In TUI, you almost always open a new process when your start a new instance.

Pretty much all GUI frameworks I have worked with behave the same, and you have to actively create singleton behavior.

Then you have Mac OS which makes it hard two create multiple processes of a GUI app. Other OSs don't necessarily behave the same.

TUI developers could implement the same kind of anti-features. This isn't a trait of GUI vs TUI.

Why would this be true? The way "single instance mode" usually works is by taking a global mutex or making a lock file. What makes this not possible for TUI apps? Isn't it just culture?

Opening more than one instance is the default, developers have to actively work to block it. At least on desktop.

Only if the program is designed to be safe to run in parallel. Which can be done with desktop apps too - and they'll run more efficiently as well, as the windows share memory.

On the contrary, multiple windows is normal and the default but tabbed interfaces became popular due to bloated GUI design. This is all down to design and not due to it being a GUI, though.

The other thing is, I've had "computer use" on my LLMs since way before they had vision and macos integration. Since in TUIs both control and data is text, they can simply send text to a pty.

There is an opencode plugin for this opencode-pty.

Terminals + browsers is all I want for most apps.

Native GTK/QT is really good, but I only want it for "system" things like a clipboard manager or file manager settings app and things like that.

I hate that you can only have one window of the Windows Settings app now.

But in the article all these GUIs are vibe coded. Tell your AI to allow more than one window. What's the problem?