15:45 <ovf> i'm sorry for continuing to go on about flags, but they just provide make for very convenient unicode test cases. :-) is is the unicode standard that says that 🏴‍☠️ occupies a single (double-width) cell despite being composed by a large number of codepoints, whereas 🇺🇳 (and others like it) seems to be taking two cells(?). (at least i'm observing the cursor moving one width at a time with the latter)

15:47 <ovf> oh, or rather it is that 🇺 (the first codepoint of the un flag) is double-width visually but the cursor moves one spot? or have i now completely confused myself about how double-width glyphs are supposed to work?

16:14 <dnkl> "normal" flags consists of two "letter" codepoints. Each of which are completely valid on their own, and has a width of 1

16:15 <dnkl> that makes flag somewhat special and unusual, and in that sense, *not* good test cases for emojis in general

16:19 <ovf> but visually the fancy letter glyphs are double-width!

16:20 <ovf> at least on my system, where i think foot is picking up twemoji (how do i check?)

16:33 <dnkl> correct, most fonts have them as double width 🤷

16:33 <dnkl> but their Unicode property is 1 cell

16:34 <dnkl> foot will log when a font is first loaded. so start a new foot instance, then paste the codepoint

16:36 <ovf> hm, no complaints (other than something something xdg toplevel icon)

16:36 <ovf> ah, sorry, i misread what you wrote. yep, twemoji.

16:38 <ovf> but anyway, this annoying discrepancy aside, i think my readline now correctly handles composite stuff (unlike gnu readline and pretty much anything else), despite not being aware of anything about unicode other than how to delimit utf8 codepoints in the byte stream

22:46 <WhyNotHugo> foot doesn't render the pirate flag as one emoji for me. It's a known issue?

22:47 <WhyNotHugo> always amused at how many *monospace* fonts have many glyphs as double-width.

22:53 <banna> hey folks, I have a bit of a question around running foot in server mode. I'm running foot 1.20.2 +pgo +ime +graphemes +toplevel-icon -assertions

22:53 <banna> my question is around how the shell is formed - I'm having a bit of an issue around the PATH

22:53 <banna> (this might be related to my login manager)

22:54 <banna> when running under systemd, would that cause a different evaluation of the login shell that foot would use?

23:27 <banna> I was actually running into something where the --server=3 flag would grumble about invalid paths, I removed that - not having done a full root-cause :\

23:27 <banna> oops

23:28 <banna> maybe not though...

23:30 <banna> I was seeing "server: failed to read socket option from passed file descriptor: Socket operation on non-socket"

23:31 <banna> more terse: "err: server.c:483: failed to read socket option from passed file descriptor: Socket operation on non-socket"

23:39 <banna> sorry for the noise -but I do not believe the PATH thing is related to foot