MNT: drop 3.5 testing for 3.1 branch by jklymak · Pull Request #12538 · matplotlib/matplotlib

jklymak · 2018-10-16T21:39:40Z

PR Summary

As per dev-list email, dropping 3.5 tests from Appveyor and Travis...

Feel free to push or close this if I did it incorrectly!

Probably todo is Circle-ci. Do we want to add 3.7 to that?

PR Checklist

Has Pytest style unit tests
Code is Flake 8 compliant
New features are documented, with examples if plot related
Documentation is sphinx and numpydoc compliant
Added an entry to doc/users/next_whats_new/ if major new feature (follow instructions in README.rst there)
Documented in doc/api/api_changes.rst if API changed in a backward-incompatible way

tacaswell · 2018-10-16T21:44:57Z

The 3.5 test was pinning all of the minimum deps, we need to move those over to one of the other tests.

tacaswell · 2018-10-16T21:45:54Z

-      dist: trusty
-      # pytest-cov>=2.3.1 due to https://github.com/pytest-dev/pytest-cov/issues/124.
-      env:
-        - PINNEDVERS='-c requirements/testing/travis35.txt'


Also remove this file (or rename it to be the 3.6 pins).

We were pinning numpy to 1.10 on the other builds?

~~It looks like the Travis 3.6 build was getting reqs from travis_all.txt, so I think this is OK...~~ Oh duh, I see we need to make sure one of the tests tests the minimums. OK. Thats a bit more work.

OK, see new file. I'm not sure if the other pinned reqs in travis35.txt were required to get it to run on 3.5 or if they were min deps we were wanting to test. I suspect the former, so I removed them.

pytest and sphinx are pinned as minimum deps also.

Why is that? Are we testing that devs can use old Sphinx and pytest? Or are those user facing pins?

I suppose so; those are the versions we specify in the docs.

Fixed, - wasn't sure if w need the pin on pytest-timeout so trying w/o

jklymak · 2018-10-17T16:05:57Z

Looks done to me, but let me know if I missed something!

tacaswell

I want to be the one to merge this so abuse the the 'request changes' feature to that end.

timhoffm

Looks good.

timhoffm · 2018-12-03T20:27:38Z

+pandas<0.21.0
+pytest==3.6
+pytest-cov==2.3.1
+# pytest-timeout==1.2.1  # Newer pytest-timeouts don't support pytest 3.4.		


Can be removed instead of uncommented.

jklymak · 2018-12-03T21:05:04Z

Sorry @timhoffm and @tacaswell - I messed up the rebase and had to do again. Clearing your approvals pending a new quick re-review

stale after bad rebase.

timhoffm

travis36minver.txt is missing

jklymak · 2018-12-03T21:51:21Z

Should there be a ci/circleci:docs-python37? I'm not really clear on the point of running two doc builds.

jklymak · 2018-12-03T22:02:59Z

Fixed now - sorry for the confusion!

Stale

anntzer · 2018-12-04T17:16:53Z

          name: "Deploy new docs"
          command: ./.circleci/deploy-docs.sh

-  docs-python35:


Let's move to a 3.7 docker build, and pin numpy 1.11 on the 3.6 build?
The CircleCI build does complement the coverage of the test suite.

Absolutely, but does doing it on 3.6 and 3.7 buy us that much?

test with oldest numpy (1.11) on 3.6, latest on 3.7.

anntzer · 2018-12-04T17:17:31Z

@@ -0,0 +1,6 @@
+# Extra pip requirements for the first travis python 3.6 build


Needs to delete travis35.txt no?

tacaswell · 2018-12-04T19:49:53Z

pushed one more commit to (I think) get the circle 37 job to actually run!

jklymak · 2018-12-04T20:31:28Z

Thanks @tacaswell. Not sure I understand the codecov complaint...

jklymak · 2018-12-04T20:42:29Z

I wonder if the azure pip install requirements need to be tweaked? And should the azure 3.6 test pin numpy to 1.11?

