For me it’s that why do I need another application open? I have my coding agents in the terminal, I should also be able to traverse and read a codebase in the terminal as well
Why do you want you coding agents to be in a terminal? A terminal is missing latex rendering, inline images, a browser showing what the agent is clicking on, being able to view a spreadsheet and then select a region to reference in the conversation, clickable links when it references a specific line number with mouse-over previews, and interactive inline visualisations. A coding agent missing any of those is a significantly worse experience, yet people are for some reason willing to give up all of them to have their coding agents run in a terminal. It's not even like you get a familiar environment to work in, nothing that you set up in your terminal carries over into your agent or IDE.
That’s just the theme matching the terminal, there are two others out of the box but I prefer the system theme. And by beautiful I mean it has a cohesive system design and theming. I prefer to to another VsCode clone
Poor wording, more so meant that when I drop into vim, vi, or emacs I have no way to (without some external knowledge) know how to open a file tree. Just a barrier to usage is all.
Okay. For what it's worth, vim tells you that you can type :help to get help every time you start it up. It would take a while until you got to the section about the file tree, though.
Yeah maybe not the best wording. It’s more meant that ideally toast can be fully used with a cursor for interactions. I don’t think you can exit vim or emacs just by clicking on a button?
You absolutely can exit Emacs by clicking a button! Even when it is running in terminal. You just need menu-bar-mode (which is enabled by default) and enabling xterm-mouse-mode (which also should be enabled by default) - https://asciinema.org/a/rprH0gLSkrUz2cqr
You can: emacs has a menu bar and you can quit emacs that way. If you write code in your .emacs.d to turn off the menu bar, then I think you can be trusted to use the keyboard to quit emacs.
You can exit Emacs by clicking a button, by making a mouse gesture, by telling it via voice command, by key-press, by modal key-chord sequence, by writing into a file, by selecting a menu item, by directly evaling (kill-emacs) command, by cookie in a file you merely visit, by a keyboard macro replayed from a register, by an abbrev whose expansion hook quits, by evaling an Org-mode source block, via a timer, via built-in solar and lunar calendars, by a shell command run inside Emacs that kills its own parent, by SIGTERM or SIGHUP from outside, by closing the last frame under the WM, by an LLM agent with an eval tool, and by pulling the power cord.
"I don't think" is never a good argument for making wide claims about anything.
Yep, I have 2 kids under 1 so this project is either built with AI or it’s not built. I put that line in because the IDE aspect doesn’t need agent management or anything. Other tools are better for that.
Then maybe sometimes it's better to take a break and not bend the natural constraints of life for once. I mean, I'm all for easing some of life with automation but it seems that a general trend is emerging where people will just keep on going no matter what, when maybe we should just stop once in a while.
I have been feeling an increasing self-imposed pressure to get more done with AI. Similar to the feeling I get when I play factorio too long. It's like the frontier of possibility has radically shifted and I feel almost guilty for not jumping into it. Tugging at me..
It would still be nice if you stated upfront that this is built with AI. The people who are interested in “no AI features” are likely to also be the ones who do not want something built with AI. It would help to be clear about the project’s stance on the matter.
Why? The source is available. That's all anyone ever needed before AI to determine whether they wanted to use it. Nothing has changed - any tool can be used to write good or bad code. The consumer should still be doing their due diligence.
I’ve been experimenting with agents a lot and have begun to think that Pi is going in the direction that a users TUI goes. If I wanted to deploy 500 agents on a single machine with minimal memory, pi wouldn’t do it. Right now I’ve confirmed it works with open router and have started piping all my tokens through it and wondering what it would be like to sit atop it.
reply