Pants #general

late-breakfast-42820

10/13/2025, 9:51 PM

I did kill the pantsd process manually as I couldn't stop it any other way when it was running a server in claude that I shut down without safely exiting

late-breakfast-42820

10/13/2025, 9:59 PM

false alarm! was some docker files from a db lol

😅 1

happy-family-315

10/14/2025, 7:12 AM

Hi, I got weird import behaviour from my pex: I created a multi stage docker image (deb + src) for my pants-built pex file. Everything is working except for an import bug, where my code tries to import geo_pandas which imports typing module from python. However instead it tries to import typing module from open cv. After that I got an exception because I do not have all dependencies installed which are required for open cv. This is also the reason I found out about the error. I'll give you the error log and the docker file. Hope someone can explain.

Copy code

2025-10-10T06:11:32.122Z
File "/usr/lib/python3.12/multiprocessing/process.py", line 314, in _bootstrap
2025-10-10T06:11:32.122Z
self.run()
2025-10-10T06:11:32.122Z
File "/usr/lib/python3.12/multiprocessing/process.py", line 108, in run
2025-10-10T06:11:32.122Z
self._target(*self._args, **self._kwargs)
2025-10-10T06:11:32.122Z
File "/usr/lib/python3.12/concurrent/futures/process.py", line 251, in _process_worker
2025-10-10T06:11:32.122Z
call_item = call_queue.get(block=True)
2025-10-10T06:11:32.122Z
^^^^^^^^^^^^^^^^^^^^^^^^^^
2025-10-10T06:11:32.122Z
File "/usr/lib/python3.12/multiprocessing/queues.py", line 122, in get
2025-10-10T06:11:32.122Z
return _ForkingPickler.loads(res)
2025-10-10T06:11:32.122Z
^^^^^^^^^^^^^^^^^^^^^^^^^^
2025-10-10T06:11:32.122Z
File "/workspaces/job/update_layer/lib/python3.12/site-packages/geofileops/__init__.py", line 10, in <module>
2025-10-10T06:11:32.122Z
import geopandas._compat as gpd_compat
2025-10-10T06:11:32.122Z
File "/workspaces/job/update_layer/lib/python3.12/site-packages/geopandas/__init__.py", line 3, in <module>
2025-10-10T06:11:32.122Z
from geopandas.geoseries import GeoSeries
2025-10-10T06:11:32.122Z
File "/workspaces/job/update_layer/lib/python3.12/site-packages/geopandas/geoseries.py", line 3, in <module>
2025-10-10T06:11:32.122Z
import typing
2025-10-10T06:11:32.122Z
File "/workspaces/job/update_layer/lib/python3.12/site-packages/cv2/typing/__init__.py", line 61, in <module>
2025-10-10T06:11:32.122Z
import cv2.mat_wrapper
2025-10-10T06:11:32.122Z
ImportError: libGL.so.1: cannot open shared object file: No such file or directory

Copy code

FROM ubuntu:${VARIANT} AS base
ARG PYTHON_MAJOR_VERSION

FROM base AS layer-deps
COPY jobs/update_layer-deps.pex /deps.pex
RUN PEX_TOOLS=1 python${PYTHON_MAJOR_VERSION} /deps.pex venv --scope=deps --collisions-ok --compile --rm all /workspaces/jobs/update_layer

FROM base AS layer-srcs
ARG PYTHON_MAJOR_VERSION
COPY jobs/update_layer-srcs.pex /srcs.pex
RUN PEX_TOOLS=1 python${PYTHON_MAJOR_VERSION} /srcs.pex venv --scope=srcs --compile --rm all /workspaces/jobs/update_layer

FROM base
COPY --from=layer-deps /workspaces/jobs/update_layer /workspaces/jobs/update_layer
COPY --from=layer-srcs /workspaces/jobs/update_layer /workspaces/jobs/update_layer

