As for the underlying question terminal emulators do seem to be REALLY hard. I don't think this has to do with language/OS complexity as much as the complexity needed to be compatible with apps that expect the terminal to work a certain way rather than just being an easy way to display text on a screen.
Would be curious to hear from people that actually work on terminal on what is it that makes them more complicated than they look if any are reading this thread though.
Terminal support in linux is built directly into the kernel. So bar minimum you need to be familiar with these syscalls, which are a little obscures, but luckily only a hand full are necessary.
Next, youll need to understand process management in the kernel, standard pipes, and job control.
The actual emulator code has to keep a mapping of terminal control characters to terminal features, and keep this working across OS's if portability is a goal. Here in lies the devil in the details. This emulation layer is tribal knowledge.
Finally you have to also be a good UI dev, who knows how to inter-op ASCII and UTF8 shell escaping.
This isnt touching on colors, GPU acceleration, and other pluses.
I think that’s the point being made here, that taken at face value it’s just ridicules that what is on the user-side minor improvements in usability over the status quo 20 years ago is still considered “an advanced feature”.
But you could say the same thing of browsers.
A terminal on the other hand - it literally has to just print text to a screen and he's right that a bunch of hype and over-engineering of a solution for it is silly.
PCs have been outputting text to screen faster than a human can process for over 60 years now (and bonus: using a lot less resources than today's Windows Console of the month).
