Allow ChainRulesCore 1.0

[oxinabox]

I haven't actually looked at the code to see if anything else is needed,
lets see what CI says

[codecov-commenter]

Codecov Report

Merging #102 (f388eff) into master (787a6a4) will decrease coverage by 0.36%.
The diff coverage is n/a.

@@            Coverage Diff             @@
##           master     #102      +/-   ##
==========================================
- Coverage   75.13%   74.76%   -0.37%     
==========================================
  Files          19       19              
  Lines         543      543              
==========================================
- Hits          408      406       -2     
- Misses        135      137       +2     

Impacted Files	Coverage Δ
src/similar_convert_copy.jl	84.09% <0.00%> (-2.28%)	⬇️
src/array_interface.jl	84.00% <0.00%> (-0.80%)	⬇️

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 787a6a4...f388eff. Read the comment docs.

[DhairyaLGandhi]

This fails Flux tests after upgrading to ChainRules 1.0, so we definitely have fixes needing to go in here. https://buildkite.com/julialang/flux-dot-jl/builds/1496#7bb3aa82-4fcd-4b7f-a3c9-e3de7006c4fa

[DhairyaLGandhi]

Actually, Flux tests are pulling in ComponentArrays@0.10.5 instead of 0.11, so the compat entry and a release would be great @jonniedie

[DhairyaLGandhi]

Also tests seem to be pulling an old version of Zygote in CI (v0.4.20 vs v0.6.19). Might be that other compat bounds need to be set.

[DhairyaLGandhi]

Could we re run ci here?

[jonniedie]

To be honest, I'm not really sure how to troubleshoot this. Running the tests locally in a fresh temp environment doesn't run into any of these problems. I also can't reproduce the Zygote @v0.4.20 thing, even by forcing all of the versions of the other packages to the versions they show in the CI run.

Any ideas?

[oxinabox]

we may just need to restart the CI.
It was maybe triggered before everything was registered?

[jonniedie]

Oh, well it looks like that did it. There's a segfault on nightly, but it doesn't seem to have to do with ComponentArrays. Is that a Zygote thing?

[DhairyaLGandhi]

I've been seeing it in many places but not sure where it's coming from.

[jonniedie]

Okay cool. Well I'll go ahead and merge this then. Nightly's gonna nightly.

[jonniedie]

Thanks for doing this @oxinabox. Sorry it took so long for me to get around to merging it.

[oxinabox]

no problem.