Should I expose function for benchmarking outside of a suite?

[yamiteru]

Currently we can run benchmarks only inside of suites.

But sometime we might want to benchmark a single function and see the results. In such a case it's kind of meaningless to create a Suite and capture the results with modes like useTerminalCompact that show the slowest/fastest.

---

Do you think this would be useful?
(if yes please specify the use-cases)

[yamiteru]

After some thinking I think it's a good idea to expose this function.

I often work on optimizing one function for some extended periods of time and I don't care about anything else like the name of the suite, what is faster/slower, etc.

I'm still not sure how to handle output modes though. Should I assume this is gonna be used only for quick and dirty testing purposes so it's always going to be something similar to useTerminalCompact and hardcode it in?

Also when it comes to CLI (#5) I think it might be cool to have a file watcher and optionally re-run on every file change. But I guess this is a bit out of scope of this discussion.

[H4ad]

Usually, people will use the benchmark to compare something with another thing, so adding one more API just for some rare cases I think don't worth the effort.

If you make the suite API easier enough to be used and accept just one benchmark, I think it will cover the use cases of the people that want to benchmark just one function.

[yamiteru]

I'm reopening this since I think this is gonna be useful for benchmarking many different things that don't share the same domain so it doesn't make sense to put them into a suite.

For example this is a good example of such a use case https://github.com/nodejs/node/blob/main/benchmark/assert/deepequal-buffer.js.

Also if we think about #5 it's most likely gonna use such a function for running individual benchmark files.

[yamiteru]

I'm closing this in favor of #5. This is no longer needed.