WORKDIR /workspaces/jobs_geo_import/

worried-piano-22913

10/14/2025, 10:40 AM

Hi everyone, is there someone who has a clue about this: https://pantsbuild.slack.com/archives/C046T6T9U/p1759762500455099

worried-piano-22913

10/14/2025, 11:39 AM

I think i just found the cause myself, but i'm not sure if it's the intended behaviour. I'll address it in the thread

mammoth-dawn-85816

10/14/2025, 4:49 PM

Hey folks, I’m struggling to understand how to define a

python_distribution

that pulls the package name and version from

pyproject.toml

— this slack mentions this in a few places, and there’s an open issue that seems to cover the same thing, but I don’t know if there’s some recommended way to use

pyproject.toml

as the authoritative source for the package being built — does anyone have advice?

brief-engine-92399

10/16/2025, 2:36 AM

Is there a way to take a pex_binary target and create a requirements.txt file from it? It's a follow up from a previous question - but I'm trying to cross-platform build docker images and it is really really slow. Like 100x+ slower when building arm from x86. I'm wondering if it would be easier to download all packages at runtime instead. I want the packages from a pex compile to do that though. Will take any advice.

thousands-plumber-33255

10/16/2025, 7:15 PM

How can i turn off deprecation warnings for a pex binary target? I tried this:

Copy code

pex_binary(
        name='deps',
        execution_mode='venv',
        include_requirements=True,
        include_sources=False,
        include_tools=True,
        layout='packed',
        env={"PYTHONWARNINGS": "ignore"},
    )

FROM base AS production-deps
ARG PYTHON_MAJOR_VERSION
COPY django/production-deps.pex /deps.pex
RUN PEX_TOOLS=1 python${PYTHON_MAJOR_VERSION} /deps.pex venv --scope=deps --collisions-ok --compile --rm all /bin/production

But that still gives me all those warnings when I run the pex:

Copy code

/bin/production/lib/python3.12/site-packages/pydantic/_internal/_config.py:323: PydanticDeprecatedSince20: Support for class-based `config` is deprecated, use ConfigDict instead. Deprecated in Pydantic V2.0 to be removed in V3.0. See Pydantic V2 Migration Guide at <https://errors.pydantic.dev/2.11/migration/>
  warnings.warn(DEPRECATION_MESSAGE, DeprecationWarning)

I can even see in the /bin/production/pex file that it injects the envs. But why is that ignored at all?

proud-planet-36170

10/16/2025, 8:45 PM

We have a project with a git submodule but it doesn't seem to be picked up by

--changed-since

Copy code

➜  platform git:(main) ✗ git diff HEAD~1 | cat
diff --git a/attendee/attendee-repo b/attendee/attendee-repo
index 8385791..672fa8c 160000
--- a/attendee/attendee-repo
+++ b/attendee/attendee-repo
@@ -1 +1 @@
-Subproject commit 8385791c33bdef6c17953f5ba356413b5d329013
+Subproject commit 672fa8c7f007562ee030cc3ec1f954a007995672

➜  platform git:(main) ✗ git diff --name-only HEAD~1 | cat
attendee/attendee-repo

➜  platform git:(main) ✗ pants \
              --changed-since=HEAD~1 \
              --changed-dependents=transitive \
              list
circleci-etl:docker-image
circleci-etl/circleci_etl/job_tests.py:../image-files
circleci-etl/circleci_etl/job_tests.py

numerous-pharmacist-91083

10/16/2025, 11:51 PM

I have a python build that uses pytest and a global

conftest.py

in my root directory. That sets up some custom markers and associated commnd line options like

--run_large

. It was all working just fine but then I upgraded to Python 3.12 which wasn't compatible with the pytest version Pants uses by default so I setup a resolve for the test tool and specified a newer version of pytest like:

Copy code

[pytest]
args = ['--log-cli-level=INFO']
install_from_resolve = "tools"

That is now using the newer version of pytest but it's no longer running my

conftest.py

even if I add:

Copy code

[python-infer]
conftests = true

I can verify that because (1) I added print statements to things like my

pytest_addoption

method and (2) the command line arguments like

--run_large

are no longer recognized. How can I fix it?

hundreds-carpet-28072

10/17/2025, 1:20 PM

What’s my best option for detecting changes to target versions in BUILD files that is general to target type?

