It's useful when you either don't want to use Microsoft Copilot, want to use a local model + harness for security reason, or want to use an agent that already has access to personal data.

I've tested it with some harnesses (pi, opencode, codex) but I guess that it should work on many others.

Not having a Microsoft developer account, I've yet to publish it to the marketplace, but it's sideloadable.

While the first prototype was written by me, I've only kept the core and let an LLM rewrite the code.

Comments URL: https://news.ycombinator.com/item?id=49790210

Points: 2

# Comments: 1

That was surprisingly easy to implement, leveraging mostly the unembedder, so I started exploring GGUFs.

First I built a simple GGUF that could compute the position of a point in a limited grid, using 4 directional tokens as input. All weights are "hand crafted", no training involved. Of course, being ollama stateless, it needs to compute everything from scratch each time from the whole conversation.

Then (with the help of coding agents) I used it as the base to build a "complete" game of snake: it can only play 57 turns (it can be fixed though) and the position of food is computed using a hash, without checking if the new position is under the snake. But it works.

Since the obvious evolution would be to run Doom on ollama, I at least got a first person roaming demo working, where the frontend only knows about the input it gives, the map comes from the GGUF.

Even if it's completely useless, it has been fun to look into how this can be done, given the obvious limitations.

Both can be played at https://ggufun.grokked.it

Comments URL: https://news.ycombinator.com/item?id=48889775

Points: 1

# Comments: 0

Apart from the fun that I got out of it, it's been there doing nothing :D

[0]: https://github.com/lucamattiazzi/magic_top_hat

the authors created a great character in rusty

So it's only working on Chromium, a way smaller attack surface than the whole Chrome users

I cannot vouch for laravel cloud or void, since I've never used them, nor I will comment on Deno/Bun since they are far more recent

it randomly retrieves a made up startup built by combining two existing startups from ycombinator

it has access to 1000 generated startups

Comments URL: https://news.ycombinator.com/item?id=41048042

Points: 1

# Comments: 0

It's just a joke, but it proves what I was always afraid of: dunder methods are way too powerful and sometimes python seems to be broken because of it.

Comments URL: https://news.ycombinator.com/item?id=40934310

Points: 3

# Comments: 2

When a similar method was invented that applies to RNA, was of course named Southern.

When the one for proteins was invented, was named Western.