Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Dependency Updates #230

Open
flashpixx opened this issue Aug 12, 2024 · 4 comments
Open

Dependency Updates #230

flashpixx opened this issue Aug 12, 2024 · 4 comments

Comments

@flashpixx
Copy link

Hello,

I like py-feat very much, but I stick sometimes in dependency issues e.g. numpy / pandas is version 1. So is it possible to update the dependencies to latest stable version(s).
In this case my question is also to use another build system e.g https://python-poetry.org/

Maybe I can try to update all components and create a PR if I find some time, but is a change of the build system in that case okay?

@ljchang
Copy link
Member

ljchang commented Sep 3, 2024

hi @flashpixx , thanks for your suggestion. We have almost finished a pretty large refactor to update and streamline dependencies. @ejolly, do you have any thoughts about switching to poetry?

@flashpixx
Copy link
Author

flashpixx commented Sep 4, 2024

Poetry supports a poetry.lock file (similar to package.lock on Node.js / NPM project), the file allows to create better reproducable deterministic builds see the documentation. If you take a look into my second issue #233 it try to create a reproducable build, because py-feat crashes, but with the lock file you can figure the exactly installed dependecy versions.

@ejolly
Copy link
Contributor

ejolly commented Sep 4, 2024

Hey @flashpixx we can definitely update dependency versions, and likely will before we release an update after we finish our refactor.

At this time we're not going to switch over to poetry for a few reasons (but thanks for the offer of help to PR it!):

  • we have several other packages we maintain and want to share standard build and development tooling across all of them and would need to switch everything to poetry
  • some of those projects rely on packages that have pre-built packages on conda or conda-forge
  • having had experience with poetry for a few other projects, what it offers in convenience comes at the cost of non-standard operation, particularly with how semantic versioning is handled compared to setuptools,,pip, and conda. This had lead to unexpected GA issues where we rebuild our package for testing on a regular basis and don't get the install behavior we expect

I've been keeping an eye on poetry for a while, but at this stage don't feel like it offers us enough benefits to warrant switching over. And the switch cost of moving between tools (e.g. poetry, pip, conda) as we move between our different projects, is a little too much for our small development team :)

Happy to revisit in the future!

@flashpixx
Copy link
Author

Conda & Poetry can work together see Medium. In general when you publish all of your projects, it may be possible to createa PR with a new build system

ejolly added a commit that referenced this issue Oct 19, 2024
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

3 participants