I think if you pass

--level=debug

(or

-ldebug

) to pants it passes that through to pex? Which version of Pants are you using?

1.25.0 I see some extra prints, nothing releated to pex

the pants-v1 python backend does not subprocess pex: it uses it as a library

pants-v2 python backend does subprocess pex, but is still nascent

There's nothing on github regrading v2 aswell ?

Not yet, but see https://pants.readme.io/docs for a rewrite of the docs specifically for V2. (Still a WIP)

I see.. thanks:) btw- just wondering if in v2 using to subprocess pex, I see pex tool doesn't have "support" to add files into the pex, only directories or packages with setup.py. but Pants does adding just like that files, how would v2 will manage to copy that behavior?

Hi all, I've created #gettingstartedv2 for providing help and discussion around getting started with using Pants with the v2 engine (which currently means Python, other language support coming soon). Feel free to @ me or @hundreds-father-404 there with questions!

Couple reasons you might want to try out V2: * precise file arguments; if you say

./pants test project/test_foo.py

, we will only test that file and nothing else * smarter caching * less verbose output * avoids irrelevant work that V1 does - should be faster * works more ergonomically with custom target types * better support for Python interpreter constraints; if one test uses Py2 and another uses Py3, you can still run them in the same Pants run, unlike with V1

Proposal for simplifying macros in BUILD files: https://docs.google.com/document/d/1mtkZGB6nZochxbbKx1zUzoJ1H3Jh_j4lEso586UPreE/edit - comments very welcome! (Currently accessible by pants-devel@, but drop me a message with your email address if you don’t have permission and want it)

hi all, just curious about the remote cache feature, which I’m setting up. does it only cache the final target, and not intermediary outputs? for example, in the below output, do the the interpreter or sources never make it to the remote cache?

17:34:25 00:01   [pyprep]
17:34:25 00:01     [interpreter]
17:34:26 00:02       [cache]
                   No cached artifacts for 6 targets.
                   Invalidated 6 targets.
17:34:26 00:02     [build-local-dists]
17:34:26 00:02     [requirements]
17:34:27 00:03       [cache]
                   No cached artifacts for 10 targets.
                   Invalidated 10 targets.
17:34:35 00:11     [sources]
17:34:35 00:11       [cache]
                   No cached artifacts for 6 targets.
                   Invalidated 6 targets.
17:34:36 00:12   [binary]
17:34:36 00:12     [binary-jvm-prep-command]
17:34:36 00:12       [jvm_prep_command]
17:34:36 00:12     [binary-prep-command]
17:34:36 00:12     [py]
17:34:37 00:13       [cache].
                   Using cached artifacts for 1 target.
                   created pex dist/load_tester.pex
17:34:48 00:24     [py-wheels]
17:34:48 00:24     [jvm]
17:34:48 00:24     [dup]
               Waiting for background workers to finish.
17:34:48 00:24   [complete]
               SUCCESS

good morning folks! anyone of any plans or appetite (or existing solutions) around building a

docker_image

build target?

Hello! We sent an email to pants-devel that I’m copying here to ensure everyone has a chance to weigh in. We’re proposing to deprecate the below in 1.27.0.dev3 and to remove in 1.29.0.dev0 given their low usage: - pants.backend.codegen.antlr.java - pants.backend.codegen.antlr.python - pants.backend.codegen.jaxb - pants.backend.codegen.ragel.java - pants.backend.codegen.wire.java - pants.contrib.avro - pants.contrib.confluence - pants.contrib.cpp - pants.contrib.errorprone - pants.contrib.findbugs - pants.contrib.googlejavaformat - pants.contrib.jax_ws - pants.contrib.scalajs - pants.contrib.thrifty Do any of you depend on any of these? If so, let us know and we’ll gladly remove them from the list.

Hi all, I’m new to Pants and I want to set my own source root patterns, it works with something like

./pants --source-source-root-patterns='["srctest/*"]' binary srctest/python/app/hello

But I cannot set this options in

pants.toml

,

[source]
source_root_patterns = [
    "srctest/*"
]

the code above doesn’t work, does anyone know how to set it correctly ?

Hi all, another question, I want to build an python_binary which has tensorflow==2.1.0 as dependency, and my BUILD file is like,

python_binary(
  name='demo',
  dependencies=[
    '3rdparty/python:tensorflow',
  ],
  platforms=[
    'current',
  ],
  source='main.py'
)

My

current

platforms is macos, but I want to build a pex file also works on linux (ubuntu), I tried several possible values for platform but None of them work Any idea about setting a proper platform value here ?

Hi all, when I build a python binary, it need lots of dependencies and use

/tmp

to do the job, but I don’t have enough space for

/tmp

, which option can point it to another directory ?

hi folks, when moving from 1.25.x to 1.26.x it seems as if

pants test

my

[python-repos] indexes;

settings as it cant seem to find a privately published dependency i cannot find anything related to this in the release notes. Any advice?

https://github.com/thebjorn/pydeps <= visualize python module dependencies. kinda cool

hey folks quick v1 question. If i wanted to take a task registered as its own goal i.e.:

TaskRegistrar(name="build-my-artifact", action=BuildArtifact).install()

and then also add it to a registered goal

Goal.register(name="build")
TaskRegistrar(name="build-my-artifact", action=BuildArtifact).install("build")

would that be possible

never mind just answereed my own question

thanks folks 🙂

(asking those kinds of questions in this channel is actually super useful, even if they seem obvious, because if it's not clear to you, it's probably not clear to more than one other person too! so answering your own question is still a great contribution ^_^ just imho!)

It’d be nice if the black lint rules swallowed positive results, at least in CI. I don’t think there’s a lot of value in seeing > 1000 lines of `

All done! :sparkles: :cake: :sparkles:
N files would be left unchanged.

Maybe the lint console rule could swallow output on positive results by default?

Where are you running this? We default to

--no-fmt-per-target-caching

so that you run everything in one single run and never have more than 3 lines of output

can see it in the lint shard: https://travis-ci.com/github/pantsbuild/pants/jobs/319264808

yeah

I think it’s only an issue for me because I haven’t worked out the git hook env problem I have

Ah, that’s because we use

--lint-per-target-caching

there because we have remote caching so it results in much better performance. A followup project I have is to get the best of both

--no-lint-per-target-caching

(generally faster perf due to the very high overhead of linters + less verbose output)and

--lint-per-target-caching

(fine-grained caching)