supposedly this required clarification

This is actually super helpful! There are multiple things this helped clarify that wasn't obvious to me: it is the Ghostty app (not libghostty), it is directly via `cat`, it is on macOS. Those are all really critical details and that helps, thank you. Is your app available (not sure what app that is) so I can reproduce these results? No need for source of course, just so I can see the relative numbers with the same inputs on the array of Apple devices I own.

It is Ghostty, the app, latest version from gh releases. Tern is in closed-beta, same as yours, but I'm open to trading invites, I'm sure it'd make us both faster in the end, which benefits everyone :) p.s. I couldn't really read your prior comment as anything but an insult, hence the reaction, but I don't honestly care much for the drama that followed it. I did praise your work just a day earlier, and have been a happy Ghostty user for multiple years now.

Thanks, when you say latest do you mean "tip" or do you mean tagged (1.3.x)? I think I saw fly by that you used tip but can't remember. Tip is dramatically faster depending on workloads. But, it doesn't matter, I can wait for your app to release and compare myself. Both apps are comfortably in the realm of "fast enough" (same thing I said for Alacritty, Kitty, etc., the enemy of speed isn't them, but iTerm2, Terminal app, etc.). As long as someone is using a "fast" terminal, I'm happy, doesn't have to be mine!! Btw Superlogical's terminal is by design slower than Ghostty. We're prioritizing other tradeoffs. I mean, I guess depending on the context, but like, if you're on a shitty LTE connection trying to view your terminal IO speed isn't the issue anymore lol.

In the prior post it was tip, as I just compared against what I ran daily (smol change adding -Dmacos-native-cpu to squeeze a bit more perf out of it) here it's tagged just to get an "official" result.

if you'd like to reproduce, please refer to the first frame ✌️

Do you have 3rd party certification for your benchmark? Did you disclose AI (artificial intelligence) usage? Have you opened a PR in libghostty yet? Please refrain from posting unauthorized benchmarks on the timeline.

I will seek the closest regulatory body to get a certification 😩

bench in public

I mean I literally told how to reproduce already, and it is using their own toolings, so the reaction I just do not understand, but can't get more clear than this

tern is so fast one might even accuse you of pulling a volkswagen /s

lol you really can't stop making yourself seem so insufferable can you.

what is bro doing

Whilst this is all very cool I'm still stumped as to why the dev world is seemingly so obsessed with terminals, terminal multiplexers, terminals within terminals and god knows what else. It's like the human computer interface is solved? I kinda disagree. Even omarchy seems like a hacker terminal setup you'd seen in a movie. 8 bit icons in file managers and what not. Yes, I know we devs live in terminals but our users don't. What is the point of a terminal that can scroll a million lines per second when there is clearly no human use for that? Is it just a dick measuring contest?

@usr_bin_roygbiv 👀👀👀

what do you think explains the performance difference? is part of it the protocol work you mentioned of 23x IO?

the clarification i need is understanding why it's faster what wizardry did u perform 👀

y’all beefing over some bullshit

@teknium pls bro 🙏👀 i think this could be a dream tern + hermes, ily4ever always but pls take a look 🚀