great-river-19742

10/19/2025, 9:45 AM

How is the final config assembled if I have multiple configfiles (eg

PANTS_CONFIG_FILES=pants.ci.toml

in addition to pants.toml). Say I have in both files a section

[blah]

. Does "[blah] from ci.toml" completely overrides the default's [blah] or does some smart merging of config entries happen?

happy-kitchen-89482

10/19/2025, 10:17 PM

https://pantsbuild.slack.com/archives/C0D7TNJHL/p1760912241192899

brainy-airline-59624

10/20/2025, 6:30 AM

I'm attempting to add a requirement to a requirement.txt file. This requirement is a raw github URL of the form git+https://github.com/...@v0.x.xxx When I add this to the requirement file, it breaks. mbp293@MBP293-Z zero-ai % pants check :: 151840.09 [ERROR] 1 Exception encountered: Engine traceback: in

check

goal in Find targets from input specs ValueError: Invalid requirement 'git+https://github.com/...@v0.x.xxx in third_party/zero-api_requirements.txt at line 9: Expected end or semicolon (after name and no valid version specifier) git+https://github.com/...@v0.x.xxx One thing I found recommended that I put this into a

python_requirement

block and it seems to grab and build this repo fine. The problem is I want to include this third_party in my default resolve so that it's a part of the lockfile I use with my IDE. Perhaps there's some simple way to do this, but I'm having a hard time solving this.

hundreds-carpet-28072

10/20/2025, 4:12 PM

I’m moving my specified dependencies from a requirements.txt to pyproject.toml to enable using

uv

lockfiles and converting the output into Pex lockfiles. Does Pants support the

[dependency-groups]

fields in pyproject.toml in any way? It would be useful to be able to generate separate resolves from these as I have in the past for dependencies related to: test, tools, docs etc.

ancient-beard-76775

10/20/2025, 8:17 PM

Hey everyone, I’m trying out Pants to build my Scala projects and hitting an error because the version of scalameta used for dependency inference doesn’t support some of the syntax I’m using. Is there a way I can override the version of scalameta that’s used, or do I need to wait for Pants itself to upgrade the version?

brainy-airline-59624

10/21/2025, 8:47 AM

I continue to get this error when trying to run a build on Github:

Copy code

tderr:
There was 1 error downloading required artifacts:
1. zero-api-python 1 from git+<https://github.com/zeals-co-ltd/zero-api.git@v0.1.631>
    pid 1487 -> /home/runner/.cache/pants/named_caches/pex_root/venvs/1/d881a8f5dff01abf8f60d8ca212d5cf43381f85e/302b21f4c5ac9f243cbb9cb5d57e98957a60ea02/bin/python

I cut out a bunch of code, but the basic problem is we have a repo, that is installing from a private github repo. We want to pass the credentials in. Every other repo in our codebase has a standardized way of doing it that basically looks like:

Copy code

- name: Configure git for private modules
        run: |
          git config --global url."https://${{ secrets.ORG_PROJECTS_TOKEN }}@github.com/".insteadOf "<https://github.com/>"

Fairly standard. I think I'm failing to pass this into PantsBuild. Any ideas on what I'm missing? No other repos use PantsBuild and so no one else at the company knows anything about this.

ancient-beard-76775

10/22/2025, 12:53 PM

Making progress in building my Scala projects, but I’ve got a unique setup that I’m not sure how to get working with Pants TL;DR: Is it possible to have a target depend on another target that uses a different

resolve

? I’m getting a

NoCompatibleResolve

error In sbt I have a project called

predef

that defines an object with the full package path

example.Predef

. Every other project then gets an additional scalac arg:

Copy code

-Yimports:java.lang,scala,scala.Predef,example.Predef

This makes it so that everything defined in

example.Predef

is available without needing any imports I’m trying to mirror this in my Pants config. In

pants.toml

I have:

Copy code

[jvm.resolves]
jvm-default = "3rdparty/jvm/deps.lock"
uses-predef = "3rdparty/jvm/deps.lock"

[scala.version_for_resolve]
jvm-default = "3.7.3"
uses-predef = "3.7.3"

