I used to use termite, but switched to kitty for ... reasons I can't remember now.
Does alacritty support ligatures yet? I've been using kitty because it does, but then I have issues with tmux in kitty ... I know, I want my cake and eat it
Does alacritty support ligatures yet? I've been using kitty because it does, but then I have issues with tmux in kitty ... I know, I want my cake and eat it
1) If you scale certain fonts using fontconfig, alacritty would ignore that.
2) No emoji colors, no ligatures (and thus no flags), also emoji size has problems
... and not to mention it's one of the slowest (yes! contrary to what they blowhard they response time to keys is very large) terminals out there.
List is endless, termite had its place, but there is no doubt that there is so much ego in OSS that one solution canot fit all
Personally I'm keeping an eye out for wezterm, it seems all the good things of alacritty without the self-congratulation part
They state explicitly on their README that they're measuring throughput, not latency. Besides that, all three of your complaints are true as far as I'm aware, and they are major dealbreakers.
Colored emoji are working fine in my alacritty terminal. Here's my fontconfig file: https://github.com/dbrgn/dotfiles/blob/master/fonts.conf
This is intriguing to me. You're not the first I've seen complaining about this, so there must be something to it.
But in my case, on a fairly low-end machine, what I type shows up instantaneously[0] on the screen, even when the CPU is busy compiling.
I'm running an i5-6500 with integrated hd 530 graphics and a 4K screen. I use i3 (so X11) and Picom with a bunch of effects.
---
[0] I have a mechanical keyboard, and if I push the key all the way down, I hear the sound of the key bottoming out after I see the character on the screen.
In these kinds of discussions, the talk is usually about perceived latency not absolute latency. If it feels like there's basically none, then it's good enough. It doesn't mean there's absolutely 0.0 ns latency.
> Different people have different sensitivity to latency
This I can get behind. But I doubt that what seems instantaneous to me would seem "terribly long a time" to someone else, especially since I look specifically for this. As you pointed out, no terminal app has absolute 0 latency, so if one particular software has terrible latency compared to another one, the difference would have to be pretty huge.
Do note that I'm not talking about being bothered by the latency. I know a lot of people who will notice the latency when pointed out to them, but they don't really care. Personally, I hate when things lag, even if I may happen to have a worse perception than average (don't actually know if that's the case).
All terminals have latency, but that doesn't mean that all terminals have equal latency. Point was that anecdotes about subjective instantaneity are pretty meaningless, especially without any reference point.
It would be much more meaningful to substantiate the discussion with actual data; the measurements done by danluu are one example https://danluu.com/term-latency/
Of course there can be quibbles about the specific methodology, but its still much better than having no data at all. Especially for something that is as measurable as latency is.
In general if you claim something is fast, back that claim up with some numbers.
Not really a surprise but something to be aware of.
cat randomfile
and see the output correctly, at least.
I switched from kitty to wezterm. I don't have an official benchmark like the below [1], but low latency is an explicit goal of the project and it seems on par with kitty.