Another approach is for your org to build the wheel and host it in an "internal artifactory". This is how Pants does our release - we build our wheels in Centos7 for maximum compatibility, and then fetch those prebuilt wheels for the release

https://pantsbuild.slack.com/archives/C046T6T9U/p1618427758021400?thread_ts=1618427717.021300&channel=C046T6T9U&message_ts=1618427758.021400 I'm still surprised by this. I would expect a Big Sur wheel to work with Catalina. Maybe unzip and look at the .deps folder, see tool tip at https://www.pantsbuild.org/docs/python-package-goal#approach-3-no-entry-point

Hi, I suddenly experience errors with

pants package ::

:

Engine traceback:
  in select
  in pants.core.goals.package.package_asset
  in pants.backend.python.goals.package_pex_binary.package_pex_binary (apps/semantic-tagging-manager-service)
  in pants.backend.python.util_rules.pex.two_step_create_pex
  in pants.backend.python.util_rules.pex.create_pex (__requirements.pex)
  in pants.engine.process.fallible_to_exec_result_or_raise
Traceback (most recent call last):
  File "/home/cue/.cache/pants/setup/bootstrap-Linux-x86_64/2.2.0_py38/lib/python3.8/site-packages/pants/engine/process.py", line 255, in fallible_to_exec_result_or_raise
    raise ProcessExecutionFailure(
pants.engine.process.ProcessExecutionFailure: Process 'Resolving 4 requirements: fastapi==0.63.0, hypercorn==0.11.2, pydantic==1.7.3; python_version >= "3.6", requests==2.25.1' failed with exit code -9.
stdout:

stderr:

I haven’t changed any setup or requirements. Any ideas why this is suddenly happening?

https://twitter.com/rust_foundation/status/1382726294136266754 Pants Build retweeted: "Rust is emphasizing collaboration because we believe that software engineering is primarily a social process rathe… https://twitter.com/i/web/status/1382726294136266754 Twitter

https://twitter.com/ThePSF/status/1382766531747061765 Pants Build retweeted: Thank you Google for becoming the first ever Visionary Sponsor of @ThePSF. The financial support and in-kind donati… https://twitter.com/i/web/status/1382766531747061765 Twitter We are happy to announce we're increasing our support for the Python Software Foundation. This includes donating $3… https://twitter.com/i/web/status/1359933923355226113 Twitter

It’s been a while since I last looked into pants. What is the current status of support for compiling extension modules w/ numpy distutils? Alternatively, can pants support arbitrary subprojects which use

setup.py

rather than BUILD files? I could probably build a wheel of the subproject and have pants treat it is an external dependency, but would appreciate if pants could automate this for me.

Happy Monday everyone! Excited to announce some upcoming conference talks of interest to the Pants community: On May 3rd at 6 AM PT I'll be speaking at PyCon Israel: - Topic: "Python monorepos: what, why and how" - Abstract: https://cfp.pycon.org.il/conference2021/talk/VKPEP3/ - Registration (free this year): https://pycon.org.il/2021/ On May 14th at 9:15 AM PT I will be speaking at PyCon US: - Topic: "Creating extensible workflows with off-label use of Python" (this is basically about lessons learned from building the Pants execution engine) - Abstract: https://us.pycon.org/2021/schedule/presentation/68/ On that same day, at 1 PM PT, @hundreds-father-404 will also be speaking at PyCon US, in Spanish: - Topic: "Cuándo usar extensiones nativas en Rust: rendimiento accesible y seguro" (also basically about lessons learned from the Pants execution engine) - Abstract: https://us.pycon.org/2021/schedule/presentation/17/ - Registration: https://us.pycon.org/2021/registration/information/ Not free, but your fees help fund the PSF, and you may be able to have your company reimburse you. We're really excited to have not one but two Pants-related talks at such a prestigious Python conference! (Our company, Toolchain Labs, will also have a virtual booth at the conference, where we'll presumably be talking about Pants and its uses for Python repos quite a bit.) Since these conferences are virtual, this is a great opportunity to participate without the cost and hassle of travel. So please consider registering, telling your friends, and promoting on social media. Cheers!

https://twitter.com/pantsbuild/status/1384246394781724674 If you contribute 5 hrs/month to Pants Build or other Python open source projects that are available for free, you're eligible for @ThePSF Contributing Member status. It's free, takes less than 15 mins to apply, and comes with voting rights. Helps support, shape & lead community. Twitter If you have a few hours a month to spare to support the Python community, consider becoming a PSF Managing/Contribu… https://twitter.com/i/web/status/1383696303352213506 Twitter

https://twitter.com/artem_pyanykh/status/1381672305466703881 Pants Build retweeted: Just published another video on YouTube: This time it's about @rustlang, async, tokio, fut… https://twitter.com/i/web/status/1381672305466703881 Twitter

Which talks/videos/post about Pex do you especially recommend?

