esafak

Apparently the plan is to move the desktop from Electron to Rune: https://github.com/anomalyco/rune/blob/dev/PLAN.md

This is interesting. Are you using an available product for managing these signatures, or did you write your own?

You're not shifting left unless you guarantee local tests always run, faithfully to production. That rarely happens. Local testing should be supplementary. We use https://hk.jdx.dev/

They ain't running in CI.

I was thinking "Builds and tests with a fast-moving Rust code base like this must be painful", then I noticed the total absence of tests: https://github.com/anomalyco/rune/commits/dev/

It presumably powers OpenCode, the agent harness.

I doubt AI labs have beating Pangram as an optimization goal.

And the sampling theorem only applies when you only know that the signal is band-limited. If you know more than that, you can go below the Nyquist rate.

My first reaction: this is weird; agentic workloads have many problems but orchestration speed ain't one of them. You spend most of the time waiting for the model! Is this for small models?

I would have lead with the durability aspect.

This is fragmentation. The advent of slop has caused many repos to stop accepting contributions, and people maintaining their own forks. I do it too, diligently maintaining repos that the original maintainers were unwilling to, or simply to add my desired features.

Since code is now cheap I think it is not the end of the world, but it does suck a bit for users, who may need to cobble together multiple forks.

I use OC with various subs, and am testing DeepSeek Harness, JetBrains Thinkrail, and Volt. Extending the harness with plugins is my main interest these days.

Sorry to be pedantic, but it's iteratively. Your README even calls it Iteration.

Thanks for sharing.

Not cost competitive. The performance difference is not commensurate with the price.

   Model          | Cached Input | Input (Miss) | Output
   ---------------|--------------|--------------|----------
   Tencent Hy4    | $0.042       | $0.83        | $2.50
   DeepSeek (peak)| $0.006       | $0.30        | $1.20
   DeepSeek (off) | $0.003       | $0.15        | $0.60

There are two more incidents:

https://www.githubstatus.com/incidents/xr3myzj7ms51

https://www.githubstatus.com/incidents/h02n9572mx9y

Get your act together, clowns.

Schmidthuber's rejoinder https://people.idsia.ch/~juergen/fast-weight-programmer-1991...

Applicant Tracking System

Looks interesting. Actively developed : https://github.com/githwxi/ATS-Postiats

edit: You are right, I did not know there was a third version. If only this thing had release notes. I have no idea the current state of the language is. Hongwei, if you're listening!

If this language is going to be a thing it needs community development.

RSI is simply what you get when human involvement tends to zero. Once they set up the training loop you get RSI. I am strongly opposed to this.

The universe presents an endless amount of training data. Some if it, like maths, is abstract; you can synthesize the data.

Tough crowd today!! In lieu of my usual critiques I'll just wish you luck.

This is the most shocking AI paper I've read in a long time. It did not even occur to me to ask the question. Cray cray stuff that opens up a whole new area of R&D.

Does this open the possibility of better decoding biological brains?

Can you add README sections for the schema and supported types?

We need token monitor apps to reveal the less biased stats.

Same price as DS4.1F: https://longcat.ai/platform/docs/pricing/longcat-2.5

On benchmarks they seem comparable. The wild cards are the cache hit rate and speed.

> PostgreSQL today; MySQL and SQLite are planned.

The database support is actually the most important part. Are the dialects plugins? If not I would make them so, so users can add support for their own database.

I would feel more comfortable if the agent wrote the software to do the surgery rather than did it itself, so that it can be vetted before deployment. And if a complication arose, to stop and defer to the surgeon (would that work?)

If only we had accurate models they could train on.

I didn't say they did. I said companies writing those job descriptions does not establish causality. You've got it reversed.

If you are working on a typical web app, you either use a canned framework or write your own. Models are trained on existing frameworks, so they don't have to hold it all in their context to reason about it; it's baked into the weights. Same goes for the humans reviewing the code. I don't see the benefit of writing your own framework for most companies.

Why wouldn't they be?? That says nothing about who's pushing it. Companies ask for it because developers use it. That's what makes hiring easier.

Go on, you're not saying anything yet.