Rename parallel feature to rayon.

[Anders429]

This is now possible because of namespaced dependencies, introduced in rustc 1.60.0.

This PR contains a temporary workaround for trybuild issue #171.

[codecov-commenter]

Codecov Report

Merging #81 (514d07d) into master (9f87ea3) will not change coverage.
The diff coverage is n/a.

@@           Coverage Diff           @@
##           master      #81   +/-   ##
=======================================
  Coverage   40.16%   40.16%           
=======================================
  Files          56       56           
  Lines        5920     5920           
=======================================
  Hits         2378     2378           
  Misses       3542     3542           

Impacted Files	Coverage Δ
src/archetype/mod.rs	10.50% <ø> (ø)
src/archetypes/mod.rs	49.65% <ø> (ø)
src/archetypes/par_iter.rs	0.00% <ø> (ø)
src/query/result/par_iter.rs	0.00% <ø> (ø)
src/query/view/assertion_buffer.rs	100.00% <ø> (ø)
src/system/mod.rs	0.00% <ø> (ø)
src/system/null.rs	0.00% <ø> (ø)
src/system/par.rs	0.00% <ø> (ø)
src/system/schedule/mod.rs	0.00% <ø> (ø)
src/world/mod.rs	0.00% <ø> (ø)

Help us with your feedback. Take ten seconds to tell us how you rate us. Have a feature suggestion? Share it here.

[Anders429]

Rerunning the miri job. I want to know whether it's flaky (which is not good) or if it's just something that is present in the branch but not in the actual pull to master.

[Anders429]

Hmm. Looks like miri passes now, which is actually worse than if it had failed again. There may be something fishy in the Schedule code.

For reference, this is the failing job: https://github.com/Anders429/brood/runs/7757239434?check_suite_focus=true

Anyway, this doesn't need to block this PR. Opened #82 to track that issue separately.