oh, it’s being smeared downward… got it. @nice-florist-55958: do you commonly see the UI smearing like that? we attempt to disable it in locations without a real TTY, but it looks like we failed in this case

@witty-crayon-22786 So far only when trying to download a binary on the internet like this. This is using PyCharm Professional’s terminal on Linux. Didn’t think to mention it, but it goes wild for a bit spamming the download attempt message and flickering. Also doesn’t gracefully exit after a keyboard interrupt until after several minutes go by. Btw at the office now and was able to successfully deploy pants on production CI servers! Will experiment with some workflows (including multiplatform) and see how it goes.

ah, got it. sorry for the trouble… you can disable the UI with

--no-dynamic-ui

, but i’ll open an issue about that. EDIT: https://github.com/pantsbuild/pants/issues/13595

Is it possible to build docker images across architectures? We are using

docker buildx

to build both x86 and arm containers and wondering if we can get this working with pants.

Getting NoSourceRootError when trying to get dependees, even though Pants recognizes the correct roots? See screenshot.

Hi team, I am trying to run this command using v1.30.5

./pants --changed-parent=origin/main list

in the directory I have defined my sentinal/root files but I am getting an error

Globs may not traverse outside of the buildroot: "../../BUILD"

as there is indeed a BUILD file in that location, but in the interim I would like for my pants environment to ignore any target outside of my current directory and only ack child targets. Is there some toml file tweak that I can do to let pants ignore anything outside of my current directory ? I tried adding

../../BUILD*

to

pants_ignore

but no luck

While working to move our python build system to pants I have encountered an error I can't make sense of - Deep down a dependency tree, my code is erroring on

os.uname()

–

def _syscmd_uname(option, default=''):
    
        """ Interface to the system's uname command.
        """
        if sys.platform in ('dos', 'win32', 'win16'):
            # XXX Others too ?
            return default
        try:
            f = os.popen('uname %s 2> %s' % (option, DEV_NULL))
        except (AttributeError, OSError):
            return default
>       output = f.read().strip()
E       ValueError: I/O operation on closed file.

/Users/ryan.king/.asdf/installs/python/3.6.15/lib/python3.6/platform.py:791: ValueError

This code works fine outside pants and I don't know what about it would fail in the pants env.

What is the best way to change the location pants bootstraps itself in and uses as a write/cache directory from ${HOME}/.cache so that we can point it to a persistent install location? I assume this is critical to get the most benefit out of Pants (right now our build servers would just reinstall it from scratch each time).

Instructions can be found here: https://www.pantsbuild.org/docs/troubleshooting#how-to-change-your-cache-directory

Thanks vm!

Having a couple issues getting started with builds. Have a test project called casper w/

./src/casper/<pkg, mod, etc.>

structure and

"src/"

pattern for root. When using

./pants run <target>

located in

./src/casper/pkg/subpkg

, I get an module not found error on

import casper.pkg.subpkg

I would have assumed Pants would place

./src

in the

PYTHONPATH

(or perhaps install the package in the hermetic env to make it self-importable)? I don't know if it makes a difference, but the import was in

__init__.py

of

pkg

. The second issue is that when I make any changes to the offending file, it seems Pants does not pick those changes up -- build and run goals keep using the same depended on file.

When trying to upgrade from

2.6.1

to

2.8.0rc5

, on the packaging step, the Linux CI fails with an error message. It looks like Pex now requires all dependencies, transitively, to be present as wheels (

--only-binary :all:

argument to the

pex

command). It works fine to produce necessary

.pex

packages with

2.6.1

atm even though some transitive dependencies are available only as

sdist

on the PyPI. I know Pex wouldn’t get the

sdist

for multiplatform packages, but the package in question is only for a single platform (

linux_x86_64-cp-38-cp38

). @enough-analyst-54434 is this intended?

Hi, I wonder whether

poetry_requirements

supports wildcard specification. In migrating a project using pants, I took the existing pyproject.toml as is, which contains, e.g.

pillow = "*"

in the declarations. After generating the user lockfile using

poetry export

and configuring the constraints file in the

pants.toml

, I received an error saying that

pillow==*

can't be satisfied while building. I've verified that poetry has correctly exported the resolved version for that package in constraints.txt. I hope I followed the right steps when using

poetry_requirements

in pants. It's not a blocker for me though as a wildcard can always be rewritten to

>=0.0.0

or some known version as the lower bound in pyproject.toml. But it would be nice if pants could have it supported.

Is there an option to change the pex build hashbang from #!usr/bin/env pyton to a pre-specified one (or set to the Python version resolved by Pants when running the goal (it is on AFS and so accessible outside of Pants' env)).

