replace derivative dependency with educe

[rorosen]

Replace the unmaintained derivative dependency with educe. See also #1583.

Motivation

The derivative dependency is unmaintained and uses syn 1 while everything else uses syn 2.

Solution

I decided to replace derivative with educe instead of derive-more, as derive-more lacks support of skipping/ignoring fields (see #311) and cannot derive Eq and PartialEq. However, I haven't tried too hard to make it work so let me know if I should try to use derive-more anyway.

Although educe can almost be used as a drop-in replacement in most places, it lacks to specify custom bounds for Eq for some reason. I decided to write the manageable impl for Eq myself where needed, i.e. in mod.rs and object_ref.rs

Edit: Noticed by CI checks that educe triggers clippy::unused_underscore_binding when deriving Debug for enums (i.e. for State<K> in watcher.rs). Hence, I added #![allow(clippy::used_underscore_binding)] in kube-runtime, not sure whether that is acceptable...

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 75.3%. Comparing base (edd384e) to head (b57cf5d).
Report is 1 commits behind head on main.

Additional details and impacted files

@@          Coverage Diff          @@
##            main   #1585   +/-   ##
=====================================
  Coverage   75.3%   75.3%           
=====================================
  Files         82      82           
  Lines       7331    7331           
=====================================
  Hits        5519    5519           
  Misses      1812    1812           

Files with missing lines	Coverage Δ
kube-runtime/src/controller/mod.rs	29.9% <ø> (ø)
kube-runtime/src/reflector/dispatcher.rs	96.3% <ø> (ø)
kube-runtime/src/reflector/object_ref.rs	69.4% <ø> (ø)
kube-runtime/src/reflector/store.rs	96.9% <ø> (ø)
kube-runtime/src/scheduler.rs	94.8% <ø> (ø)
kube-runtime/src/utils/delayed_init.rs	97.1% <ø> (ø)
kube-runtime/src/watcher.rs	44.2% <ø> (ø)

[clux]

Thanks a lot for doing this. Finally had the chance to look over this. Was worried about Eq, but it seems all good to me!

[clux]

Marking as a fix as I don't think this will need to be classified as a breaking change. Though, likely one for the next minor anyway.