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

Pants

its a classfile or a resource and we have a product for those

need to finish my first cup of coffee before getting into this.

not to say the products couldn't be improved

hm. there will also not be a need for annotation processors to (re)write out a file mapping classes to resources

correct - that protocol was only invented to support batch, with 1 target at a time, not needed

<@U07LYMLQ1>: I was out yesterday, but saw your sample - I've sent you <https://github.com/viktortnk/pantsbuild-sample/pull/1> to switch to <http://pantsbuild.github.io/setup/>

<http://pantsbuild.github.io/setup/> needs more polish, but works fine now, I've used it for other sample repos <https://github.com/pantsbuild/issues-1751> and <https://github.com/pantsbuild/issues-1779>

<@U04S45AHA>: in Whisk's repo I got 
```
PANTS_PACKAGES=(
  pantsbuild.pants==${PANTS_VERSION},
  pantsbuild.pants.contrib.scrooge==${PANTS_VERSION}
)
```

what is the best way to support contrib modules

like you are for now, like I said, the setup repo needs work

<@U04S45AHA>: code in PR doesn’t bootstrap Pants for me

```
./pants: line 65: /Users/viktortnk/.pants.d/bootstrap/pantsbuild.pants==0.0.35-latest/bin/pants: No such file or directory
./pants: line 65: exec: /Users/viktortnk/.pants.d/bootstrap/pantsbuild.pants==0.0.35-latest/bin/pants: cannot execute: No such file or directory
```

Not sure, I just tried again via `rm -rf ~/.pants.d/bootstrap/ &amp;&amp; ./pants`, which worked.  I'm on linux though, so there may be OSX issues

OK - well I'd reject that pull request then!

No worries, I'll find some time to get a ci setup for that script and robustify - its a needed thing and it ought to just work.

<@U04S45AHA>: the only thing preventing <https://rbcommons.com/s/twitter/r/2453/> from living in contrib is that it needs to install a GroupTask member. worth breaking the 4th wall to move it there?

(to be more specific: having its `register.py` reflect to find the existing task and add the member at the head)

can anyone think of a case where the `test.junit` task runs, but `resources_by_target` are never requested? (apparently it's possible, and i see the fix, just wondering how to test the fix)

<https://rbcommons.com/s/twitter/r/2489/>