jerch

Here a few numbers with the new approach. Again done on 100k scrollback, from shriniking/enlarging between 55 -65 cols. They should be JIT'ed (did some warmup before profiling), and show...

Something weird is going on during enlarging in the demo. If I increase cols by one I get the following `onResize` dispatching:  Note that I changed cols to 87,...

Runtimes w'o zoomlevel madness: - shrinking ~85ms (~3x faster) - enlarging within bounds ~93ms (~2.7x faster) - enlarging out of bounds ~250ms (same speed)

Sure, for example this contains my terminfo entry for xterm: ```bash $> infocmp # Reconstructed via infocmp from file: /lib/terminfo/x/xterm xterm|xterm-debian|X11 terminal emulator, am, bce, km, mc5i, mir, msgr, npc,...

Would it be possible to move some of the work to idle times with `requestIdleCallback()`?

Just some brainstorming on this (as I have lit. no clue how its organized currently on either webgl or canvas end): - Put standard glyphs (as of ASCII print range,...

@Tyriar Is there a middle ground possible? Its just a gut feeling, but kinda both seem slightly off to me - like canvas is too close and webgl too far?...

About lazy resizing of the underlying array buffer - see https://github.com/xtermjs/xterm.js/pull/4144 @Tyriar I currently dont like the idea to place lines * `DebouncedIdleTask` objects on the terminal buffer, also placing...

@Tyriar Yeah WeakMap and WeakSet are very awkward and mostly not of much use. Will see if its useful here (have currently benchmark issues, seems master has a bad perf...

> I think the resize experience is best done by the embedder as it's easy to wrap `resize` calls in a debouncer. We could recommend this in the .d.ts docs...