Hacker News: TomatoCo

New comment by TomatoCo in "Artemis II safely splashes down"

TomatoCo — Sat, 11 Apr 2026 01:06:26 +0000

There should be an opposite thruster for each axis. I wonder if the short bursts were due to heating limits.

New comment by TomatoCo in "Artemis II safely splashes down"

TomatoCo — Sat, 11 Apr 2026 01:00:03 +0000

The space shuttle, too, was able to communicate. I imagine the smaller the craft the smaller the angle you can "speak" out of and, below a certain size, it just doesn't work.

New comment by TomatoCo in "A cryptography engineer's perspective on quantum computing timelines"

TomatoCo — Thu, 09 Apr 2026 02:01:24 +0000

You can't do software updates securely, but it strikes me that compromising the revocation process is a good thing. Suppose you can use a key to sign a message saying "stop using this". If someone else breaks that key and falsely signs that message, what are the downsides?

You revoke a cert because you lose control of it; if someone else can falsely revoke that cert, doesn't that truthfully send the exact same signal? That you lose control of it?

New comment by TomatoCo in "I Built an Open-World Engine for the N64 [video]"

TomatoCo — Sat, 28 Mar 2026 14:48:53 +0000

This reminds me of Magicore Anomala, a side scrolling game being made for the 1985 Atari. I wish there was a way to know how people contemporary to the release of the Atari or the N64 would react to seeing these modern engines.

New comment by TomatoCo in "We broke 92% of SHA-256 – you should start to migrate from it"

TomatoCo — Sat, 28 Mar 2026 14:37:46 +0000

Wouldn't that be the play, though? Get a buttload of bitcoin, turn it into real money, then destroy bitcoin. If you found a break in bitcoin you wouldn't rely on keeping your wealth in bitcoin and then hoping nobody else discovers it.

New comment by TomatoCo in "Don't post generated/AI-edited comments. HN is for conversation between humans."

TomatoCo — Wed, 11 Mar 2026 20:25:21 +0000

I think translation should be the only exception. It might even need to be, given how all automated translators use LLMs these days. The only alternative I see is to have people post in whatever language they're most comfortable in and then everyone else has to translate for them which just feels inefficient.

And of course, a more limited exception for posts about LLM behavior. It might be necessary for people to share prompts and outputs to discuss the topic.

New comment by TomatoCo in "Rendezvous with Rama"

TomatoCo — Mon, 09 Mar 2026 23:41:19 +0000

Conservation of angular momentum. Once everything is in it, and it's spun up, it won't stop.

New comment by TomatoCo in "UUID package coming to Go standard library"

TomatoCo — Sat, 07 Mar 2026 17:13:48 +0000

You have to generate random bytes with sufficient entropy to avoid collisions and you have to have a consistent way to serialize it to a string. There's already a standard for this, it's called UUID.

New comment by TomatoCo in "Never Bet Against x86"

TomatoCo — Sat, 07 Mar 2026 00:55:40 +0000

Now that Google and Apple have to (more-or-less) allow other app stores, I wonder if Valve is bankrolling FEX with the intent of selling games on mobile?

New comment by TomatoCo in "7zip.com Is Serving Malware"

TomatoCo — Sat, 14 Feb 2026 19:54:56 +0000

As a power user I agree, but how do you avoid it being like the Vista UAC popups? Everyone expects software to auto update these days and it's easy enough to social engineer someone into accepting.

New comment by TomatoCo in "Recoverable and Irrecoverable Decisions"

TomatoCo — Fri, 13 Feb 2026 04:20:06 +0000

As long as the wreck doesn't go past the exit they, technically, haven't missed it yet.

New comment by TomatoCo in "Autonomous cars, drones cheerfully obey prompt injection by road sign"

TomatoCo — Sat, 31 Jan 2026 22:23:51 +0000

If you put on a reflective vest they might.

New comment by TomatoCo in "Extremophile molds are invading art museums"

TomatoCo — Wed, 28 Jan 2026 22:03:03 +0000

Does neutron radiation have the same degradation? I know there's neutron embrittlement for metals but do more plastic materials suffer the same?

New comment by TomatoCo in "Is It Time for a Nordic Nuke?"

TomatoCo — Mon, 26 Jan 2026 17:12:27 +0000

Or a fleet of TELs roaming the uninhabited regions.

New comment by TomatoCo in "Oneplus phone update introduces hardware anti-rollback"

TomatoCo — Sun, 25 Jan 2026 23:53:28 +0000

Which is why I mentioned that carriers or Google might have that as a requirement for partnering with them. iPhones are rarely stolen these days because there's no resale market for them (to the detriment of third party repairs). It behooves large market players, like Google or carriers, to create the same perception for Android phones.

Thieves don't do that research to specific models. Manufacturers don't like it if their competitors' models are easy to hawk on grey markets because that means their phones get stolen, too.

New comment by TomatoCo in "Oneplus phone update introduces hardware anti-rollback"

TomatoCo — Sun, 25 Jan 2026 21:24:51 +0000

My understanding is there was a bug that let you wipe and re-enable a phone that had been disabled due to theft. This prevents a downgrade attack. It's in OnePlus's interest to make their phones less appealing for theft, or, in their interest to comply with requirements to be disableable from carriers, Google, etc.

New comment by TomatoCo in "Oneplus phone update introduces hardware anti-rollback"

TomatoCo — Sun, 25 Jan 2026 21:22:40 +0000

Lots of CPUs that have secure enclaves have a section of memory that can be written to only once. It's generally used for cryptographic keys, serials, etcetera. It's also frequently used like this.

New comment by TomatoCo in "LLM Structured Outputs Handbook"

TomatoCo — Sat, 17 Jan 2026 17:05:54 +0000

You can't do it as part of whatever's calling it because this changes the sampler. The grammar constraints what tokens the sampler is allowed to consider, only passing tokens that are valid by the grammar.

New comment by TomatoCo in "The chess bot on Delta Air Lines will destroy you (2024) [video]"

TomatoCo — Tue, 13 Jan 2026 01:45:09 +0000

Ah, the Zap Brannigan school.

New comment by TomatoCo in "Sampling at negative temperature"

TomatoCo — Sun, 11 Jan 2026 22:47:35 +0000

I'd assume that's just an optimization? Why bother sorting the entire list if you're just gonna pick the top token, linear time versus whatever your sort time is.

Having said that, of course it's only as deterministic as the hardware itself is.