Use backports.functools_lru_cache instead of functools32 #7871

Merged
merged 1 commit into from Feb 16, 2017

Conversation

Projects
None yet
5 participants
Contributor

felixonmars commented Jan 19, 2017

It's better maintained and more widely used (by pylint, jaraco, etc).

QuLogic requested a review from mdboom Jan 19, 2017

tacaswell added this to the 2.1 (next point release) milestone Jan 20, 2017

Owner

tacaswell commented Jan 20, 2017

I am mixed on this. A bit worried that changing the deps is going to just cause havoc for down-stream packagers as this is a new for 2.0 dependency, but sympathetic to using a more widely used / better maintained backport.

Contributor

felixonmars commented Jan 21, 2017

I guess the havoc would be limited, since most distros have pylint packaged, they will have backports.functools_lru_cache already :)

Contributor

anntzer commented Jan 23, 2017

Apparently pylint only started depending on backports.functools_lru_cache recently (PyCQA/pylint@02f6823). For example, a quick search on pkgs.org suggests that none of CentOS, Debian Jessie, Mageia and Ubuntu (in their latest versions) provide that package right now.

Contributor

HolgerPeters commented Jan 25, 2017

Not a linux distributor, but I bundle packages nonetheless and https://pypi.python.org/pypi/backports.functools_lru_cache seems to be a more mature package on a quick glance.

Contributor

NelleV commented Jan 25, 2017

I am fine either way. @tacaswell I think you should be the one making the decision on this change, as you are the maintainer 😈

Owner

tacaswell commented Jan 30, 2017

I am 👍 on this for 2.1. Can you also put a file in https://github.com/matplotlib/matplotlib/tree/master/doc/api/api_changes with basically the first message from this PR?

@tacaswell

modulo and api_changes entry.

@felixonmars felixonmars Use backports.functools_lru_cache instead of functools32
It's better maintained and more widely used (by pylint, jaraco, etc).
460e19d
Contributor

felixonmars commented Jan 31, 2017

API changes file added :)

Contributor

felixonmars commented Feb 2, 2017

Hrm, the test failure looks irrelevant.

@NelleV NelleV merged commit 4cfbb2f into matplotlib:master Feb 16, 2017

4 of 5 checks passed

codecov/patch 0% of diff hit (target 62.24%)
Details
codecov/project 62.65% (+0.4%) compared to 70831bd
Details
continuous-integration/appveyor/pr AppVeyor build succeeded
Details
continuous-integration/travis-ci/pr The Travis CI build passed
Details
coverage/coveralls Coverage increased (+0.4%) to 62.651%
Details
Contributor

NelleV commented Feb 16, 2017

Thanks @felixonmars !

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment