How clear is the desired scope of SRFI 205 at this point?

John added a ton of stty settings on top of the latest draft.

I would be interested in something at the level of the minsi library I
wrote: "just give me a raw terminal and let me read and write ANSI
sequences". If SRFI 205 wants to be more fine-grained, that is fine.

If there's wider interest in the minsi-level stuff, we can make another
SRFI out of it (with SRFI 205 as one possible implementation strategy)
and make sure the API plays nicely with SRFI 205, avoiding redundancy
and inconsistency as far as possible. I can also just keep it as a
library in snow-fort. Just wrote a Chibi wrapper for it.

On top of this stuff, we should draft the character cell SRFI, with ANSI
codes, Windows Console, and others as possible backends.

I'm interested in surveying current features of terminal emulators and
extracting a useful, widely-supported subset. Help making this survey
(or pointers to existing surveys) would be greatly appreciated.

In "what can we do with terminals today" news:

* The notcurses library is the cutting edge in terminal ricing.
Demoscene vibes: https://www.youtube.com/watch?v=b4lmMADP1lA. Notcurses
is used for the s7 REPL in git master.

* Sixel graphics are making a comeback and there are some libraries to
help convert image files to sixels. Here
https://www.youtube.com/watch?v=0SasrQ7pnbA a hardware VT330 takes two
minutes to draw a snake photo.

* Gnuplot vector graphics, also on a hardware VT330:
https://www.youtube.com/watch?v=afNuDH7QpYA