michaelbjames/elm-benchmark

Elm code benchmarking suite

elm-benchmark's Issues

Code review

Goals:

• Have one central library called Benchmark that contains all of the basics.
• Combine types, runner, and benchmark. Make LineGraph independent of the Result type. This is primarily so the docs work, but maybe line graphs can be made reusable as well?

I also could not tell what was going on with setup functions. I think these should either be simpler or in some submodule for complex stuff, like Benchmark.FancySetup or something.

What is the future plan for runMany?

I've been casually looking through the code of this project to see if it can easily be extended with functionality to measure the message throughput of a signal.

When I looked at the runMany function in Runner.js, I found this comment:

/**
 * [...]
 *  This is intentionally a monolithic function. The hope is to clean this
 *  up when we need more functionality and when more features have been
 *  added to Elm. It is possible that Commands will help.
 */

So I was wondering: is the runMany function written so it might some day be ported to Elm?

Naming

What if it was like this:

Benchmark
    logic
    render
    renderStatic
    run

Benchmark.DeferredSetup
    logic
    render

With the intention that people import DeferredSetup qualified as DS or something.

Please publish this package

This package looks great, but I had no way of knowing about it until someone offhandedly mentioned it on the mailing list. Please publish it to package.elm-lang.org!