fix AD demos on Julia nightly

[simeonschaub]

This was broken by JuliaLang/julia#38136

[simeonschaub]

Are we not testing on nightly currently?

[nickrobinson251]

Are we not testing on nightly currently?

I thnk we're supposed to be (https://github.com/JuliaDiff/ChainRulesCore.jl/blob/decfca42702ff11a6c4b97c3563498e19415407a/.github/workflows/JuliaNightly.yml)

But doesn't look like it's actually running. Maybe @oxinabox knows what the situation is?

[simeonschaub]

Ah, so it's not run on every PR, but instead just once per day on master. I think it would make sense to at least test nightly on one platform on every PR, so problems are noticed right away.

[nickrobinson251]

I think it would make sense to at least test nightly on one platform on every PR, so problems are noticed right away.

opened #300 :)

[simeonschaub]

How does CD here work nowadays? Should I bump the version as well and then register afterwards? Or do we now have a tool for that?

[oxinabox]

Should I bump the version as well and then register afterwards?

Yes, do that

[codecov-io]

Codecov Report

Merging #299 (df97168) into master (decfca4) will decrease coverage by 1.75%.
The diff coverage is n/a.

@@            Coverage Diff             @@
##           master     #299      +/-   ##
==========================================
- Coverage   89.61%   87.85%   -1.76%     
==========================================
  Files          13       13              
  Lines         462      420      -42     
==========================================
- Hits          414      369      -45     
- Misses         48       51       +3     

Impacted Files	Coverage Δ
src/differentials/thunks.jl	56.52% <0.00%> (-7.48%)	⬇️
src/ruleset_loading.jl	94.73% <0.00%> (-3.14%)	⬇️
src/differentials/composite.jl	79.80% <0.00%> (-2.65%)	⬇️
src/rule_definition_tools.jl	93.57% <0.00%> (-0.74%)	⬇️
src/accumulation.jl	96.66% <0.00%> (-0.40%)	⬇️
src/differentials/abstract_zero.jl	93.75% <0.00%> (-0.37%)	⬇️
src/differential_arithmetic.jl	96.15% <0.00%> (-0.10%)	⬇️

---

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 decfca4...df97168. Read the comment docs.