https://twitter.com/pantsbuild/status/1384330136955219969 Pants 2.5.0.dev2 is released! https://pypi.org/project/pantsbuild.pants/2.5.0.dev2/ This is the first release to support running Pants with Python 3.9. To use Py39, please update your ./pants script by using the version at https://t.co/gDmsb4FBS3. Twitter

https://twitter.com/pantsbuild/status/1384336023883190278 Meet team members in May, sharing what we learned and re-conceptualized while creating Pants v2 engine. @benjy will be speaking about monorepos at @pyconil and about off-label use of Python at @pycon. @EArellanoAZ will be speaking about Rust at Pycon. Intrigued? Details are next. Twitter

https://twitter.com/pantsbuild/status/1384801475738492930 PythonNotPython Twitter Twitter

I recently updated to Pants 2.4.0 in our project, which uses a constraints file which we update pretty regularly. Previously we were using Pants 2.2.0. With 2.2.0, we were pulling in

setuptools

50.3.2, but with 2.4.0, we're now pulling in

setuptools

53.1.0. However, we're also using the

python_awslambda

packaging type in our codebase, which is based on

lambdex

, which apparently has a

setuptools

constraint of

<50.4,>=50.3.0

, which is unsatisfiable with

setuptools==53.1.0

. I can pin

setuptools

to the old 50.3.2 and still build lambdas, but this was rather unexpected. I see a few old open PRs on the lambdex repository that seem possibly related (such as https://github.com/wickman/lambdex/pull/8). Is this a known issue? Is there a better way to handle such conflicts than what I'm doing? Thanks in advance.

https://twitter.com/pantsbuild/status/1384920017033125888 Monorepos and performance: highlights from Pants Build maintainer @benjy's conversation with @semaphoreci podcast https://blog.pantsbuild.org/monorepos-and-performance/ Twitter

What would people think of Pants adding an automated mechanism to validate that artifacts built with

./pants package

are valid? For example, something like this:

pex_binary(name="main", entry_point="main.py", validate_argv=["--help"])

Then Pants would run

dist/../main.pex --help

for you as part of

./pants package

https://twitter.com/pyblogsal/status/1382290422970654720 Pants Build retweeted: Python 3.10 🐍 will try to offer you suggestions from existing attributes on AttributeErrors if you mistype the name… https://twitter.com/i/web/status/1382290422970654720 Twitter

https://twitter.com/pyblogsal/status/1384881060396285955 Pants Build retweeted: We just gave those IndentationErrors some extra love. In Python3.10 🐍 you will be able to know what kind of block w… https://twitter.com/i/web/status/1384881060396285955 Twitter

Is there a good way to see orphaned files in pants? basically files that only end up in lib and test targets but are never put into a binary

After upgrading to Pants 2.4.0, I'm seeing some odd failures in our CI environment:

16:26:31.31 [WARN] Completed: lint - isort failed (exit code 2).
.cache/pex_root/venvs/d8105fd40948579fedc63bb8ddd363d251b7db7a/5dc7bde7966ce85a9f3820e78694dc0e8d63768d/pex: line 7: syntax error near unexpected token `('
.cache/pex_root/venvs/d8105fd40948579fedc63bb8ddd363d251b7db7a/5dc7bde7966ce85a9f3820e78694dc0e8d63768d/pex: line 7: `    venv_dir = os.path.abspath(os.path.dirname(__file__))'

I'm working on getting access to the build nodes to dig in further, but it looks like after the failing pants run that

pex_root

directory isn't present anymore, making it tricky to debug. Any ideas on how I might be able to debug this? Thanks 🙇

(this doesn't happen locally on my workstation, so it appears to be some kind of environmental issue, but any pointers would be appreciated)

https://twitter.com/pantsbuild/status/1386525428467179524 👏👏👏👏👏👏 Twitter if you ever want to do things concurrently, here's how (a thread) Twitter

👋 looking at our CI I noticed Pants (2.1.1) is taking a long time resolving the 3rd party constraints (~3min) and can see many lines like this one:

18:18:28.00 [INFO] Long running tasks:
  110.81s 	Resolving 3rdparty/python/constraints.txt

I could not figure out why, do you have any suggestion how to investigate/solve this?

Hi Team: Is it possible to disable dependency inference per build target?

Does pants support visibility for libraries? For example:

visibility = ["//visibility:public"]

I don't see any reference to that pants documentation.

I am interested in trying out Pants to run tests for an existing machine learning project in which third-party dependencies are specified via a conda env.yml file that is used to create a conda environment. To use Pants, would I need to re-specify the dependencies in requirements.txt format, or could I somehow get Pants to make use of the existing conda environment?

Is there currently a solution to lint/autoformat BUILD files? My project has gotten into a space where we mix 2 and 4 spaces as indents and would love to fix to be 2 spaces, but if there is a tool that only does 4 I can be flexible.

Hi. Just getting started with Pants and need to write some custom rules. I'm a little confused about how to chain rules together. For example, suppose my rule consume target

my_target

and generates some new files. As I understand it, I can return a custom`Goal` subclass that contains some information about those files but that does not generate new targets. Now suppose I want to make sure my generated code is safe via

bandit

. Since I don't have a new target and

bandit

's

@rule

doesn't take a

MyCustomGoal

how do I tell

bandit

that it depends on (or should check) these custom files?