senderista

And repeat that work every time you sync with upstream?

Sorry, I agree I was unclear and your restatement accurately reflects my intent. My biggest concern with Zig safety is that ReleaseFast turns asserts into assumptions, which is equivalent to injecting UB at every failed assert site. I doubt many users are aware of that "feature".

Part of the safety you get with Rust is runtime bounds checking, which is equivalent to Zig in "safe" mode.

Yes, I recommend googling "mirror symmetry" if you're unfamiliar.

What about this common stereotype: EA advocates that one should take the most lucrative job possible, regardless of any moral harm it may cause, as long as it enables you to do good which (perhaps barely) outweighs the moral harm your job caused in the first place? How fair or unfair is that stereotype?

Indeed, "race" is a folk concept and has at best limited correspondence to genetic population structure.

We are all mixed race.

https://www.goodreads.com/en/book/show/35749414-who-we-are-a...

I wouldn't consider donating until the project leader has demonstrated considerably more personal maturity.

Seems like "parallel construction" where you show you actually understand and can explain the issue independent of an LLM would be the way to go. No need to mention how you found the bug as long as you can explain what the bug is, why it matters, and how to repro.

It does have SMP support but it's unverified and I'm not sure how widely used or well-tested it is.

Even Bertrand Russell briefly advocated this. "Better dead than Red."

I believe Nash himself didn't consider it particularly noteworthy either, and was far prouder of his work in pure math (like the Nash embedding theorem).

Did Hunter Biden work for his father's business?

Not a single mention of AWS Kinesis? That seems like the closest competitor.

Minor correction: it is not true that no major object stores support appends--Azure Blob Storage has from the beginning.

lobste.rs has always been invite-only.

Ultimately it's the fact that compile time was not a first-class consideration during the design of Rust's important features. There's only so much you can do to mitigate the consequences.

Santorini may have helped cause the Minoan collapse, not the LBAC which occurred c.400 years later.

I don't understand your chronology. The LBAC was c.1200 BCE, while the Harappan collapse was c.1900 BCE IIRC.

Agree, look at how many incorrect concurrent algorithms have been published over the years, where seq_cst is assumed. It's easy enough to mess up concurrency without any weak memory orderings. There are only a few people I know of who I would trust to informally reason about weak memory orderings in nontrival algorithms (Dmitry Vyukov, for one).

You seem to be implying that it's a toy, which it very much is not.

https://sel4.systems/use.html

Have you seen Lampsort?

https://bertrandmeyer.com/2014/12/07/lampsort/

A TLA+ spec defines both a model and properties (global invariants). How do you know that the properties the LLM specifies are the ones you care about?

Which style guides? That makes no sense to me because 1) if you're writing a lock-free algorithm/data structure you presumably both know what you're doing and care a lot about performance, 2) many lock-free algorithms don't even require any seq_cst operations (or equivalent fences), and 3) weak memory orderings can be essential to getting acceptable performance in critical paths.

I would also note that aside from formal methods, LLMs are absolutely not trustworthy but the top frontier models can reason to some degree about weak memory orderings, and can at least find concurrency bugs which can be later confirmed by human expert review (preferably after eliminating false positives via adversarial LLM review of the findings).

Sounds like your dream workflow could replace you with a PM?

You can use LLMs to make your software better, but it's easier to use them to make it worse.

As an introvert kids' birthday parties are purgatory for me.

Astra is less nitpicky IME than Fable.

Uh, CockroachDB?

Oracle seems to be doing their best to expedite their own crushing. I wish them success.