4 ms·
that's nice, but .. wouldn't it be better to implement kitty graphics directly at this time ? https://sw.kovidgoyal.net/kitty/graphics-protocol/ https://sw.kovi
by makapuf 2y ago
that's nice, but .. wouldn't it be better to implement kitty graphics directly at this time ? https://sw.kovidgoyal.net/kitty/graphics-protocol/ https://sw.kovidgoyal.net/kitty/graphics-protocol/
- jetbalsa 2y agohow long until its just X11 with some extra steps :V
- db48x 2y agoThere are already a couple of ways to run an X server inside of a sixel–capable terminal: https://www.youtube.com/watch?v=UOTMGdUPYRo https://www.youtube.com/watch?v=UOTMGdUPYRo
- alchemist1e9 2y agoI love everything about kitty except the refusal to implement sixel. Maybe someone has an independent patch that can be applied?
- cwyers 2y agokitty is a product of a single open-source project, Sixel is a product of DEC with multiple independent implementations. DEC is out of business and there's very little risk of anyone coming out with a new version of VT333/VT340 reference manual, so Windows Terminal/conhost can have its own implementation of Sixel and ensure interoperability with other implementations as it sees fit, as opposed to committing to either chase compatibility with the Kitty project or committing to shipping someone else's code in perpetuity.
- mid-kid 2y agoKitty's standard is described for other people to implement both emulators and clients for. Since this relationship already exists, there's enough friction between emulators implementing new features, and clients actually using them (in a backwards-incompatible way), that I believe one can be relatively sure their emulator will support clients well into the future. Either way, I agree that kitty's protocol should become a versioned standard, but I also don't think that people will be very quick to change it at this point.