Ergonomic, hermetic, user-friendly software build system for Python, Go, Java, Kotlin, Protobuf, Scala, Shell. Pants lets you fearlessly scale up your codebase!

Pants

<@U04S45AHA>, <@U051221NF>: are y'all on `1.8.0` ?

It's unlikely we'd be on stable for any stable period of time.

i'm going to run through the motions of 1.8.1 for <https://github.com/pantsbuild/pants/pull/6278> in particular

but... yea. twitter is off of 1.8.x already so.

will just need to make extra noise about the rcs and see whether the community chimes in.

<@U6ZRNH0TC>: so i know that the cherry-pick of <https://github.com/pantsbuild/pants/pull/5521> requires a bit of munging. i can do the cherry-pick itself if you haven't started editing the `1.9.x` branch

i haven’t and won’t be able to get to it before 2 so go for it i was just about to message you

or if you don’t get to it then i can easily pick it up

ok, i'll do the cherry-picks that i can and get that all green and then get back to you

ok, cherry-picked the two that i could and removed their labels.

1.9.x running here: <https://travis-ci.org/pantsbuild/pants/builds/421743052>

1.8.x on the other hand looks more problematic: <https://travis-ci.org/pantsbuild/pants/builds/421713148>

back, i can check out the failures on 1.8.x

<@U6ZRNH0TC>: don't worry about it... that one is on me.

<@U6ZRNH0TC>: maybe you can cut the next 1.9.x release candidate? or get the path globs PR landed, and cherry pick that before doing the candidate?

i merged <https://github.com/pantsbuild/pants/pull/6278> just now, and i am near to a solution for <https://github.com/pantsbuild/pants/issues/6394> (i think), so i'm going to fix that before making 1.9.0rc2

not from py3 changes as far as i can tell, john's comments are pretty clear. i don't know what precisely triggered this regression but once i fix it i might, and if it's py3 i'll let you know for sure

Okay, thanks! I’m trying to land <https://github.com/pantsbuild/pants/pull/6372> atm, will significantly improve our ability to detect Py3 regressions

this is awesome! i left a comment that you can 100% ignore

Yay thanks for the review! Also I don’t have write access to merge, so will need help with that once we feel ready

Alternatively to fixing in a followup PR, would you be comfortable if I push that cosmetic change but we ignore CI because it’s purely cosmetic?

We can also re-run CI if preferred. This isn’t immediately blocking anything

the way i have always seen <@U06A03HV1> handle this is to make cosmetic changes in a followup if necessary instead of going through CI again

(my comment was about a cosmetic change on this PR, for context)

imo don't worry about this comment i just wanted to bring it up at some point

That’s fine by me. I am planning on fixing `test_build_graph` non-determinism this week, so can include in that

great! although I would probably do any non-code (style) change in a separate PR from any real changes

by the way, if i had to guess, i would think the march 17 commit from benjy which makes everything a list in the `fingerprint()` method would be what caused the `--binaries-path-by-id` regression, that and the (assumed) fact that nobody overrides that in `pants.ini` caused it to take this long

python 2 json.dumps flatly lies and says it coerces keys to strings

so i bring that up just because i doubt py 3 edits caused this

<@U6ZRNH0TC>: has someone actually requested <https://github.com/pantsbuild/pants/issues/6394> ...?

no but it's a regression and if anyone has overrides for that var in pants.ini they will start seeing errors