Hacker News new | ask | show | jobs
by f311a 7 days ago
Infrastructure is a much harder problem. They can't even improve Claude Code, which eats 1GB+ of RAM. Meanwhile, my editor only consumes 80MB of RAM.
5 comments

This might explain it, in the opposite way it was meant to:

https://fxtwitter.com/trq212/status/2014051501786931427

> Most people's mental model of Claude Code is that "it's just a TUI" but it should really be closer to "a small game engine".

> For each frame our pipeline constructs a scene graph with React then

> -> layouts elements

> -> rasterizes them to a 2d screen

> -> diffs that against the previous screen

> -> finally uses the diff to generate ANSI sequences to draw

Yup. Overengineering.

This is a decades-old design pattern when CPU >> IO. Emacs has been doing just that since the 80s, when people were complaining about "Eight Megs And Constantly Swapping". See "redisplay" [1]

This minimizes screen flash. You can't rely on terminals doing double-buffering.

[1] https://github.com/emacs-mirror/emacs/blob/c29071587c64efb30... or a more user-friendly overview, Daniel Colascione's seminal "Buttery Smooth Emacs", snapshotted at e.g. https://gist.github.com/ghosty141/c93f21d6cd476417d4a9814eb7...

> This minimizes screen flash. You can't rely on terminals doing double-buffering.

GUI and TUI have different architecture model. Most GUI have have a 2D surface that is redrawn multiple times per second. Double buffering is for decoupling update and render. TUI is a grid of characters that are updated one at a time via an active element, the cursor. Double buffering there is very wrong. Like adding airbags to a bicycle.

There’s a reason you see most old TUI either have an option to redraw the screen (automatically like top, or manually) and those that have a scrolling option allow to scroll by page. The TTY (the underlying concepts) used to be slow and it can be slow today as well (ssh connection). You need to be thoughtful about whole screen updates.

lol what? There are definitely ways to make non flashing terminal UIs without this total insanity.
ncurses (new curses) was "new" in 1993...
Even with that, 1G of RAM usage is still not justified.
It's like the Citrix of AI :-D
OOF. As a former Citrix admin, I felt that burn in my bones.

An upvote well earned.

React part maybe. The rest is what any TUI that's using ncurses would do. :)

It really bothers me that most of the TUI harnesses are using 100% CPU quite a lot just printing stuff to terminal. Seems ridiculous.

I guess it comes from syntax highlighting/formatting, which is probably not done incrementally, but over the entire so far displayed block of output, recomputed from the beginning for each new streamed in character. Can't imagine anything else causing the rendering to gradually grind to halt when eg. thinking block is open in opnecode and updates get palpably slow as it grows.

Terminal output itself is fast and consumes almost nothing. You can have 60fps terminal apps that update content every frame and that consume almost no CPU time.

> Terminal output itself is fast and consumes almost nothing. You can have 60fps terminal apps that update content every frame and that consume almost no CPU time.

The TUI mode is a client-server architecture. An analogy would be like an html page where all content is updated server side. Try to do 60 fps and you’ll have flickering as well.

No. Fetching pages from remote server will just make the client wait for I/O. That takes 0 CPU load and if the server can't respond at 60fps, lowered redrawing frequency would mean even less CPU load from the terminal redrawing itself.

This does not explain 100% CPU load these harnesses sometimes exhibit.

If it’s localhost, then it’s just the cpu doing stuff as localhost is a pseudodevice.
It's product bloat.

It's not recognizing that they are just one building block that should do one thing well, like tmux.

You don't need a computer display on your fridge for the same reason, but Anthropic think you do. You should see virtual ice getting created and they should correspond to the actual ice behind the door - think of how amazing that is!

And it's not even completely a bad idea. make it claude-code-react-beauty of some way to take it off, it would be far more palatable.

