Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

I built a GPU back end for Emacs

en.andros.dev

Post reply on HN
Print view View original

81–90 of 117 posts

Re: I built a GPU back end for Emacs

#81

Post by jadyoyster »

wrs wrote: post #59
arikrahman wrote: post #27

Aside from the wonderful contribution to Emacs, I have the utmost respect for how straight-forward you were with the 100% LLM generated code. The enlightening conversation on GPU freedom that ensued was also informative.

This is a way of using LLMs that doesn’t seem appreciated enough: prototyping a refactor or other new approach to see how it goes. Even if you then rewrite it from scratch, it’s a big timesaver because you’re using the new knowledge to do the “right thing” and avoid backtracking.

This is exactly how I use LLMs. "I wonder if it's possible to ...", and if I get something that seems to work or passes the test suite that gives me lots of added confidence that an approach is worth spending proper engineering time on.

Or sometimes you get suggested a method or library you've never heard of, opening your horizons.

Re: I built a GPU back end for Emacs

#82

Post by d-us-vb »

jwr wrote: post #42

Emacs user for >32 years now. It's a pity this won't get merged. There is a big usability/accessibility factor to consider here: I really wish I could have something like the Ghostty cursor_blaze.glsl shader for highlighting where the cursor is when you switch windows/buffers/apps. Most people think GPU equals silly toys like video in a text window, but there is much more to it than that. [and yes, I know about beaco…

I don't think video in a text window is a silly toy. Sounds like another useful tool in the toolbox.

Re: I built a GPU back end for Emacs

#84

Post by slashdave »

andros wrote: post #12
ghosty141 wrote: post #9

Earlier quoted context omitted.

Very much this. I really dislike the somewhat corporate and characterless sounding tone of AI. I really liked the part where he tried to get it into upstream emacs, most of these type of projects never even get there. But yeah things like "bring honest numbers" sounds just weird.

Please, have some mercy, English is not my first language! Thanks for the advice, I'll keep it in mind.

Ignore the detractors, I, for one, enjoyed reading it

Re: I built a GPU back end for Emacs

#85

Post by stackghost »

salutis wrote: post #79
stackghost wrote: post #50

Earlier quoted context omitted.

>an ideological position on "GPUs" as a category is bizarre even by Late Soviet FSF standards. The FSF and the GNU project are both paralyzed by their inability to move on from Stallman. He may have been a visionary 40 years ago but now he's an obsolete dinosaur who hasn't written a line of code in decades and has absolutely no idea how modern computers work. He can't update his own website. He evidently doesn't seem…

Please do not spread lies. "obsolete dinosaur who hasn't written a line of code in decades" The most recent code change in Emacs by RMS was on 2026-04-22 (0fb9d096e38), so ~3 months ago. He can write complex code (without AI), such as the `cond*` macro (707 LOC, all macro code), authored and pushed to Emacs on 2024-08-02 (18491f48d97).

I stand corrected on the "writing lines of code in decades" portion but not on the rest.

I think RMS does far more harm to the movement than he does good, and this ignorant, luddite take on GPUs is a poignant example.

Re: I built a GPU back end for Emacs

#86

Post by panza »

sph wrote: post #8

https://github.com/tanrax/emacs-gpu/blob/main/.github/assets... This massive speed-up on 4K screens makes me want to try it. The wayland pgtk version has such terrible latency I have to use the X11 build to avoid gnashing teeth during my working hours. And I think it's the X11 version that uses cairo, so the actual speedup in my case might be even larger. I reported the issue years ago, the pgtk maintainer confirmed,…

My experience with Emacs, 4K and Wayland was so abysmal that I spent a weekend making my config fully compatible with a terminal.

Not ideal, but I'll keep running it through Ghostty until it's fixed.

Re: I built a GPU back end for Emacs

#87

Post by golly_ned »

jwr wrote: post #42

Emacs user for >32 years now. It's a pity this won't get merged. There is a big usability/accessibility factor to consider here: I really wish I could have something like the Ghostty cursor_blaze.glsl shader for highlighting where the cursor is when you switch windows/buffers/apps. Most people think GPU equals silly toys like video in a text window, but there is much more to it than that. [and yes, I know about beaco…

I'd be thrilled with a second thread much less GPU.

Re: I built a GPU back end for Emacs

#88

Post by tmtvl »

_f9cu wrote: post #55

What is a backend for GPU?

A backend is a part of a program which receives commands from the frontend (the user-facing part of the program) and performs some actions based on the input received before sending the result back to the frontend.

A GPU (graphics processing unit) is a piece of hardware which is very good at doing a lot of calculations with a lot of numbers very quickly.

Making a backend for the GPU means programming the backend in such a way that it can run on the GPU, thereby (hopefully) getting the program to run faster and more efficiently.

Re: I built a GPU back end for Emacs

#89

Post by agentultra »

joshjob42 wrote: post #48
drob518 wrote: post #10

Well done. Here’s hoping that a hand-written GPU backend gets developed based on this wonderful proof of concept. There’s no reason to not take advantage of the state of the art hardware when it’s available. And screens are only moving toward 4k and higher (6k or 8k).

Or that GNU updates a policy that will very rapidly go from probably net silly/mildly contestedly useful to completely ridiculous in a year or two. Not allowing LLM code will be basically turning down the work of the worlds best programmers running at 50x speed in couple years, and will functionally doom any software project that enforces such a policy.

Doubt that. If you couldn’t write it yourself you can’t get an LLM to do it for you. So unless you’re attracting the 50 best software developers you’re getting middling to miserable results.

Re: I built a GPU back end for Emacs

#90

Post by stebalien »

zeendo wrote: post #67
andros wrote: post #20

Earlier quoted context omitted.

Good news on the Wayland front: there is already a working branch (wayland-pgtk-backend) that adds a PGTK binding on top of the same EGL/GLES driver, pending merge to main. If you want to try it before that, the branch is available on the repo.

I can't find that branch here - https://cgit.git.savannah.gnu.org/cgit/emacs.git Am I looking in the wrong place?

https://github.com/tanrax/emacs-gpu/tree/wayland-pgtk-backen...

The main emacs-gpu branch doesn't have wayland support.

Post reply on HN