The "cache read miss" is now the compiler step. Make it more explicit

[sylvestre]

We are getting:

Average cache write               0.115 s
Average compiler write            5.737 s
Average cache read hit            0.000 s

was:

Average cache write               0.115 s
Average cache read miss           5.737 s
Average cache read hit            0.000 s

[codecov-commenter]

Codecov Report

Base: 30.58% // Head: 30.41% // Decreases project coverage by -0.16% ⚠️

Coverage data is based on head (da125cd) compared to base (a28f6c4).
Patch coverage: 22.22% of modified lines in pull request are covered.

Additional details and impacted files

@@            Coverage Diff             @@
##             main    #1527      +/-   ##
==========================================
- Coverage   30.58%   30.41%   -0.17%     
==========================================
  Files          48       48              
  Lines       16704    16704              
  Branches     7935     7931       -4     
==========================================
- Hits         5109     5081      -28     
- Misses       6206     6208       +2     
- Partials     5389     5415      +26     

Impacted Files	Coverage Δ
src/compiler/compiler.rs	35.95% <20.00%> (+0.21%)	⬆️
src/server.rs	31.37% <25.00%> (-0.73%)	⬇️
src/test/utils.rs	36.36% <0.00%> (-4.05%)	⬇️
src/cache/disk.rs	35.61% <0.00%> (-1.37%)	⬇️
src/util.rs	34.74% <0.00%> (-1.30%)	⬇️
src/test/tests.rs	32.65% <0.00%> (-0.69%)	⬇️
src/compiler/c.rs	38.26% <0.00%> (-0.46%)	⬇️
src/compiler/gcc.rs	55.26% <0.00%> (-0.42%)	⬇️
src/cache/cache.rs	38.99% <0.00%> (-0.39%)	⬇️
... and 6 more

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

☔ View full report at Codecov.
📢 Do you have feedback about the report comment? Let us know in this issue.

[Xuanwo]

Others LGTM

[Xuanwo]

Maybe use BoxFuture?

[sylvestre]

sorry but why ? :)

[Xuanwo]

Oh, no strong reasons, just to avoid repeating Pin Box things 🤣

[drahnr]

I think we should split the item into read miss, compile and cache write

[sylvestre]

I think we should split the item into read miss, compile and cache write

what do you mean? it is what we have currently

[drahnr]

I misinterpreted the changeset, if it's only compilation I'd name it as such compile, without the write suffix.
Question here is what does the compilation metric really give us?

[sylvestre]

I misinterpreted the changeset, if it's only compilation I'd name it as such compile, without the write suffix.

Good point, fixed

Question here is what does the compilation metric really give us?

It is free and interesting to see how long a user can save. And it is better to have it than not.