Thanks for the PR.

1. I've just added this to address please enable event handler as 'passive' #129.
2. I don't think it's necessary to bring in a new package for what is already being done in a few lines of existing code.
3. I've looked into re-adding rAF (it used to be used) but there has never been an obvious performance gain. Even the React folks have been unable to find a reason to use it for batching since the current state update batcher works well. If you can provide an example that shows definitively that using requestAnimationFrame for updating state is going to make the list more performant I'd love to see it.

Re: batching, I believe this is something that really should be handled at the application level - you can inject your own batcher - rather than in individual components.

I've seen performance gains from batching at rAF, but our application processes a lot of potentially redundant data updates so it is a bit special.

@caseywebdev what might definitive proof for (3) look like? Is there a test case or methodology you can suggest? A rather unscientific test of holding down the PGDN key with raf enabled results in significantly less CPU time processing events, and effectively eliminates long frame renders:

Without raf on the left, with on the right. The selected period is small for display purposes, but the results are consistent over any time period.

I just did some profiling on one of our most intensive react-list implementations (~10k rows x 100+ columns) and I found scrolling FPS was the same or even slower when using the requestAnimationFrame batching. This optimization feels like it might be case-specific, in which case @STRML's suggestion of customizing your batcher might be the best bet.

@pdf If you're looking for an RAF batching implementation, see ours.

@STRML I'm aware of the raf batcher, needs work though, I'll take a look at the commit you've just made on that fork.

@caseywebdev I'm curious as to your methodology - I was just using the example page from this repo to test. What sort of FPS were you getting?

Yeah, I found a lot of issues with the base config with regards to page visibility and a tendency to stampede timers. I've found the above implementation to work quite well for us.

Hey all

Just curious , it seems like a good thing to merge?

Bump

No need to bump... I'm not sold on RAF for this and it's also out of date.

Ahh, thanks for commenting