[Python-Dev] PEP 428 (pathlib) now committed (original) (raw)

anatoly techtonik techtonik at gmail.com
Sat Nov 23 16:15:08 CET 2013

I'd vote for a different perspective on path handling. For me the pathlib is not the good way to go. Especially with copying ill behaviour of old os.path functions.

We definitely need a "task force page" dedicated to "working with paths in Python" to collaborate. ML + PEP with privileged write access is ineffective.

While I like the problem decomposition what I see in pathlib, I don't think it supports the greatest power of Python that it allows you not to think about cross-platform differences - http://clanmills.com/files/dist/doc/cross_platform.html#python-batteries-included I've got the impression that while using pathlib you should care about those.

I am +1 for iterating few more times to research best compromise between the most simple possible and most user friendly interface (wart-free). But every iteration should be historically accessible like in blog report, not like in Mercurial PEP diff. There is still problem to dedicate time though.

-- anatoly t.

On Sat, Nov 23, 2013 at 4:32 PM, Serhiy Storchaka <storchaka at gmail.com> wrote:

22.11.13 18:44, Antoine Pitrou написав(ла):

I've pushed pathlib to the repository. I'm hopeful there won't be new buildbot failures because of it, but still, there may be some platform-specific issues I'm unaware of. Congratuate Antoine! Does it means that issues #11344 (Add os.path.splitpath(path) function) [1] and #13968 (Support recursive globs) [2] have no chance? Both are ready for commit and waits for reviews almost a year. Are the os.path and glob modules deprecated now? [1] http://bugs.python.org/issue11344 [2] http://bugs.python.org/issue13968

Python-Dev mailing list Python-Dev at python.org https://mail.python.org/mailman/listinfo/python-dev Unsubscribe: https://mail.python.org/mailman/options/python-dev/techtonik%40gmail.com

More information about the Python-Dev mailing list