Volodymyr Agafonkin
Volodymyr Agafonkin
@mroch this is ready for review. While the performance didn't regress, there's no point in optimizing the code much because almost all time is spent in `flow-parser`. Preliminary benchmarks show...
@thehogfather apologies — can't get to it at the moment but hopefully I'll find some time in a week or two. The only comment left unaddressed is this one: >...
Yeah, that's something I've been worried about myself — it consistently shows up as the biggest bottleneck on practical applications such as triangulation polygons in OSM data for map rendering...
@smallsaucepan perhaps RBush isn't the best library to use in this case, because it's meant for dynamic cases where you have to add or remove points often. For static cases...
You're right, I should look into this.
@perliedman thoughts?
I think I agree on this. Mixins like this throughout the code don't play well with tree-shaking, so we might want to look into this in other places as well....
Hi @smallsaucepan, this sounds good to me! Perhaps instead of 3, we should do the same treatment to `geokdbush` as I did for `kdbush` (first-class types), but that can wait...
ESLint focuses on JS linting, but we could explore https://stylelint.io/
@Falke-Design just pushed a fix for that in https://github.com/Leaflet/Leaflet/commit/25fce1684789da91e3c8db10e308b7f6274e66e6