Hacker News: bencyoung

New comment by bencyoung in "Rust Memory Management: Ownership vs. Reference Counting"

bencyoung — Mon, 27 Apr 2026 14:30:45 +0000

Unless your memory allocator runs a form of garbage collection, which most of the advanced ones do! Worst memory performance issue I've ever seen was in a C++ program where the deallocation of a large object graph from one spot completely trashed the performance of the application across many threads...

New comment by bencyoung in "Things I've Done with AI"

bencyoung — Tue, 10 Mar 2026 09:44:50 +0000

Sounds like the Chronophage clock in Cambridge: https://en.wikipedia.org/wiki/Corpus_Clock. It it's purely mechanical but has odd pauses in the ticks etc

New comment by bencyoung in "Permission Systems for Enterprise That Scale"

bencyoung — Tue, 06 Jan 2026 10:09:35 +0000

Sorry for the delay! It's fairly simple. 1. You have a column on your objects you want secured as an LTREE[] 2. You add a GIST index on that column

The values should be the different hierarchy paths to access the object starting with a "type" e.g departments.root.deptA

When you run a query, depending on how you want to access you use a <@ query. E.g. I'm a user with root access to all depts "col <@ 'departments.root'::ltree" or I'm a user in dept A "col <@ 'departments.root.deptA'::ltree" etc

New comment by bencyoung in "Permission Systems for Enterprise That Scale"

bencyoung — Tue, 06 Jan 2026 10:03:24 +0000

Sorry for the delay! It's fairly simple.

1. You have a column on your objects you want secured as an LTREE[] 2. You add a GIST index on that column

The values should be the different hierarchy paths to access the object starting with a "type" e.g departments.root.deptA

New comment by bencyoung in "Permission Systems for Enterprise That Scale"

bencyoung — Wed, 24 Dec 2025 11:14:10 +0000

If you're using Postgres then using the ltree module is great for permission systems. Available in RDS too

New comment by bencyoung in "The state of SIMD in Rust in 2025"

bencyoung — Thu, 06 Nov 2025 10:00:00 +0000

Not majorly odd, just an area I thought Rust would be hot on when it comes to performance...

New comment by bencyoung in "The state of SIMD in Rust in 2025"

bencyoung — Thu, 06 Nov 2025 09:59:13 +0000

Having worked in HPC a fair bit I'm not a fan of autovectorization. I prefer the compiled code's performance to be "unsuprising" based on the source and to use vectors etc where I know it's needed. I think in general it's better to have linting that points out performance issues (e.g. lift this outside the loop) rather than have compilers do it automatically and make things less predictable

New comment by bencyoung in "The state of SIMD in Rust in 2025"

bencyoung — Wed, 05 Nov 2025 19:46:45 +0000

Odd that c# has a better stable SIMD story than Rust! It has both generic vector types across a range of sizes and a good set of intrinsics across most of the common instruction sets

New comment by bencyoung in "Kafka is Fast – I'll use Postgres"

bencyoung — Wed, 29 Oct 2025 15:54:20 +0000

Cheapest MSK cluster is $100 a month and can easily run a dev/uat cluster with thousands of messages a second. They go up from there but we've made a lot of use of these and they are pretty useful

New comment by bencyoung in "Kafka is Fast – I'll use Postgres"

bencyoung — Wed, 29 Oct 2025 15:20:30 +0000

Kafka is great tech, never sure why people have an issue with it. Would I use it all the time? No, but where it's useful, it's really useful, and opens up whole patterns that are hard to implement other ways

New comment by bencyoung in "JIT: So you want to be faster than an interpreter on modern CPUs"

bencyoung — Tue, 14 Oct 2025 08:26:38 +0000

JIT compilation can be faster for compiled languages too, as it allows data driven inlining and devirtualization, as well as "effective constant" propogation and runtime architecture feature detection

New comment by bencyoung in "Writing Memory Safe JIT Compilers"

bencyoung — Fri, 26 Sep 2025 13:28:03 +0000

I think this is exactly the same approach PyPy used 15 or so years ago! Partially evaluate the language runtime

New comment by bencyoung in "Perhaps my last post – we'll see (2016)"

bencyoung — Thu, 25 Sep 2025 12:25:54 +0000

I almost did a PhD with David but ended up working at Transversal instead, which was a company he co-founded to do some interesting work in the search engine space! It's what got me into software development as a career so I'm always grateful to him

New comment by bencyoung in "The underground cathedral protecting Tokyo from floods (2018)"

bencyoung — Mon, 14 Jul 2025 10:16:13 +0000

The Fens in East Anglia in the UK has a lot of interesting pumping tech. The latest can do 100m3/s (https://www.edie.net/st-germans-pumping-station-keeps-fens-f...). If all the pumps failed there would be hundreds of km2 underwater within days or weeks

New comment by bencyoung in "Destination: Jupiter"

bencyoung — Wed, 04 Jun 2025 11:32:27 +0000

I think it's more that western cultural cliches become invisible to western audiences rather than moving on. E.g. the "superhero" is definitely a western cliche. "A lone operative defies the rules to do the right thing because might is right if you're right. Individual exceptionalism triumphing etc". Somewhat shallowly examined in some films but still turns up all over the place.

New comment by bencyoung in "Run a C# file directly using dotnet run app.cs"

bencyoung — Thu, 29 May 2025 11:06:06 +0000

Dotnet is getting a fully interpreted mode in 10 or 11 so I wonder if they'll switch to that for things like this

https://github.com/dotnet/runtime/issues/112748

New comment by bencyoung in "Building my own solar power system"

bencyoung — Wed, 21 May 2025 12:21:13 +0000

We have https://www.myenergi.com/ for our car charger and it seems to be able to integrate batteries, charging and panels like you suggest, only you have to go all in. We have parts of it and are tempted to use more, but the lock-in angle is a bit off-putting

New comment by bencyoung in "Dead Stars Don’t Radiate"

bencyoung — Tue, 20 May 2025 13:01:42 +0000

Ok I drilled down a bit and looks like you are right, although I'm still not sure I've built up a clear understanding! (https://physics.stackexchange.com/questions/187917/thought-e...). In fact that question (series of onions) is exactly how I visualised it...

I'm not quite sure from that discussion why an event horizon is equivalent to a body moving outwards at the speed of light but it does make some sense. GR is always fun!

I still don't have a good idea of the "slow moving crossing the event horizon" case" but I'll read around it some more

Maybe the difference is between "free fall through an event horizon" vs "hover" (as much as is possible) at an event horizon

New comment by bencyoung in "GitHub Copilot Coding Agent"

bencyoung — Tue, 20 May 2025 12:14:03 +0000

Seems amazing similar to the changes a junior would make (jump to the solution that "fixes" it in the most shallow way) at the moment

New comment by bencyoung in "Dead Stars Don’t Radiate"

bencyoung — Tue, 20 May 2025 12:10:20 +0000

Thanks! Hmm, I think we're talking about slightly different things but it's been too long since I studied it to put it in the right words :)

I completely agree that spacetime can be "flatish" for a large block hole, but the event horizon does still represent a boundary right?

Consider the edge case of crossing the event horizon itself at some speed <I agree that your head will pass through the future light cone of your feet, and so could do somethign to affect your head (by emitting something falling slower than your head), but I'm not sure any light rays could follow that path.