<https://github.com/pantsbuild/pants/issues/7426> ...
# development
a
https://github.com/pantsbuild/pants/issues/7426 it's possible to modify the v1 engine to fix the caching bug endemic to the current c++/python_dist() support, but just migrating to v2 would also work. i'm probably going to prioritize the migration of the native backend to v2 in my free time because why not
w
i know that's a rhetorical question, buuut... the answer is probably: because we don't have caching or a target API for v2 yet
a
oh right ugh
process execution caching
i don't super care about the target api i can develop that as i go
w
that is probably true.
i started sketching a bit over the last few evenings.
a
it's just a little frustrating because we can't really land a v2 native ruleset until process execution caching works
which is fine because i've wanted to do that recently anyway
sketching sounds very good
i'm going to dump some issues in the native backend project: https://github.com/pantsbuild/pants/projects/11
oh ha we already have https://github.com/pantsbuild/pants/issues/7136 i'll update it
w
i'll try to post the target API thoughts tonight. last night it just resulted in some cleanup.
a
that's great
i just posted the todo for v2 porting to the native backend in 7136 (very simple)
i'm already diving into the process execution stuff again when debugging remoting flakiness so focusing on local process execution caching is a good convergence