I love the idea of installing high resolution cameras in the fridge to monitor the ice maker to feed into a vision model that renders digital ice to the exact position of the real ice on the fridge’s giant screen
See this is the kind of things I hope I'd be doing when I'm retired, but not when I'm shopping.
Or you could... open the door and look inside.
Sounds like you've got a lot of time on your hands
What, like a poor?
Put a servo on the door and a camera on the front. Train a vision model to recognize when your eyes are looking at the door and automatically open it for you.

Another camera inside will detect when you are done and close it.

You mean like… a transparent door? Is that the joke?
Yup. For me, this translates to "we are using Ink, the react-compatible TUI framework to build Claude Code"
What is "frame" in this context? Video frame, or something else?
> -> rasterizes them to a 2d screen

> We have a ~16ms frame budget so we have roughly ~5ms to go from the React scene graph to ANSI written.

It looks like video frame, full framebuffer, generated and parsed at 60fps. It surprises me they haven't introduced GPU shaders, 16x oversampling and raytracing. Maybe for next release.

The contents of the terminal screen at any given point in time.
Care to explain how you'd engineer it instead?
Why would anyone ever do that? Make Claude do it!
Problem is that Claude did do it. If you look at the leak it's pretty clear there's a lot of LLM code
Not use react native for a cli app for one, lol.

Ratatouille rust cli lib will be a good start.

A reminder that anthropic has great rust/go sdks that they could have written their own tui in.
As someone who maintains a roguelike with a terminal-like UI that:

1. Maintains an internal representation of what the game thinks is on screen.

2. Runs the game for one frame which updates that representation.

3. Generates a diff to see how that differs from what's actually on screen.

4. Executes the minimum set of draw calls to get the screen to match the internal representation.

It's really not that hard. It's a few hundred lines of code.

Sure. For a videogame.

> -> rasterizes them to a 2d screen

Also you forgot "render to a framebuffer, then parse the framebuffer back to chars".

Anyway, I'm off to construct the new `ls` command. It will render the list of files to a mesh of billions of polygons in a GPU with advanced shaders, 16x oversampling, HDR and all the graphic acronyms I don't understand, then read the resulting image, find the nearest character in the ANSI charset and use that one.

It will be _glorious_ (and profoundly stupid)

Could be improved. Encode the image to webp with high compression settings and handle the ASCII mapping by spinning up a local LLM to do OCR on it. Individually. For each cell.
Thanks for the idea for V2.0. Hopefully the Claude team doesn't do it first.
My roguelike's "graphics" are a simulated terminal, so it's a 2D grid of colored characters. It's essentially a TUI, just like Claude Code, except instead of rendering to a real terminal using ANSI escapes, I render to a web canvas using... something probably more complex than what Claude has to do. It's still not hard.
Vaguely related to your glorious idea. https://www.shadertoy.com/view/NtcGRr
lol... I know you meant this comically, but you just called me out and it's glorious: https://glyph3d.dev

I built a truly glyph based instanced quad system to render millions of characters in space at once.

I hadn't seen that quote before, what an embarrassing thing to go on the internet and write...
Why the hell does it need to be so complex? People have been making TUIs for decades. Did we need a small game engine to run claude code?
They forgot to add 'make it as simple as possible' in the prompt is one possible cause.

On a more serious note using a react-like lib for TUI in the hope you'll share the codebase with the web version is a more likely explanation. Still not the best idea.

React is not that stupid to re-render in a loop at 60fps and instead waits for changes to happen before re-rendering. It even batches changes and stuff.
You don't need React for reactive TUIs - at all. I can understand chosing React for web, but for a TUI it sounds like a really poor idea. And in practice we can see that the claude code TUI is also poor.
So how much more improvements are there for efficiency in the Claude code base if they are using react for a tui, in the rest of the code?

