07:57 <dnkl> Arnavion: nice, thanks for testing. Will merge later today if nothing else turns up

09:13 <rapha> dnkl: good morning ... i just wanted to apologize in case i came across as pushy. the intended purpose of the questions was just to get a feeling for the complexities involved. this specific problem is sort of a big thing among people (both technically minded oes as well as mere users) in my field of studies.

09:45 <cbb> rapha: https://gitlab.freedesktop.org/terminal-wg/bidi/-/issues/1#note_114453

09:46 <cbb> that comment mentions some of the complexities

09:47 <cbb> it seems like the kind of feature that would never be quite right in a terminal

09:49 <dnkl> rapha: no worries. And fwiw, the lack of details in my replies were partly due to me not having a full and complete understanding of the problem. I decided to not dig deeper once I understood the problem was complex enough that I didn't want to support it in foot

09:59 <rapha> okay, fair enough :)

10:08 <rapha> cbb: what does the commenter mean by "cells of purely spaces"?

10:22 <rapha> dnkl: to recap for myself, supporting even the simplest use-case of RTL ("contiguous strings of RTL-script characters should be rendered in the order they would be read in, but the cursor, selections, etc. all behave as though everything was LTR") in a terminal requires introduction of high levels of completey all throughout the rendering engine (is that what you call it?) - would that be an apt

10:22 <rapha> description?

10:30 <dnkl> rapha: at the very least throughout the rendering engine. But quite possibly (likely) also requires changes to the VT parser. But yes, otherwise a correct description (to the best of my knowledge)

11:05 <rapha> okay, cool

11:06 <rapha> then i'll jot this conversation down as a note to be able to pick back up on it should i ever find the time to try my hand at an rtl-capable fork of foot :P