[scalac]
args = ["-deprecation", "-encoding", "utf8"]

[scalac.args_for_resolve]
uses-predef = ["-Yimports:java.lang,scala,scala.Predef,example.Predef"]

3rdparty/jvm/BUILD

I have:

Copy code

jvm_artifact(
  name = "jvm-default-org.scala-lang_scala3-library_3",
  group = "org.scala-lang",
  artifact = "scala3-library_3",
  version = "3.7.3",
  resolve = "jvm-default",
)

jvm_artifact(
  name = "uses-predef-org.scala-lang_scala3-library_3",
  # ...same group, artifact, and version as above...
  resolve = "uses-predef",
)

And in a target (

shared/src/main/scala/example/BUILD

) that should get the extra scalac arg I have:

Copy code

scala_sources(
  resolve = "uses-predef",
  dependencies = ["predef/src/main/scala/example"],
)

When I run

pants check ::

I get this error:

Copy code

NoCompatibleResolve: The selected targets did not have a resolve in common:

jvm-default:
  * 3rdparty/jvm:jvm-default-org.scala-lang_scala3-library_3
  * predef/src/main/scala/example/Predef.scala

uses-predef:
  * 3rdparty/jvm:uses-predef-org.scala-lang_scala3-library_3
  * shared/src/main/scala/example/Example.scala

Targets which will be merged onto the same classpath must share a resolve (from the [resolve](<https://www.pantsbuild.org/2.29/reference/targets/deploy_jar#resolve>) field).

brief-engine-92399

10/23/2025, 2:23 AM

May be more of a pex question, but any idea why if I try to run pex --compile locally after a pants package of a pex_binary, what I see in the venv makes sense, but when I try to run it in a

docker_image

, packages are missing?

gray-apple-58935

10/24/2025, 11:40 AM

I ran into similar issue, debugger misses all breakpoints in

pants test --debug-adapter ...

, if I run

pants run --debug-adapter ...

, top level code is executed but not test body. I put more context here - https://github.com/pantsbuild/pants/discussions/22788 is this a bug or have I missed some setting in

launch.json

pants.toml

silly-queen-7197

10/24/2025, 8:30 PM

Does pants change

pants test foo#bar

into

pytest foo -k bar

? I realize I don’t know how to do

pytest foo::bar

fast-school-44220

10/24/2025, 8:35 PM

May I ask a history question? How did pants come to use the term "targets" for primary sources? This is causing some Abbot & Costello style communication issues on my team (coming from Make style tools) sometimes.

fast-school-44220

10/24/2025, 11:27 PM

annoying: GitHub identifies pants BUILD files as being written in Starlark.

mammoth-dawn-85816

10/27/2025, 9:25 AM

Hey folks! I’m having trouble with in-monorepo (source layout) dependencies declared with poetry in pyproject.toml, when running

pants test ::

. The structure of the dependency isn’t flattened, so the references to it fail (when

pytest

works correctly, outside of pants). What am I missing? •

src/libs/package_1

depends on

src/libs/package_2

, ◦ declared with a line like

package_2 = { path = "../package_2", develop = true }

in the

[tool.poetry.dependencies]

section of my

pyproject.toml

◦ Code for package 2 lives in

src/libs/package_2/src/package_2/*.py

◦ both have a single BUILD file in the package root, with

python_sources

python_tests

and

poetry_requirements

(sources depends on poetry, tests on both) •

pytest

, run outside of pants, executes tests in

src/libs/package_1/tests/test_*.py

◦ It correctly loads

package_2

source files and tests succeed ◦ I believe this is because of the line

packages = [{ include = "package_2", from = "src" }]

inside

[tool.poetry]

of my package_2

pyproject.toml

. •

pants test src/libs/package_1::

fails with

ModuleNotFoundError: No module named 'package_2'

◦ When I keep & inspect the sandbox, the

src/libs/package_2/

dir is empty except for

src/package_2/*.py

◦ I believe

package_1

is unable to read the source in

../package_2/*.py

because it’s actually in

../package_2/src/package_2/*.py

but there’s no

pyproject.toml

there to include that — does this seem correct? •

pants run tools/whatever::