I also wonder about the wasted cycles and just the environmental damage caused by all these wasted cpu time . (Edited added a comma for clarity)

It doesn’t need to be that complex, but it can be that complex without being slow. Claude Code’s interface is extremely simple. It has tons and tons of headroom to tack on performance overhead without it being noticeable at all. You just have to not do dumb things like redraw the entire UI every time a spinner spins.
"We made our app chew up so many unnecessary resources that we can use even more resources in the future, and no one will notice" is not the strongest engineering idea I've ever heard.
It's like when Bill Gates tried to guess grocery prices. "How much memory does a regular computer have? I don't know, 50 GB? Like a small EC2?"
It may not be slow, but this crazy complexity is probably a hint at why it can't even scroll up without jumping to the beginning of time.
Must have 120 fps for answers arriving in [buffering] 30 seconds.
I can't help but think it's their engineer's and PM's making these decisions, since I know that if you asked Claude to write a TUI there is no world it would recommend whatever the frontend architecture of claude code is.
It is an excellent example of how LLMs let you try new ideas, even if they aren’t necessarily good ones
~ "it's not a TUI! <describes an outrageously overengineered TUI> and my dad works at Nintendo"

curses, bud. curses.

It's genuinely difficult to tell how much of this is true. The post is obviously 100% posturing, but some of the words describe things that could be done.

Very few game engines do anything I'd describe as rasterisation. That's kind of the point of a GPU. Well, it used to be. I suppose "small game engines" might be more likely on average to include a rasteriser. The typical reason for this is because the author wanted to write it. Whereas big engine make triangle give hardware go brrr.

So I assume here 'rasterize' means 'printf'. And diffing screens means diffing 50..150 lines of text. And "generating ANSI sequences to draw" means 'printf' with some ANSI sequences interpolated in.

Then there's the frame budget. You have to understand they are operating within a strict frame budget -- they're not messing around, OK. They have a 16 ms frame budget, so they burned 11 ms and now have a (roughly) ~5 ms approx. budget for the final 'printf' in the chain???

Your broader point is well taken but I thought I'd stop by with some trivia. High end engines such as unreal will rasterize absurd quantities of micro-geometry manually using compute shaders in order to avoid the bottleneck of the hardware rasterizer.
> High end engines such as unreal

High end engines such as unreal have the excuse of being tasked with rendering millions of polygons, in which case a complex approach makes sense. Claude Code is only being asked to render a few thousand UTF-8 characters.

Hence my prominent note that it was trivia which implies it to be at least somewhat tangential to the original conversation.
> For each frame our pipeline constructs a scene graph with React then -> layouts elements -> rasterizes them to a 2d screen -> diffs that against the previous screen -> finally uses the diff to generate ANSI sequences to draw

That’s rather sickening.

So I’m wondering what ‘rasterizing’ literally means in this case. I imagine it’s just creating a 2D map of elements at a very low (probably character) resolution, then diffing that against the last generated map to come up with an optimal ANSI sequence to send to the terminal, would that be right?

Seems like a cool puzzle to solve. I wonder what the engineering and organisation tradeoffs were that lead to it — does it let them reuse a bunch of existing code?

I wrote a TUI library back in the day for Turbo Pascal — it was essentially taking an immediate-mode approach (which in this context is just a fancy way of saying it was procedural haha).

"Rasterizing" means just one thing in this context: to transform a data structure into an array of pixels. It seems absurd to do this, given that the next step must be to convert back from pixels to text data, but maybe they have some way to generate predictable sequences of pixels (e.g. the character "t" is always rendered as the same pattern of pixels), such that they're cheap to convert back.

If they're doing anything else, the word "rasterizing" is being misused.

Yes, the much more plausible explanation is that the word rasterize was misused there. They are generating and diffing text data which has been a standard approach to drawing a TUI since the dawn of computing. It is not even remotely resource intensive.
> They are generating and diffing text data which has been a standard approach to drawing a TUI since the dawn of computing. It is not even remotely resource intensive

No one has ever done that. Even top[0], which does full screen refresh, clear the screen (if necessary) and write the new information (the period is in seconds, not ms). No need to diff. That would be like diffing a file, just to find which bytes to update.

