> I disagree with this sentiment (I didn't downvote you though). Most serious terminal users are in fact concerned with millisecond latency.
SOME are concerned with millisecond latency, I agree. My point wasn't that there aren't users who care. It's that IF you care, then you shouldn't be using a multimedia terminal emulator. Putting rich content into the terminal will always add overhead. It doesn't matter if your terminal emulator is written in Electron or Rust, graphics is always going to be slower than text. So if you care about millisecond latency, then the problem with this terminal isn't the Electron part.
I also disagree with your point that most serious terminal users care about millisecond latency. A lot probably do but I'm about as serious a terminal user can get; I've literally written and use my own shell because existing shells weren't powerful enough for me yet I didn't want to leave the terminal. The reason I don't care about millisecond latency is because you literally don't notice millisecond latency. More on that below.
> Consider also that typing with 1ms additional latency per keystroke is noticable - even though it sounds counterintuitive.
1ms is literally imperceptive. There's plenty of evidence to back that up but I'll quote a SO answer:
"The 100 ms threshold was established over 30 yrs ago. See:
Card, S. K., Robertson, G. G., and Mackinlay, J. D. (1991). The information visualizer: An information workspace. Proc. ACM CHI'91 Conf. (New Orleans, LA, 28 April-2 May), 181-188.
Miller, R. B. (1968). Response time in man-computer conversational transactions. Proc. AFIPS Fall Joint Computer Conference Vol. 33, 267-277.
Myers, B. A. (1985). The importance of percent-done progress indicators for computer-human interfaces. Proc. ACM CHI'85 Conf. (San Francisco, CA, 14-18 April), 11-17."
source: https://stackoverflow.com/a/2547903
> And note the other threads here with people noting that even the VS Code terminal can get very slow.
...under some very specific workflows. I literally explained that workflow in my post as well:
"as long as the terminal reads from its PTY fast enough not to block those processes, which in all bar a very few edge cases it will be..."
I do acknowledge that Electron is going to be slower than native code. Of course it's going to be slower. The question isn't whether Electron is as performant as native code, the actual point the GP made before their edit (a very cheeky edit on their part too) was that processes run slower in an Electron terminal. Which is horsehit in all bare a few edge cases.