@fingolfin could you please mark the ci jobs for 1.0-1.3 as no longer required?

Usually, dropping support for earlier version of Julia is not considered as breaking.

Changing Julia version compatibility must be a non-breaking feature

https://github.com/SciML/ColPrac#dropping-support-for-earlier-versions-of-julia

Codecov Report

Merging #221 (0af536b) into master (7f18f15) will increase coverage by 12.63%.
The diff coverage is 100.00%.

@@             Coverage Diff             @@
##           master     #221       +/-   ##
===========================================
+ Coverage   76.81%   89.45%   +12.63%     
===========================================
  Files          11       10        -1     
  Lines         772      512      -260     
===========================================
- Hits          593      458      -135     
+ Misses        179       54      -125     

Flags with carried forward coverage won't be shown. Click here to find out more.

📣 We’re building smart automated test selection to slash your CI/CD build times. Learn more

Usually, dropping support for earlier version of Julia is not considered as breaking.

Changing Julia version compatibility must be a non-breaking feature

https://github.com/SciML/ColPrac#dropping-support-for-earlier-versions-of-julia

This is true for packages with a x.y.z versioning scheme. As Aqua.jl is on version 0.y.z, I can either release this as breaking, or as a patch release. But the julia version should not be restricted on a patch release.

And there will be more breaking changes released together with this in one batch.

As requested I marked the ci jobs for 1.0-1.3 as no longer required.

Oh, I'll also add the new 1.4 test for Windows and macOS to the list of required tests.

Oh, I'll also add the new 1.4 test for Windows and macOS to the list of required tests.

Can you add "enforce PR labels" as well?

@lgoettgens BTW you could have just restarted that one failing job, instead of restarting all :-)

@lgoettgens BTW you could have just restarted that one failing job, instead of restarting all :-)

I know, but I wanted to make sure that this is not something failing on a small percentage of runs, and more runs would maybe find that :)

Windows-1.4 seems to need a bit more time than 10 minutes. I tried to lengthen the timeout for windows jobs.