0

I'm writing a python package that depends on a relatively old library that breaks in a very specific case: when running a container in read-only mode, a .wh..wh..opq file is added to a directory where the library isn't expecting it, the library tries to read and process all the files, and breaks.

Not a big deal, a very short PR could fix it. However, the library hasn't accepted a new PR in years. I'd prefer to use the library as it comes from pip or conda (so that my package is easier to distribute), so it seems like my current options are:

  1. Find a hacky way to remove the .wh..wh..opq file from the problematic directory at runtime
  2. Make a PR and hope it gets merged
  3. Fork the library and take it upon myself to maintain and distribute it

For my case, option 1 is probably fine--but all of the options seem pretty bad. So in general: What are the best practices in a situation like this, i.e, when unmaintained but otherwise functional libraries could benefit from extremely minor bug fixes?

New contributor
Will G is a new contributor to this site. Take care in asking for clarification, commenting, and answering. Check out our Code of Conduct.
1
  • 4
    You don't have to maintain it officially. You can simply bundle its code with your own project and pseudo-maintain it only for your own purposes. May 15 at 7:28

1 Answer 1

5

Fork the library and do your own fixes on the fork, no need to maintain it outside your own needs. You can use the fork directly in your code.

Do make PRs back into the original library with your fixes, as long as it makes sense (if you do a major refactor it might be difficult to merge back just bits) maybe they will get merged one day.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.