v1.0 #728
v1.0 #728
Conversation
Codecov ReportAll modified and coverable lines are covered by tests ✅
Additional details and impacted files@@ Coverage Diff @@
## master #728 +/- ##
==========================================
- Coverage 89.57% 89.00% -0.58%
==========================================
Files 11 11
Lines 969 973 +4
==========================================
- Hits 868 866 -2
- Misses 101 107 +6 ☔ View full report in Codecov by Sentry. |
| matrix: | ||
| version: | ||
| - '1.6' | ||
| - 'lts' |
There was a problem hiding this comment.
| - 'lts' | |
| - 'min' | |
| - 'lts' |
(They are not identical)
There was a problem hiding this comment.
It tests the minimum compatible version.
| SpecialFunctions = "1, 2" | ||
| StaticArrays = "1.5" | ||
| julia = "1.6" | ||
| julia = "1.10" |
There was a problem hiding this comment.
Why? Is it broken on < 1.10?
There was a problem hiding this comment.
No, but if we aren't testing on Julia v1.6 it's best to explicitly drop support. It also allows us to remove weakdeps as deps and use any features from v1.10+.
There was a problem hiding this comment.
My point is that the package is tested successfully on 1.6, it's that the PR in its current form dropped tests on 1.6 (but adding min would fix that). Releasing 1.0 doesn't require making any changes to the package or its dependencies.
There was a problem hiding this comment.
To be clear, I'm not generally opposed to making such changes but I think they should be done separately. Releasing 1.0 should just involve updating the version number + removing deprecations (if existent) IMO. Otherwise it's also easy to forget to remove VERSION-specific code, old workarounds, etc.
There was a problem hiding this comment.
I don't really care, push any changes you want to this PR
There was a problem hiding this comment.
No, I don't want to. I want to keep them in separate PRs. And I don't think any of the changes in this PR apart from the 1.0 bump are necessary for releasing 1.0.
There was a problem hiding this comment.
Then make your own PR
2 participants
I also updated the CI to test on lts and bumped the julia version