Getting permission denied errors when trying to use black, pylint, etc. (img 1). I was able to do this earlier on in the experimentation phase, so not sure what is happening here now. Reinstall did not fix the issue. The file clearly has executable permissions, so not sure (img 2). Tried unzipping it too because it looked like a pex archive, but unzip complains it isn't (img 3).

Is it possible to declare 3rd party dependencies in a way analogous to sitecustomize.py? We have several internal lib dependencies not on our PyPI proxy, and we would prefer not to have to explicitly bring them into the repo as that would cause a lot of bloat and be tedious to maintain. In our current setup we make these packages importable in our main virtual environment (and known/indexable by our IDE) by updating sitecustomize.py with a sys.path.append("resolved/path/to/dependency") statement (we have a library that does the abs path resolution of various AFS release links). Is there a canonical way to do that in Pants such that A) Pants will infer the dependency in any build sources B) package the dependency as it does a PyPI requirement? Even if we placed the dependency in ./3rdparty, what distribution format does Pants expect in order to package it or will it just copy whatever you have in the top-level directory (i.e., assume it is already "installed")?

When placing a package with a compiled binary into

3rdpary/pkg/bin.so

, Pants only copies the

*.py

files when running the test goal. So

import pkg

results in

__init__.py

being invoked and then an import error from

from .bin import *

because

bin.so

is not in the sandboxed environment. Confirmed by dropping into pdb and looking at the folder structure in the execution env. Any ideas? Is that expected behavior or?

I'm going through and adding

mypy

to my project. But continuously getting this error:

𐄂 MyPy failed.
10:29:21.85 [ERROR] Completed: Typecheck using MyPy - MyPy failed (exit code 2).
test/python/gcloud/conftest.py: error: Duplicate module named "conftest" (also at "test/python/editor/score/conftest.py")
test/python/gcloud/conftest.py: note: Are you missing an __init__.py? Alternatively, consider using --exclude to avoid checking one of them.
Found 1 error in 1 file (errors prevented further checking)

So now (due to: https://github.com/pantsbuild/pants/issues/13615) I'm adding

overrides={"conftest.py": {"skip_mypy": True}}

to all my

python_test_utils

but it feels like I'm doing something wrong?

Hmm… why is this not compatible? [SOLVED]

You have set interpreter constraints (`CPython<4 OR CPython>=3.6.*`) that are not compatible with those used to generate the lockfile (`CPython<4,>=3.6`). `flake8` determines its interpreter constraints

Dang, ok I see the

OR

now…

For symmetry, it would make sense if the

C1, C2

where presented as

C1 AND C2

to match the

OR

case… that would be clearer 🙂

Is there a page or part of the docs that describe what pre-commit or CI might look like for a "complete" build-and-test of a Pants monorepo? I ask because obviously there's

fmt

,

lint

,

check

,

test

, but what about

validate

? or

update-build-files --check

? Edit: The 2.8 docs have me covered 🎉

How do we tell the file watcher to ignore certain files? My current case is, while addressing https://github.com/pantsbuild/example-django/issues/12 by putting the .sqlite3 files in the cwd (so the repo root), as the migration runs it modifies those files, which causes it to rerun... *.sqlite3 is already gitignored and therefore pants-ignored, but that appears to not be sufficient.

The tricky thing is we probably want to preserve how you wrote them in your pants.toml, e.g. so you can grep it

[python]
interpreter_constraints = [">=3.6,<4", "==2.7.*"]

We use

' OR '.join()

I think grep-ability is less important than conceptual clarity in the error message, imo.

Hi All - I'm doing some work to help evangelize Pants in the Django space and have put together a blog post on how to get started on Pants with a brand new Django app. Would appreciate some love on the social 🙂 • Reddit • Twitter Sorry for the shameless plug!

👋 we recently moved some apps to Pants/PEX, and we noticed a big increase in the container's memory usage (from ~300MB to ~1.5GB). We found out this was caused by linux caching after unzipping the file before running the app. Did anybody else experience the same issue? and if so, how did you solve it?

Hi folks! Don't forget to add your organization to Who Uses Pants https://www.pantsbuild.org/page/who-uses-pants It's fast and easy to do at https://pantsbuild.org/register

Hello. This is not directly related to pants but PEX itself. We are trying to build some pex’s with poetry instead of pants (for reasons I will not get into at the moment). We are having some issues. Created an issue here: https://github.com/pantsbuild/pex/issues/1518 Any insight?

Commented. I think you're just trying too hard! SImpler approach noted there.