I'm on release duty today. It looks like 2.11.x has 1 pending fix that Benjy is on top of in addition to a Pex upgrade I'd like to get in that fixes a Pex lock file issue. Let me know, or add to the milestone if you have others that should block release. I'll probably start that this afternoon assuming Benjy's fix is in and cherry-picked. I'll probably spin up the 2.12.0.dev2 release at that same time unless anyone yells.

Yes, thanks. I can try this morning to remove some of the deprecation is that we keep punting on for dev2.

Ok, great. I had 4 dep PRs on my plate when I looked at it this morning after a glance. I'll assume you're doing those then instead until I hear otherwise in which case I plan on doing them.

The subsystem API deprecation must be punted on again, as tool chain has not yet released its new plug-in. I think the others can be handled

That's an interesting case. For anyone except TC we would just be breaking them (unknowingly but with fair warning by Pants own terms). I think this is a smell. It seems to say our deprecation policy and aggressiveness is too hard for real users.

true, @bitter-ability-32190 it was a big deprecation to rewrite your plug-in to use a whole new API. What do you think about delaying until 13?

I'm not sure I agree about us "just breaking them" I expect almost all plugins are in-repo, which means people can fix when they upgrade Pants versions. TC is the odd duck who publishes a plugin. That said, yeah punting costs us nothing 🙂

@witty-crayon-22786 the changes in target.py a bit complex and I suspect you already had ideas for what the new interface should look like. Would you have a couple minutes to please make those changes?

I expect almost all plugins are in-repo

Yeah, that's almost certainly true today. That does not seem like the road to a vibrant ecosystem long term though.

Yeah, perfect time now to tweak the process, if needed (our plugin at work are not in-repo btw)

At some point we need an API folks can rely on to get that ecosystem. We need not be Linux and never break user space. But we could be a heck of a lot closer to that end of the spectrum than we are. Today we aren't forced into the general thorny issue of API design that imposes on you.

i’ve been tracking (what i think are) issues which block trying to make a push for plugin API stability: https://github.com/pantsbuild/pants/projects/19 would be interested in people’s thoughts on the list, but until we’ve fixed some of the fundamental stuff (particularly around the rule graph), i think we’d be hamstringing ourselves to declare stability.

He already has the cherry pick up for the fix

I have 0 opinions on the matter; so I'll hold back on 2.11.x unless I get a clear positive go, go, go!

would be interested in people’s thoughts on the list

Might be worth a different 🧵

yeeep. sorry.

My cherrypick is in on 2.11.x, thanks

Sounds good. Per Stu above I'll be not doing a 2.11.x today as it turns out.

@enough-analyst-54434: i was suggesting releasing without it, so the fact that it has landed is icing. sorry that that was confusing

Gotcha. Probably confusing because I didn't read well and I'm also waiting on myself with the Pex upgrade & CP. I will cut a 2.11.x rc then once my self wait terminates.

I think that 2.12 should be fixed already https://github.com/pantsbuild/pants/pull/14922. We only did not cherry pick to 2.11. Sorry that I did not finish with the other deprecation beyond the lock file generator. The presentation today took more time than expected. Feel free to bump those deprecation if you would like, they all seem pretty involved