| TheJulia | So, I'm just going to note that the log has many different points, but this is also the TC room and it might help project focused individual contributors to get more direct and constructive feedback as it relates to specific projects, their decisions, and even issues or disconnect which exist. That doesn't mean there is not value to some of the items in the log, just we can't expect people to read minds or channel logs they might not | 00:17 |
|---|---|---|
| TheJulia | focus on. | 00:17 |
| mnaser | mikal: Yeah that's why we wrote an OpenStack database exporter which is literally a mistake in the fact that it exists and I agree things like queue depth or VMs waiting scheduling it whatever kinda metrics still not being exposed via prom is tough.. esp if we're proactively avoiding it | 00:33 |
| *** elodille1 is now known as elodilles | 12:06 | |
| fungi | mikal: not that i necessarily disagree with your points, but proxmox started development 5 years before openstack was formed, and had a 1.0 release 2 years before openstack's first release | 13:54 |
| fungi | (there are probably better examples) | 13:56 |
| fungi | the cpython community now has guidelines for maintainers enforcing their ai contribution policies: https://discuss.python.org/t/updated-triage-guidelines-and-process-for-our-ai-policy/109717 | 18:21 |
| cardoe | Well I missed some fun here. | 18:37 |
| cardoe | mikal: I'd -1 pretty hard if we made nova just work with KVM. Today it works with Ironic as a backend and that's what I use. | 18:37 |
| cardoe | There's recently been some work in oslo.middleware for supporting otel. | 18:37 |
| frickler | fungi: interesting, though I'm not sure how to judge whether responses are actually being made by people using their own words. but we could consider defining some similar hashtags for statistical purposes, even without support for comments from a bot | 18:41 |
| fungi | frickler: right, i was contemplating that model (using hashtags specifically) in light of some of the cut-and-paste code review responses or templates clarkb proposed | 18:41 |
| fungi | though my reason for linking in here is more about the fact that they set guidelines for how their maintainers are recommended to respond to some more common llm-generated change related problems | 18:42 |
| fungi | not as much about the mechanism they implemented for that | 18:43 |
| fungi | however, at a more meta level, it does come across as a sort of botfight too ;) | 18:46 |
| fungi | i don't want to see our gerrit turning into moltbook | 18:47 |
| JayF | I'll note that I have an AI policy "interpretation" draft in the resolutions; that'd be a good place to discuss things like this if we think we need additional technical clarification. | 18:49 |
| JayF | From my POV, I just want the TC to set a standard. I have preferences but none of them are as strong as the preference to not have "AI Discourse"(tm) in every other code review I do :) | 18:49 |
| mikal | I'm with JayF. I don't mind there being a policy. I mind an undocumented policy being enforced. | 18:51 |
| JayF | I would argue we don't have an undocumented policy, we have a policy that has room for interpretation, which leads to unnecessary friction. | 18:54 |
| * JayF thinks most of that friction was created by AI megacorps using Co-Authored-By to drop ads in to everyone's commit messages, but that's my $0.02 only and that rounds to $0.00 in the US now :D | 18:54 | |
| fungi | there is also sometimes a fine line between clear enough guidance to let people agree on a compromise and move past the issue without constantly relitigating, vs drowning in bureaucratic red tape and arguing over every little detail | 18:55 |
| fungi | i feel like our community is already drowning in bureaucracy in many ways, so always worry about each new policy addition making that situation worse | 18:57 |
| JayF | Like I said when I filed it: my main goal in going the route of a resolution was to avoid the appearance of modifying project-team-guide to "create" policy. If the TC agrees that the interpretation is obvious and just needs documentation, I'm fine to do that :) | 18:59 |
| gouthamr | JayF: thanks for proposing it. i wrote it down as one of the PTG action items.. | 19:02 |
| gouthamr | we can help refining it and getting it in | 19:03 |
| gouthamr | link: https://review.opendev.org/c/openstack/governance/+/1012080 | 19:03 |
| TheJulia | fungi: That is, indeed, a very fine line. Unfortunately a root cause is people not finding the same foundational basis upon which a line is drawn as well so its at different places. Goes back to the building mind share and consensus aspect and sometimes requires listening as much as talking by sharing mutual context. | 20:55 |
| enricokern | o/ | 21:34 |
| enricokern | so hope everyone calmed down after yesterday nights discussion :P | 21:35 |
| gouthamr | i don't think there were tempers | 21:46 |
| fungi | there's no crying in openstack | 22:42 |
| TheJulia | Passion yes, crying... highly unlikely. | 23:04 |
Generated by irclog2html.py 4.1.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!