Hacker News: yencabulator

New comment by yencabulator in "Fire at Starlink site in Bulgaria after suspected sabotage in Poland"

yencabulator — Wed, 14 Oct 2026 20:51:15 +0000

> Starlink, the commercial product, was always off limits for military operations (the type that require connectivity at 100+ MPH).

Lots of Ukrainian use of satellite communications is where people are hiding in a makeshift bunker. Those don't usually travel at 100+ MPH.

New comment by yencabulator in "Fire at Starlink site in Bulgaria after suspected sabotage in Poland"

yencabulator — Wed, 14 Oct 2026 20:47:34 +0000

The people who started the fire are not going to be FSB employees, they're useful idiots that got paid for some vandalism who won't even know who hired them. Plausible deniability.

New comment by yencabulator in "Passkeys are a trap"

yencabulator — Wed, 14 Oct 2026 18:37:10 +0000

> sharply restricting -if not preventing- freedom to sync the attested credential

They already did that! That was the first round! It is only through community outcry that "Credential Exchange Format (CXF)" came to existence.

https://github.com/keepassxreboot/keepassxc/issues/10407

This one is specifically about 1Password: https://news.ycombinator.com/item?id=37836783

There was a similar thing about what Google did with Android (basically making only the built-in manager viable, for a while), but I can't find a good link.

New comment by yencabulator in "Google told to halt work on datacentres in Finland over environmental concerns"

yencabulator — Wed, 14 Oct 2026 16:34:21 +0000

There's huge areas of Finland that are very much protected (e.g. you can only walk on two trails in the whole preserve). Do an image search for Kevo trail to get an idea.

https://en.wikipedia.org/wiki/Protected_areas_of_Finland

New comment by yencabulator in "Google told to halt work on datacentres in Finland over environmental concerns"

yencabulator — Wed, 14 Oct 2026 16:25:16 +0000

Finland is the second lowest-corruption country in the world.

https://www.transparency.org/en/cpi/2025

New comment by yencabulator in "Fingerprints are the new cookies"

yencabulator — Wed, 14 Oct 2026 16:21:09 +0000

The cookie banners are intentionally designed so that they cause maximum moan and groan in the people who care, and the maximum naive allowance in people who don't.

New comment by yencabulator in "Hackers activate automatic defrosting in supermarket fridges across Europe"

yencabulator — Wed, 14 Oct 2026 15:58:04 +0000

> not doing a strong action that would be discussed at a diplomatic level.

That's exactly what's happening.

https://brusselsmorning.com/von-der-leyen-proposes-european-...

New comment by yencabulator in "Americans Trust in Federal Government Sinks to New Depths"

yencabulator — Wed, 14 Oct 2026 15:23:17 +0000

Supreme Court voting 5-4 means it was very close to being "legally sound" the opposite way. Not very sound, then.

New comment by yencabulator in "US imposes sanctions on ICC hours after former judge wins Nobel Peace Prize"

yencabulator — Tue, 13 Oct 2026 23:10:59 +0000

Nurnberg style trials are for the humans behind the actions, not for countries.

New comment by yencabulator in "OpenAI engineer arrested in bugging of executive conference room"

yencabulator — Tue, 13 Oct 2026 22:40:31 +0000

HR formed a team -- you might say a squad -- to fire them.

New comment by yencabulator in "11 year old unsoundness bugs found in Rust borrow checker"

yencabulator — Tue, 13 Oct 2026 22:05:00 +0000

Even better, configuring the built-in static analysis to forbid it is a one-liner.

New comment by yencabulator in "11 year old unsoundness bugs found in Rust borrow checker"

yencabulator — Tue, 13 Oct 2026 22:04:02 +0000

On a typical Linux box, overcommit and eager or confused OOM killer will ruin your day anyway.

For no_std style work where running out of memory is actually a hard limit, you have two choices:

1. disable the global allocator and use APIs that take Allocator, to force yourself to only consume APIs that can return errors from allocation. (The ergonomics aren't always great, and fattening all types with a pointer to Allocator isn't great; might be able to use a zero-sized Allocator, with trickery. Say goodbye to many libraries.)

2. use static analysis and try to avoid ever hitting a code path like Vec::with_capacity that doesn't return a Result. Make sure the tooling also analyses your dependencies. (Tooling for the static analysis is painful. And still, say goodbye to many libraries.)

I do wish there were more conveniences for alloc-can-fail APIs, like the failable version of Vec::push needs both try_reserve and push_within_capacity (nightly only).

New comment by yencabulator in "11 year old unsoundness bugs found in Rust borrow checker"

yencabulator — Tue, 13 Oct 2026 21:53:59 +0000

Sure, but it's worth acknowledging that that specific wart is being fixed.

https://doc.rust-lang.org/std/alloc/trait.Allocator.html

New comment by yencabulator in "11 year old unsoundness bugs found in Rust borrow checker"

yencabulator — Tue, 13 Oct 2026 21:39:01 +0000

No, people who don't actually use Rust just like to pretend it's something more than it is, for some performative outrage or something.

Here's the current list of similar (unsoundness-causing) open compiler bugs in Rust:

https://github.com/rust-lang/rust/issues?q=state%3Aopen%20la...

New comment by yencabulator in "What is Codemode"

yencabulator — Sun, 11 Oct 2026 22:46:20 +0000

The sandbox is the point.

New comment by yencabulator in "Making Git Hooks Visible"

yencabulator — Sun, 11 Oct 2026 21:56:36 +0000

Meanwhile, in JJ, per-repo config is also stored in ~/.config, to avoid this attack:

https://www.jj-vcs.dev/latest/design/secure-config/

New comment by yencabulator in "MXC - a sandboxed code execution system"

yencabulator — Sun, 11 Oct 2026 20:55:17 +0000

> vm can only connect to an IP if it came from a DNS lookup of an allowed name

That sounds like more than just the named destination is reachable, if an allowed name uses that same cloud load balancer/CDN.

New comment by yencabulator in "New gTLD Application for .lan"

yencabulator — Sun, 11 Oct 2026 17:00:25 +0000

Meanwhile, I watched a very prominent Microsoft consultancy shop install a client's ActiveDirectory as .ad. You know, the country of Andorra? That has an actual real ccTLD?

New comment by yencabulator in "New gTLD Application for .lan"

yencabulator — Sun, 11 Oct 2026 16:55:37 +0000

They should finally set aside .home and .corp as reserved domains. They're already considered not a valid gTLD:

https://en.wikipedia.org/wiki/.home

https://icannwiki.org/Name_Collision

New comment by yencabulator in "New gTLD Application for .lan"

yencabulator — Sun, 11 Oct 2026 16:50:35 +0000

Can we somehow, as a society, eject all lawyers?

> AGB Q118. What is the meaning/definition of the applied-for gTLD string?

> Lan commonly refers to a broadly recognized term used across a wide range of contexts.

> 1. Describe the mission and purpose of the applied-for gTLD, including the intended registrants and users, and the related activities that have been or will be carried out to achieve this purpose.

> The mission of .LAN is to provide a namespace associated with the applied-for string.