A third way of using Linux

As far as I remember, when I wrote my email reader/writer in C, ncurses was not even a thing, so I made my own windows and dialogs in text mode with VT-something codes or something like that. That was not on Linux, but on DOS at the time, but anyway. I did not wait for graphical stuff to be common.

I used my software for months to read and write mails, interact in mailing lists, newsgroups, etc.

Funny thing is when a sysop sent me a mail once: basically, he said to me that one of the machines he had the responsibility of back then warned him about a non-conform header. He asked me the name of the email software I was using to report the bug to the author. I thanked him for my first report ever, for his diligency, etc. and of course, I fixed it promptly ;)

The article has it backwards: GUIs are the third way of using Linux. TUIs are the 2nd way. I mean, stuff like pico, nano, vi, Emacs - these are all TUIs.

Perhaps it's because I grew up on DOS that I think that way, but one of the reasons I switched from Windows to Linux decades ago was to use TUIs.

I stick to CLI's and TUI's and use zellij as my window manager. This way the configs and muscle memory I develop on my work macbook are transferrable to the linux box I use for personal projects, and visa versa.

It seemed a bit limiting at first, but after adapting to it I feel like it was the right move. GUI's represent a sort of explosion in the set of valid states for software to be in, and I find that software which opts out of trying to support that explosion tends to work better, likely because the maintainer just has less to worry about.

What a ridiculous article. TUI is GUI with fixed cells, that's it.

Emacs text interface, Emacs GUI interface, and ... sorry what's the third one? :-)

Looking at tuios, using a mouse in a TUI completely defeats its purpose, for me personally.

Other ways I often use:

- Using a web browser to connect to a web app running in a Linux VM. I write all my custom tools that way.

- Using VS Code to connect remotely, using ssh

I'm exploring aspects of these "other ways". I've been pushing Kitty terminal graphics compositing and then layering libraries over that. I've also been working on remoting Golang programs over the network as well as compiling them to WASM and embedding that in browsers. This WebGPU browser shader is an example of that mashed together [1]. Same code for native and browser.

Shout out to TUIOS [2] that is in the article; author has a lot of neat libraries and his `sip` protocol is the basis of that remoting work.

[1] https://nimblemarkets.github.io/ntcharts/demos/shaders/

[2] https://github.com/Gaurav-Gosain/tuios

Like the days of Norton Commander

I was hoping the third way would be somehow related to a type of hands-free neuralink so I can code at the speed of thought with my eyes closed laying in bed. Someday the tech will exist but, by then I'm not sure coding will be a thing to do anymore. I'll probably still do it anyway though.

> What gives you the benefit of a visual interface, without the bloat that comes with Wayland or X11?

fbdev?

Early Reuters and Bloomberg terminals had a text-centric aesthetic.

AFAIK on a PC, when booted into text-only mode, you are still using a bit-mapped framebuffer. From my experiments you can mix graphics and text.

Well, if we count claude as tui.

Ah yes, website talks about "the bloat of x11 and Wayland", and doesn't even scroll correctly on Firefox mobile likely due to some JavaScript fuckery.

Practice what you preach. And practice first.

Isn’t that what resolution aware ttys and screen accomplish?