vidarh

30 ms·

781.
by vidarh 8mo ago
The large scenes looks like about 4 screens wide? But they're not full height - looks like about 2/3, so let's say ~24KB total including color data. I don't think it should be a problem. The walk + scroll is slow enough
782.
by vidarh 8mo ago
For a game like that, while I agree with the Amiga version looks good, frankly the Amiga port still feels like a good example of why there were lots of complaints about "lazy" ports for the Amiga that didn't take proper advan
783.
by vidarh 8mo ago
If it is fast enough, and cheap enough, people would very happily reroll specific subsets of decisions until happy, and then lock that down . And specify in more details the corner cases that it doesn't get just how you want it. Peo
784.
by vidarh 8mo ago
"brief" also has a similar meaning in English, in (several) alternative meanings of "brief" to mean a precis, summary, position paper of various kinds. Both are derived from latin breve/brevis - "short". Y
785.
by vidarh 8mo ago
I think the automation makes a significant difference though. I'm building a tool that is self-improving, and I use "building" for a reason: I've written about 5 lines of it, to recover from early failures. Other than th
786.
by vidarh 8mo ago
Too late to edit, but one minor self-correction: BigRequest changes the allowed length and format of every single protocol request. For small requests they are the same, but if length is set to 0, an extra 4 octets are inserted to allow e
787.
by vidarh 8mo ago
It's the kind of code you should expect if you don't run a harness that includes review and refactoring stages. It's by no means the best LLMs can do.
788.
by vidarh 8mo ago
> a missile that can hit something a hundred kilometres away At no point did I mention "a hundred kilometres away"
789.
by vidarh 8mo ago
That's purely convention. It doesn't need to be the case. There's no functionality that enforces or depend on that.
790.
by vidarh 8mo ago
> imposible to fix (within the confines of the existing protoco) X11's extension mechanisms can - and has - been used to enable backwards incompatible protocol changes. E.g. BigRequest changes the length and format of every single
791.
by vidarh 8mo ago
> It's a firework-grade rocket with no payload that can't even ignite reliably. Doesn't matter at that cost. > To imitate even a TBM or a MBRM, you need similar kinematics, even if you're running without a payload.
792.
by vidarh 8mo ago
100 harmful rockets for $10k is off by orders of magnitude, but that was entirely not the point. The point is that if you're in a asymmetric position where you can't do much damage directly, then whatever you can do to make the
793.
by vidarh 8mo ago
> The US has an ocean to cross. But this is exactly the point: This approach allows for insurgents or parties subject to overwhelming but expensive force to strain the logistics and budgets of their opponent. This is something that would
794.
by vidarh 8mo ago
So mount a battery on a truck, fire, move, fire again. Do that on a few hundred trucks, and drive around places without any real target and you have the other side burning through their arsenal real quick. This isn't a rocket that is
795.
by vidarh 8mo ago
They did not defend it or make excuses for it. They argued about the very specific claim of carpet bombing in Iran , before pointing out the entire campaign is illlegal and calling the choice of targets "not legally legitimate".
796.
by vidarh 8mo ago
You don't need to win any wars with it if you can use them to sow confusion, obscure the firing of more serious rockets, and/or trigger a sufficiently more expensive response. It clearly needs more work, but if an amateur can get
797.
by vidarh 8mo ago
TFA is literally about a $96 rocket. > propulsion, warheads, arming and safety, QA, traceability, climate and shelf life stability. You're entirely missing the point. These do not need to be reliable for the scenario I hinted at.
798.
by vidarh 8mo ago
They should be sweating, because if the other side can fire 100 rockets for $10k that are close enough to not immediately and obviously be off target, and you don't know whether a more expensive one with actual explosives is hiding
799.
by vidarh 8mo ago
I use Haiku to classify my mail - it's way overkill, but also doesn't require training unlike a classifer. I recieve many dozens of e-mails a day, and it's burned on average ~$3 worth of tokens per month. I'll probably
800.
by vidarh 8mo ago
Most of my complex information is text, for which TUI's work just fine. For the rest, I typically use a browser.
801.
by vidarh 8mo ago
None of the TUI's I use look like that. They all have very different structure to what a typical GUI look like. E.g. a focus on condensing more information in few text cells, and usually not displaying anything extraneous, typically ra
802.
by vidarh 8mo ago
> if we replace the real coding experience learning curve with LLMs from day one. People will learn different things. They will still learn. Most developers I've hired over the years do not know assembly. Many do not know a low-leve
803.
by vidarh 8mo ago
I really don't want my TUI's to look like GUI's rendered in low res. The appeal to me of a TUI is that it is built specifically to be a TUI, and that means eschewing complexity and detail , and favouring compact text.
804.
by vidarh 8mo ago
I made the argument multiple times that the right answer to many prompts would be a question, and it was allowed under some rare circumstances, but far too few. I suspect in part because the provider also didn't want to create an easy
805.
by vidarh 8mo ago
> Did you also notice the evolution of average developers over time? I mean, if you take code from a developer ten years ago and compare it with their output now, you can see improvement. This makes little sense to me. Yes, individual de
806.
by vidarh 8mo ago
I've spent 30 years seeing the junk many human developers deliver, so I've had 30 years to figure out how we build systems around teams to make broken output coalesce into something reliable. A lot of people just don't realis
807.
by vidarh 8mo ago
I worked on a project that did fine tuning and RLHF[1] for a major provider, and you would not believe just how utterly broken a large proportion of the prompts (from real users) were. And the project rules required practically reading tea
808.
by vidarh 8mo ago
For most cases we don't need nearly that comprehensive verification. This is expecting more off AI written code than we ever bother to subject most human written code to. There's a vast chasm there we only need to even slightly st
809.
by vidarh 8mo ago
But the predictor would need to be almost as bad as pure chance for the odds to tilt away from one box. Unless you think it is materially worse than chance, it's a bad bet.
810.
by vidarh 8mo ago
You appear to be shadow banned. Letting you know since I didn't see anything egregious on a quick scan. Maybe contact HN and plead your case. I vouched for your reply below, and to answer in the meantime: Yes, it's runtime, but th

More ›