lib: improve esm resolve performance#46652
Conversation
|
Review requested:
|
GeoffreyBooth
left a comment
There was a problem hiding this comment.
On quick glance this seems fine, but we need to land #44710 first. Just putting a hold on this until that gets landed.
d5f008b to
76fb5d6
Compare
aduh95
left a comment
There was a problem hiding this comment.
I'd like to see a benchmark for this, but the theory looks solid. This can land before or after the off-thread PR, the conflicts should be minimal :)
|
Shouldn't it be factorized with the |
I agree, but I prefer to do it in a different pull request, and keep this minimal. |
|
Before we land this shouldn’t there be some attempt at confirming that this does, in fact, make things faster? And especially not slower. Even just running a benchmark on someone’s local machine. |
There was a problem hiding this comment.
To strengthen @GeoffreyBooth 's point
Making sure benchmarks are run in some form before this lands. Please dismiss my review when that happens @anonrig (or whomever)
|
@GeoffreyBooth Benchmark CI: https://ci.nodejs.org/view/Node.js%20benchmark/job/benchmark-node-micro-benchmarks/1332/ Results |
|
Landed in 3e9ed7e |
I couldn't find a benchmark or had the time to write it, but in theory, it should be a lot faster.
cc @nodejs/performance @nodejs/loaders
Refs: nodejs/performance#39