I was excited to learn about two new terminal emulator app which seemed to have a lot of cool new features, warp and wave. Then I looked closer and found that both are a no go for me.
Warp is closed source and you need to create an account to use your terminal. Jebus Christus, no, thanks, but no.
Wave is an Electron app. While that’s better than not having a Linux version, I’ve seen how Electron apps behave. They are the ones which hog all memory and get killed by the OS first. So that’s a no from me too.
I guess I keep my Tilix for now.
Might I add the idea that your terminal emulator must support your shell is utterly ridiculous?
https://docs.waveterm.dev/reference/faq#what-shells-does-wave-terminal-support
https://docs.warp.dev/getting-started/using-warp-with-shells
Also Wave might be FOSS but if you look at the footer in their website it says it’s backed by venture capital… how would you estimate the chances it gets closed, paywalled or otherwise enshittified?
Holy crap, no thanks. That’s legit awful.
TBH I am starting to come around to the idea of a tightly integrated shell and terminal emulator support. There are just things you cannot do with these being separate things. I am very tempted to explore the idea from the other end though - writing a shell that has a emulator built into it (like screen/tmux basically are). But I do think that this integration is needed for any per command features that is not just printing out a prompt.
It would be interesting to see what could be done with this type of integration but will likely break support for existing shells. Unless you maybe launch a shell for each command you run or something 🤔. Would like to seem more people experimenting with stuff like this and see what new things we could drive forward. We have been stuck with the current tty system since like the 80s to support devices that just dont exist anymore.
Philosophy aside, the practical issue with your terminal emulator having to support your shell is… that one does not use just one shell: what happens whenever you start a repl or an whatever program that has interactive sessions (say, for example, psql or parted)?
I can’t think of any, but I’m not the most creative person… what do you have in mind?
Having something that is like (say) tmux+fish could make sense, but only if it’s something that outweighs the lost flexibility of being able to combine <whatever shell you like> + <whatever terminal multiplexer you fancy>.
REPLs are basically shells. They behave the same way in every essential way. So the real question is support vs non-supported shells. But then that is easy - non supported shells fall back to just what a normal commands do ATM to process input/output. Other applications like TUIs are also easy to deal with as they already enter a different mode called raw mode - when a application requests that it can do what they currently do - switch to a new buffer and give full control to that one application.
Having smarter scroll back that knows the difference between a prompt/command and the output would let you do quite a few things that would be nice to have. Such as collapsing the output so you can only see the commands, keeping the command at the top of the screen even as other output scrolls off the top so you can always see what was running. Extra support for other UI elements could be nice to have as well - like tooltip support for blocks or similar.
All the shell - or really any application - needs to do is tell the terminal which bits of the output are witch. Like mark the start/end of the prompt, command and command output. Then the fallback is basically ignore the markers and print things out like it currently does.
And those are just random thoughts I have had over the last few days. These can be implemented in backwards compatible ways I believe and don’t need special support for specific shells - just needs to expand the VT100 protocols to be able to send more information between the terminal and shells/applications that are running. Much like how color, cursor positioning, entering/exiting raw mode etc are already done. Though I think some tight specialized integration might be a good start to explore these ideas before the protocols are formed.