Thursday, 2026-04-02

gouthamrin my research on how the release went for today's open infra live, i discovered that skyline was one of our most "active" project teams21:57
gouthamrhttps://review.opendev.org/q/project:%5Eopenstack/skyline-.*+status:merged+after:2025-10-01+before:2026-04-0121:57
gouthamr257 changes merged in Gazpacho21:57
gouthamrinteresting i thought, for something we've kept having governance problems in.. feels like people see value in it and existing maintainers are doing their best voluntarily21:58
fungiwhen it comes to methodology, for one thing you can use merged-after and merged-before to get more accurate (because your query will also show changes that were merged before the after date by comments and the like)22:00
fungibut also i usually measure cycles from the beginning of rc1 thursday in the previous cycle to the beginning of rc1 thursday in the cycle in question22:01
JayFI still am refering to Hibiscus as "next cycle".22:01
JayFIT never feels like it starts until PTG.22:01
gouthamrfungi: ah, useful tip, it removed one change and gave me 256.. but i'll use that for queries 22:02
gouthamrugh, you're right about RC date22:03
gouthamrJayF: haha, true.. all contentious work merges after22:04
JayFLike I've said many times: OpenStack governance is not a good fit for all OpenStack-integrated projects. There are lots of successful projects that don't need the thick layer of governance we apply.22:04
JayFI wonder if gouthamr's findings are a nod in that direction22:04
fungigouthamr: merged-before will also help you not lose changes in your count if it's been a while since the before date22:04
gouthamrack22:05
gouthamri don't know JayF.. new contributors were finding it hard to land changes in Skyline.. it was mostly work the three existing contributors committed.. but, that's changed a lot this cycle.. so all the reach out we've done could have helped... 22:06
gouthamrbut no amount of us goading people can help if they can't/won't do it.. which is why we need the new contributors to become maintainers 22:06
clarkbIsn't skyline part of the governance? If anything I would argue that the friction of playing along with the existing expectations has created more problems contributing than less22:11
JayFclarkb: that's basically what I'm trying to say but you used different words22:12
JayFand had less of a nod to a potential solution :) 22:12
clarkbI think you're saying they do less governance and merge a lot of code. I think I'm saying they don't do what we expect them to so its hard for anyone not them to contribnute code22:13
clarkbbut maybe I've misinterpreted22:13
clarkbfor example: https://opendev.org/openstack/skyline-console/src/branch/master/skyline_console/static22:13
gouthamri wasn't there in the prior skyline discussions and need context.. 22:19
clarkbI'm just pointing out that if someone fomr openstack were to see ^ they would panic (and indeed I believe this has happened before with zigo maybe?)22:21
gouthamryes, i vaguely recall a ML post22:21
clarkbbut things are improving and I think that is a result of trying to make things moer similar so that more peopel can dive in and make sense of it22:21
gouthamr++ agreed... i think we can break the communication barrier as well if more people got involved and could talk to each other through code commits.. like show me how you'd do it differently.. 22:24
gouthamrdon't commit minified js bundles, we've had this long history of packaging problems that could be done better by this other method here.. 22:24
gouthamrfrom what i recall they had a very short "time to market" goal, but then their company folded up their efforts before they could build a community around it :/22:25
clarkbits also not a great idea to commit binary data in git if you can avoid it. I think nova has an opnwrt vm image in its history that will never go away now for example22:25
clarkband that is a good point to make. Its not like the other projcets were perfect. They learned these lessons the hard way and it can be furstating to watch others repeat those mistakes rather than take the easy road and learn from them22:26
clarkbbut sometimes personal experience is the best learning method :)22:27
gouthamr:D 22:27
JayFclarkb: yeah, I guess you're right that is an issue regardless22:53

Generated by irclog2html.py 4.1.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!