remove that version and move to tags? | 20:42 |
| dhellmann | mestery: I didn't know what octavia or dragonflow were | 20:43 |
| mestery | :) | 20:43 |
| dhellmann | I should probably get their ACLs updated | 20:43 |
| mestery | dhellmann: Ack! | 20:43 |
| dhellmann | mestery: yeah, that should be safe | 20:43 |
| mestery | cool | 20:43 |
| mestery | dhellmann: Is this one ok to merge? ofagent had a 2015.2 version prior to this: https://review.openstack.org/#/c/193129/2 | 20:44 |
| mestery | jenkins passes, but I suspect this is an example of one we want to WIP until we release a new ofagent | 20:44 |
| dhellmann | mestery: https://review.openstack.org/193305 | 20:45 |
| dhellmann | mestery: yeah, let's hold off on that one | 20:45 |
| mestery | cool | 20:45 |
| mestery | dhellmann: The octavia patch merged already: https://review.openstack.org/#/c/193302/ | 21:02 |
| *** openstackgerrit has quit IRC | 22:55 | |
| *** jgriffith has quit IRC | 22:55 | |
| dhellmann | mestery: we can live with it | 22:57 |
| *** dims_ has joined #openstack-relmgr-office | 23:01 | |
| *** dims has quit IRC | 23:04 | |
| *** openstackgerrit has joined #openstack-relmgr-office | 23:06 | |
| *** jgriffith has joined #openstack-relmgr-office | 23:06 | |
| *** dims_ has quit IRC | 23:53 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!