Add a note about polled events and the later package

[DavisVaughan]

Documenting that switching from 200ms to 50ms helped with posit-dev/positron#7235

We did the switch in this commit 452b84f

[lionel-]

hmm if I set it to 1250ms or even disable polled events entirely I still get a short time.

[DavisVaughan]

@lionel- I think I know why. Winston added later::run_now() to his original example, which messes things up if you're trying to rely on the R_InputHandlers being flushed. Try this

# Put these in a block to ensure they're run without delay in between.
{
  later::later(\() { finish <<- Sys.time(); cat("Finished!")}, 0)
  start <- Sys.time()
}

# Wait for a moment after running block above.
# Then run:
finish - start

This respects the 1250ms delay, and gets faster with the 50ms delay.

(As he noted, we don't see the cat("Finished!") output in this case)

[DavisVaughan]

It's definitely tied to this delay, we run the input handlers here

ark/crates/ark/src/sys/unix/interface.rs

Line 100 in 3d42829

pub fn run_activity_handlers() {

Which is called from process_idle_events()

ark/crates/ark/src/interface.rs

Lines 1893 to 1902 in 3d42829

	fn process_idle_events() {
	// Process regular R events. We're normally running with polled
	// events disabled so that won't run here. We also run with
	// interrupts disabled, so on Windows those won't get run here
	// either (i.e. if `UserBreak` is set), but it will reset `UserBreak`
	// so we need to ensure we handle interrupts right before calling
	// this.
	unsafe { R_ProcessEvents() };
	crate::sys::interface::run_activity_handlers();

Which is what we do on each 50ms tick

ark/crates/ark/src/interface.rs

Line 924 in 3d42829

Self::process_idle_events();

[lionel-]

Also add link to #791?

[lionel-]

No need, this is unrelated to the {later} issue.

@DavisVaughan Should we use an even smaller timeout?