Attempt to fix some of our Travis failures on v3.0.x. Supersedes#13206
MarkInfo objects were removed in pytest 4.0. The other solution would be to backport the PR that fixes this from our master branch.

PR Checklist

Has Pytest style unit tests
Code isFlake 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

NelleV changed the title~~FIX Starting xvfb as a service to avoid travis failures~~Backport #13205 on branch v3.0.x (Add xvfb service to travis)

Jan 18, 2019

Add xvfb service to travis

ae56447

NelleV force-pushed thetravis_failures_v30x branch from06e99aa toae56447Compare

January 18, 2019 05:02

TST downgrading pytest

c129091

NelleV force-pushed thetravis_failures_v30x branch froma1ab49b toc129091Compare

January 18, 2019 06:11

tacaswell added this to thev3.0.3 milestone

Jan 18, 2019

tacaswell mentioned this pull request

Jan 18, 2019

Backport PR #13205 on branch v3.0.x (Add xvfb service to travis)#13206

Closed

Copy link

Member

tacaswell commentedJan 18, 2019

Restarted 3.5 which was the only one that failed and it failed on the Qt tests. I wonder if there is some roll-over going on where we are getting a mixed population of workers, some the new way works and some the oldway works?

Copy link

Member

dstansby commentedJan 18, 2019

Yeah, the 3.5 build is getting axvfb: unrecognized service (https://travis-ci.org/matplotlib/matplotlib/jobs/481215980#L1323)

dstansby added the topic: testing label

Jan 18, 2019

Copy link

Member

tacaswell commentedJan 18, 2019

It is because we force the dist back to trusty, lets see what happens if we drop that....

CI: do not pull py35 test run back to trusty

dcfe153

This is to make sure we have the xvfb as a service

QuLogic approved these changes

Jan 18, 2019

View reviewed changes

Copy link

MemberAuthor

NelleV commentedJan 19, 2019

fui! Great! Thanks a lot for fixing that last problem.

timhoffm approved these changes

Jan 19, 2019

View reviewed changes

timhoffm merged commit4a63ccb intomatplotlib:v3.0.x

Jan 19, 2019

timhoffm added a commit that referenced this pull request

Jan 19, 2019

Revert "Backport#13205on branch v3.0.x (Add xvfb service to travis) (…

d2736de

…#13216)"This reverts commit4a63ccb.

Copy link

Member

timhoffm commentedJan 19, 2019

Hm, maybe I should not have squashed the commits. Should I revert and merge without squashing?

Copy link

MemberAuthor

NelleV commentedJan 19, 2019 via email

No, it's good!

…

On Sat, Jan 19, 2019, 7:30 PM Tim Hoffmann ***@***.*** wrote: Hm, maybe I should not have squashed the commits. Should I revert and merge without squashing? — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub <#13216 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AALR3uTlBGk9gxbqGJ3dAX87GXTf2geYks5vEte_gaJpZM4aHOSy> .

Copy link

MemberAuthor

NelleV commentedJan 19, 2019 via email

Thanks!On Sat, Jan 19, 2019, 7:36 PM Nelle Varoquaux <nelle.varoquaux@gmail.comwrote:

…

No, it's good! On Sat, Jan 19, 2019, 7:30 PM Tim Hoffmann ***@***.*** wrote:> Hm, maybe I should not have squashed the commits. Should I revert and> merge without squashing?>> —> You are receiving this because you authored the thread.> Reply to this email directly, view it on GitHub> <#13216 (comment)>,> or mute the thread> <https://github.com/notifications/unsubscribe-auth/AALR3uTlBGk9gxbqGJ3dAX87GXTf2geYks5vEte_gaJpZM4aHOSy>> .>