dafelst

This is exactly what I do - I run my agent inside an isolated podman container and mount the workspaces on a shared volume.

It is yet to be determined whether or not the bun rewrite is a success, though.

This is so cool, what a great project!

I remember using a single save game and soft locking myself in Space Quest 2 by not getting the glowing gem, and in Space Quest 1 I think by not picking up the shard of broken glass, and being so pissed off about having to replay.

The LSL dip and sunscreen issues you fixed were particularly egregious, by that point I feel like they should have fixed things.

While Sierra should absolutely be recognized for spawning an entire genre of games that were legitimately fun, it is crazy that they were fine with shipping games with these broken states, when it was eminently preventable.

You can read the manual all you want, but if you don't know basic algebra, trig, calculus, etc, you are not going to have any idea how to apply or use much of anything that the manual describes with regards to actually doing math with a graphing calculator. There is a base level of knowledge required.

Feels appropriate for bugs in a formal proof system:

> Beware of bugs in the above code; I have only proved it correct, not tried it.

-Knuth, 1977

Great concept for a horror film

I am very happy with my BenQ monitors, they are great value and high quality.

It's a proxy for reach, which is what GP is talking about.

Oh this is much more novel (comparatively), that makes a lot more sense. Kind of an amalgamation of a bunch of optimizations.

Thanks for sharing.

It's kind of insane that such an obvious optimization can be patented, I have to imagine that it has been invented independently dozens if not hundreds of times.

I drove on this for the first time yesterday, once from 405 S to 85th W, then back on 85th E and 405 N. Super easy, the lanes are fairly nicely signed and there is ample signage. No issues at all.

I strongly prefer the original to your edit.

You don't need to be the gold standard to make obscene amounts of money. Product quality and financial success are only loosely correlated at best.

Tightly integrated sure, but I wouldn't say tightly coupled, their framework (GPUI) is available standalone and is being used to create other apps.

https://www.gpui.rs/

FWIW Slint was founded by a group of long time Qt alumni, so brings a lot of that know-how into the space.

I have not, what is it?

Saunas are very cheap to buy and/or build, certainly within the budget of an average research grant.

More victims of AI.

Not actually of "AI is replacing jobs", more "oh shit we are spending too much and the product isn't good enough for us to ever make a return on our absurd over-investment".

There is good, useful content in this article, but it is seriously overshadowed by the LLM-isms indicating that a nontrivial part of it is AI generated.

I'm obviously channeling my inner boomer here, but as soon as I start seeing the tells of AI authoring, I just give up on the article altogether. To the author, please consider that I and many others want to see your thought process, warts and all, there is no need to hide behind the facade of the LLM screed.

But guys, what you don't understand is that the code IS the contract!!! That means you don't even NEED regulation!!

Why are people running AI bots to make comments on HN?

Yeah, the concept of "nemawashi" (根回し) is very important there, this idea that all the groundwork and decision making is agreed upon before the meeting happens.

The term literally comes from the concept of "preparing the roots", that is, the process of softening the ground and trimming around the roots of a tree (often a bonsai) in preparation for moving it safely.

Trump is too useful to Ellison right now, he isn't going to derail that gravy train over a few tens of millions of dollars.

TACO

The term "agent" with regards to law enforcement substantially predates "agent" in the context of AI.

High capacity, super reliable box that you could run your entire business stack on, if you could afford it.

Sure, if the memory error is an immediately crashing one like a null per deref, but if is (for example) a memory corruption (e.g. an out of bounds write or a write-after-free) then this would be super helpful in exposing where those are happening at the source.

So you're saying that just because they can't pay everyone, they should pay no one?

The big US credit score trio, Transunion, Equifax and Experian, have all had multiple, massive data leaks. This is not very good at all.