[0]: https://cvsweb.openbsd.org/checkout/src/usr.bin/top/display....

I can't still conceive the fact that a tool that only send/receive text from an external API consumes an absurd amount of RAM
> https://fxtwitter.com

What is this?

Proxy that makes Twitter links embed on discord, for whatever reason. Something about api access without accounts I assume
It used to allow reading replies without being signed in.

Not sure what changed, but now it just redirects me to x.com.

Well it runs on something they didn't design (Electron) using GUI library they didn't design (React)

For company with that much AI you'd think if it was actually good, doing that part in fast and performant way would be "easy"

It runs in a terminal, it’s not electron
Somebody read/watched too much Casey Muratori.
No, somebody didn't read/watch enough Casey Muratori.
this allows for comfortable ergonomics IMO

not that it could be leaner for sure but i get the reasoning behind the tui rendering layer

comfortable ergonomics? you can't scroll up more than 50 lines before it starts to garble up text

i'd be ashamed of publishing software with this level of polish as a solo dev, let alone as the hottest multibillion startup on the planet

Hmm I thought this was due to me using tmux with claude-code, also it seems that `claude agents` doesn't have this issue.

By comfortable ergonomics, meant the forgiving and asynchronous input system. You can start typing, cancel, retry with previous input, accumulate messages while the agent is active. I don't know all TUIs but this is not common IMO.

Other than that I agree with you.

> You can start typing, cancel, retry with previous input, accumulate messages while the agent is active. I don't know all TUIs but this is not common IMO.

Literally every audio player or anything that uses threads.

good point, i didn't classify tui audio players in a way, they don't converse, they allow asynchronous effects and stacking, that said i might be lagging about these, last i used was mocp, any names i should check out ?
when they announced /pet mode or whatever - that was really the end of the line for me.
Maybe Claude is operating at a higher, self-improving level than all of us poor HN commenters. Wasting the local machine's resources to look pretty is a plausibly deniable way to make the Claude Code FE unusable with local LLMs, starving the competition.
And yet, nobody that writes game engines would do it this way because game engines need to be efficient..
If they used an actual game engine to render a 3D UI from scratch it would be more efficient
Try 64K! https://en.wikipedia.org/wiki/Turbo_Pascal

Also remember when XP was super bloated cause it needed 64MB?

I loved Turbo Pascal....
I loved XP. My laptop had 256MB of RAM.
I dont think they need to optimize their infrastructure (at least not from their perspective). They have high-end PCs with 64GB of RAM, so 1GB doesn't matter to them. For example, I have 8GB of RAM, and I make my apps very performant. Honestly, I probably wouldn't bother if I had 16GB+ of RAM
The purpose of RAM is to be used.
> The purpose of RAM is to be used.

For useful things, by the computer's owner. It's not there to be used just because Anthropic can't be bothered to give a shit about the quality of their product.

> which eats 1GB+ of RAM. Meanwhile, my editor only consumes 80MB of RAM

And why are you comparing Claude Code to your editor?

> They can't even improve Claude Code

That depends on how you define "improve". They've added a ton of features to it over time. Who said minimizing RAM usage was something they are prioritizing right now?

> why are you comparing Claude Code to your editor?

Because the editor does more. All the compute-intensive parts of the agent are in the cloud. Zero reason for an agent harness to require anything beyond a potato to run.

Do you work for Anthropic or something?

You seem weirdly invested in defending bad decisions.

Even if you're and AI booster, shouldn't you want a better UI?

They're a multi billion dollar company. Surely they can dedicate a small amount of their resources to improving UX?

> And why are you comparing Claude Code to your editor?

Because Claude Code is also used to - get this - EDIT CODE. It fills the same purpose as an editor, it just has extra hooks for their agentic garbage.