Hi,
On Fri, Dec 26, 2025 at 01:25:24PM +0100, Jochen Sprickerhof wrote:
> Hi Hiltjo,
>
> * Hiltjo Posthuma <[email protected]> [2025-12-26 13:00]:
> > I'm not sure, because st does not support all the options (double
> > underscore,
> > curly underscore, etc).
>
> st does not use the defined styles but shows all options differently as
> defined in:
>
> https://git.suckless.org/st/file/st.c.html#l1372
This is a different thing. When you have several parameters in a
CSI command then you apply every one of them to the command:
CSI 3;4 m (SGR 3;4)
is equivalent to
CSI 3 m (SGR 3)
CSI 4 m (SGR 4)
but it seems like in this specific case they are hacking the CSI parsing
and adding a special case for CSI 4 m where it parsers the next parameter
as a special parameter for it. The link [1] that you provided says:
This is done by re-purposing some SGR escape codes that are
not used in modern terminals
I would like to know why the author considers that bold or italic
are not used in modern terminals, but this extension is breaking
how sequences are used. If we add this feature I would use a different
SGR number or even a different command. But adding this terminfo entry
just as SGR 4 can create a lot of confusion to the programs using it
and expecting the other SGR values not being modified.
[1] https://sw.kovidgoyal.net/kitty/underlines/
Regards,
