Hacker News: MarsIronPI

New comment by MarsIronPI in "IPv6 traffic crosses the 50% mark"

MarsIronPI — Thu, 16 Apr 2026 12:42:55 +0000

66.6% uptime anyone?

New comment by MarsIronPI in "Anna's Archive loses $322M Spotify piracy case without a fight"

MarsIronPI — Wed, 15 Apr 2026 21:43:21 +0000

I really hate people using terms like "theft" or "piracy" to refer to the act of artificially-limited copying. Copying something does not take anything away from the person you're copying from inherently. Theft does. And piracy is like theft, but with murder and kidnapping too.

New comment by MarsIronPI in "CPUs Aren't Dead. Gemma2B Out Scored GPT-3.5 Turbo on Test That Made It Famous"

MarsIronPI — Wed, 15 Apr 2026 18:34:14 +0000

It's almost like Qwen 3.5 9B is 4 times larger.

New comment by MarsIronPI in "CPUs Aren't Dead. Gemma2B Out Scored GPT-3.5 Turbo on Test That Made It Famous"

MarsIronPI — Wed, 15 Apr 2026 18:33:32 +0000

> A weekend of focused work, Claude as pair programmer, no ML degree required

It's not caught up if you're using Claude as your pair programmer instead of the model you're touting. Gemma 4 may be equivalent to GPT-3.5 Turbo, but GPT-3.5 isn't SOTA anymore. Opus 4.5 and 4.6 are in a different league.

New comment by MarsIronPI in "Anna's Archive loses $322M Spotify piracy case without a fight"

MarsIronPI — Wed, 15 Apr 2026 17:55:13 +0000

Get the metadata from Anna. At least now it's freely available.

Also it's unfair to call Anna greedy. There can't be much money in giving stuff away for free.

New comment by MarsIronPI in "Anna's Archive loses $322M Spotify piracy case without a fight"

MarsIronPI — Wed, 15 Apr 2026 17:47:22 +0000

I'm pretty sure piracy also includes threats to phisical security by the pirates. No such thing here. Let's not dilute the meaning of the word.

New comment by MarsIronPI in "Anna's Archive loses $322M Spotify piracy case without a fight"

MarsIronPI — Wed, 15 Apr 2026 17:45:31 +0000

I've been against the UK trying to shove its regulations everywhere and I'm just as against the US doing it.

New comment by MarsIronPI in "Someone bought 30 WordPress plugins and planted a backdoor in all of them"

MarsIronPI — Tue, 14 Apr 2026 12:02:23 +0000

It may be better from a DX perspective, but it's pure pain for distros like Debian who don't want to use cargo at build time to fetch arbitrary dependencies and instead use vetted system versions.

New comment by MarsIronPI in "Someone bought 30 WordPress plugins and planted a backdoor in all of them"

MarsIronPI — Mon, 13 Apr 2026 19:51:27 +0000

> I used to think that HN is full of enlightened open minded people who are open to correcting misconceptions if presented with new evidence, and adopting better practices.

Well, I don't think the average HNer has much of a say in how WordPress is operated, or even uses WordPress by preference.

New comment by MarsIronPI in "Someone bought 30 WordPress plugins and planted a backdoor in all of them"

MarsIronPI — Mon, 13 Apr 2026 19:49:05 +0000

Rust is like this too. Every time I open a Rust project I look at Cargo.lock and see hundreds of recursive dependencies. Compared to traditional C or C++ projects it's madness.

New comment by MarsIronPI in "MiniMax M2.7 Is Now Open Source"

MarsIronPI — Mon, 13 Apr 2026 11:09:06 +0000

I disagree. Fine-tuning, while useful, feels more like patching executables than source code. Besides, just because most people don't compile e.g. Android for themselves doesn't mean that Android should only be distributed in binary form.

New comment by MarsIronPI in "Google removes "Doki Doki Literature Club" from Google Play"

MarsIronPI — Mon, 13 Apr 2026 01:11:33 +0000

Seems like we should not be centralizing control like this.

Take control of your computing, user.

New comment by MarsIronPI in "MiniMax M2.7 Is Now Open Source"

MarsIronPI — Sun, 12 Apr 2026 11:51:43 +0000

Even the MIT-licensed weights are just that: open weights. Let's not call the weights "source", because they're emphatically not. I can't retrain Qwen from the ground up with different pre-training algorithms, for example.

New comment by MarsIronPI in "AI assistance when contributing to the Linux kernel"

MarsIronPI — Sat, 11 Apr 2026 13:06:16 +0000

If consent to use of your code in AI training can be revoked at any time, that makes training impossible, since if anyone ever withdraws consent, it's not like you can just take out their work from your finished model.

New comment by MarsIronPI in "Iran demands Bitcoin fees for ships passing Hormuz during ceasefire"

MarsIronPI — Thu, 09 Apr 2026 12:55:55 +0000

https://news.ycombinator.com/newsguidelines.html

New comment by MarsIronPI in "Iran demands Bitcoin fees for ships passing Hormuz during ceasefire"

MarsIronPI — Wed, 08 Apr 2026 18:55:21 +0000

> IMHO don't see how Israel-US can politically survive those two scenarios.

Is that such a bad thing?

New comment by MarsIronPI in "Battle for Wesnoth: open-source, turn-based strategy game"

MarsIronPI — Tue, 07 Apr 2026 13:10:43 +0000

Essentially it makes it so that if your attack is 4 swings attacking a unit on 50% defense territory, it will hit twice and miss twice every time. Any remainders are dealt with randomly, but the seed is based on the save, so it removes save-scumming (which has always tempted me, and I don't like).

New comment by MarsIronPI in "Battle for Wesnoth: open-source, turn-based strategy game"

MarsIronPI — Tue, 07 Apr 2026 12:05:56 +0000

Regenerating? Luanti doesn't ever regenerate terrain. Do you mean generating the first time? Or loading? I've had it happen that I run into the edge of the mapblock and can't walk into the next one because it's loading or still generating. Maybe give it another chance, because I can't picture the behavior you're describing.

New comment by MarsIronPI in "Anthropic expands partnership with Google and Broadcom for next-gen compute"

MarsIronPI — Tue, 07 Apr 2026 02:36:44 +0000

We're already seeing this with Qwen 3.5 and Gemma 4. They're better than GPT-3.5 and they run on smartphones and old laptops.

New comment by MarsIronPI in "Miracle – A hackable window manager for Linux"

MarsIronPI — Tue, 07 Apr 2026 02:33:30 +0000

Huh. I'm all for hackable software, but why WASM instead of something more introspective? Dare I say... Lisp?

Ok, ok. I concede the point: using WASM as a generic ABI is a neat idea.