tacaswell · 2018-12-04T21:01:57Z

As mentioned on the call we need to look at moving all the tests over to azure so I'm less worried about getting it's pinnings right yet.

tacaswell · 2018-12-04T21:05:01Z

In test_animation we no longer hit the 'false' path on that conditional. Not sure on the other one yet but I suspect it is something similar.

jklymak · 2018-12-04T21:33:36Z

Fixed test_animation. No idea how to fix backend_ps...

jklymak · 2018-12-04T22:38:09Z

Fixed itself 😉

jklymak added this to the v3.1 milestone Oct 16, 2018

jklymak closed this Oct 16, 2018

jklymak reopened this Oct 16, 2018

tacaswell reviewed Oct 16, 2018

View reviewed changes

jklymak added status: work in progress and removed status: work in progress labels Oct 16, 2018

tacaswell requested changes Oct 18, 2018

View reviewed changes

jklymak mentioned this pull request Nov 22, 2018

Scatter color: moving #10809 forward #12422

Merged

6 tasks

tacaswell previously approved these changes Dec 3, 2018

View reviewed changes

timhoffm previously approved these changes Dec 3, 2018

View reviewed changes

timhoffm added the status: needs rebase label Dec 3, 2018

jklymak force-pushed the mnt-drop35-travis branch 2 times, most recently from 28352de to 2ad5f6e Compare December 3, 2018 21:03

timhoffm approved these changes Dec 3, 2018

View reviewed changes

timhoffm previously requested changes Dec 3, 2018

View reviewed changes

MNT: remove py 3.5 from testing

09ce51d

jklymak force-pushed the mnt-drop35-travis branch from 2ad5f6e to 09ce51d Compare December 3, 2018 21:39

jklymak removed the status: needs rebase label Dec 3, 2018

anntzer reviewed Dec 4, 2018

View reviewed changes

jklymak added 2 commits December 4, 2018 10:30

MNT: add circleci37

b045ff2

MNT: remove trvis35.txt

6a6f414

jklymak force-pushed the mnt-drop35-travis branch from 7590697 to 6a6f414 Compare December 4, 2018 19:30

Update config.yml

2bb34e5

MNT: fix test version check

85eef19

jklymak force-pushed the mnt-drop35-travis branch from 0031c6c to 85eef19 Compare December 4, 2018 21:53

tacaswell merged commit e2045c3 into matplotlib:master Dec 5, 2018

timhoffm mentioned this pull request Dec 17, 2018

Dropping support for Py3.5 and numpy 1.10 #12358

Closed

		@@ -0,0 +1,6 @@
		# Extra pip requirements for the first travis python 3.6 build

Conversation

jklymak commented Oct 16, 2018

PR Summary

PR Checklist

Uh oh!

tacaswell commented Oct 16, 2018

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

jklymak Oct 16, 2018 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

jklymak commented Oct 17, 2018

Uh oh!

tacaswell left a comment

Choose a reason for hiding this comment

Uh oh!

timhoffm left a comment

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

jklymak commented Dec 3, 2018

Uh oh!

timhoffm left a comment

Choose a reason for hiding this comment

Uh oh!

jklymak commented Dec 3, 2018

Uh oh!

jklymak commented Dec 3, 2018

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

tacaswell commented Dec 4, 2018

Uh oh!

jklymak commented Dec 4, 2018

Uh oh!

jklymak commented Dec 4, 2018

Uh oh!

tacaswell commented Dec 4, 2018

Uh oh!

tacaswell commented Dec 4, 2018

Uh oh!

jklymak commented Dec 4, 2018

Uh oh!

jklymak commented Dec 4, 2018

Uh oh!

Reviewers

Uh oh!

Assignees

Labels

Projects

Milestone

Development

Uh oh!

5 participants

jklymak Oct 16, 2018 •

edited

Loading