(which depends on

package_1

) successfully runs using

package_1

(and its dependency on

package_2

) I’ve tried including a

resources(name="pyproject", sources=["pyproject.toml"])

package_2

, (with both the package_2

python_sources

depending on it, and package_1's

python_tests

) — with no change in the outcome, despite the package_2

pyproject.toml

being present in the sandbox. I think I’m either missing something stupid; or the approach I’m taking with using

poetry

for dependency management is confusing matters. Can anyone suggest debug steps?

quaint-piano-62770

10/27/2025, 2:32 PM

I am trying to generate lockfiles where my requirements have `torch`/`torchvision` but the

+cpu

wheels. For this I add

find_links

option to the

[python-repos]

section in the

pants.toml

file

Copy code

[python-repos]
indexes = [
  "<https://pypi.org/simple>",
  "%(env.PIP_EXTRA_INDEX_URL)s"
]
find_links = [
  "<https://download.pytorch.org/whl/torch/>",
  "<https://download.pytorch.org/whl/torchvision/>"
]

But I get this error when trying to run `pants generate-lockfiles`:

Copy code

File "/Users/narayan/.cache/pants/named_caches/pex_root/venvs/1/32a2d4a7e846b6328d7526fa44a431609f09ae7c/2fa36aa9e14d5074d008690effb919f0552cfb0a/lib/python3.11/site-packages/pip/_vendor/msgpack/fallback.py", line 968, in _pack_map_pairs
pip:     self._pack(v, nest_limit - 1)
pip:   File "/Users/narayan/.cache/pants/named_caches/pex_root/venvs/1/32a2d4a7e846b6328d7526fa44a431609f09ae7c/2fa36aa9e14d5074d008690effb919f0552cfb0a/lib/python3.11/site-packages/pip/_vendor/msgpack/fallback.py", line 819, in _pack
pip:     raise ValueError("Memoryview is too large")
pip: ValueError: Memoryview is too large

If I run one resolve at a time it works,

pants generate-lockfiles --resolve=[resolve-name]

but obviously this is inconvenient. Any solutions?

hundreds-carpet-28072

10/28/2025, 10:56 AM

Could somebody provide a link as to the exact spec that Pex uses for its lockfiles? I can’t seem to find it searching docs/github.

billowy-tiger-59247

10/28/2025, 4:57 PM

Hey folks! We are dealing with an interesting behavior that got me trying to understand how Pants deals with test dependencies at test runtime. What happens in this case: • service resolve: dependency A==1.0.0 • pytest resolve (using the

pytest.install_from_resolve

): dependency A=2.0.0 Which dependency will be picked up during test execution? The one from the original resolve or the one from the pytest resolve?

numerous-pharmacist-91083

10/28/2025, 11:03 PM

I'm facing some very slow Python build times. Any time any dependency is changed the build takes forever, primarily on steps like

Building 21 requirements for requirements.pex

which I think is building a pex isolated environment for unit tests and/or mypy runs. i think the main culprit is a few big libraries like PyTorch and torchvison but, for obvious reasons, I can't eliminate those. It's gotten so bad that I just added a single dependency and my CI is now timing out after 2 hours even though I'm caching the pants cache in my CI. Is there anything I can do to speed this up? I'm looking at options like maybe running with

execution_mode = 'venv'

or run_against_entire_lockfile
but I don't have a good enough understand of how Pants, pex, testing, pip, etc. all interact and since each experiment takes about 2 hours some advice would be greatly appreciated.

ripe-architect-1001

10/31/2025, 1:59 PM

Maybe a stupid question but for using

ruff

for example if I just want to use it but I don't need any additional configuration I just need to add the backend, I don't need a config section?

curved-manchester-66006

10/31/2025, 2:12 PM

🧪 There are recent RCs on all current branches. If these resolve an issue you were experiencing please take them for a spin. • https://github.com/pantsbuild/pants/releases/tag/release_2.27.1rc0 • https://github.com/pantsbuild/pants/releases/tag/release_2.28.1rc0 • https://github.com/pantsbuild/pants/releases/tag/release_2.29.1rc1