mgaunard

4 ms·

31.
by mgaunard 2mo ago
The problem with the approach is that it's not real JIT-compilation, it's just assembly templates with basic substitutions. By not using LLVM, you're missing all the optimizations it does.
32.
by mgaunard 2mo ago
It's a committee, many people are involved, all with different opinions, but any decision requires consensus. Why would anyone expect big broad changes? And historically, whenever they did happen anyway through forced compromise, they
33.
by mgaunard 2mo ago
The website is listing Monohon; that's a good one. Old Street, walk-in only.
34.
by mgaunard 2mo ago
We already tried the fully semantic web in the days of XHTML, where the document is purely semantics and CSS purely transformation. It didn't work in practice, because people want to structure their document based on how they expect it
35.
by mgaunard 2mo ago
London does have a number of independent Ramen spots -- not many though.
36.
by mgaunard 2mo ago
Kanada-Ya is a chain with at least 5 locations in London, and while they're fairly good, they're still somewhat focused on Deliveroo-type transactions. It remains quite different from a genuine ramen shop in Japan.
37.
by mgaunard 3mo ago
Any function can throw an exception unless it is marked noexcept. Your code shouldn't need to make assumption about whether exception are being thrown or not. Your mistake is thinking that a function potentially throwing means you need
38.
by mgaunard 3mo ago
tl;dr: we only want the USA to dominate because we think their military would never abuse AI (disproven a few months ago)
39.
by mgaunard 3mo ago
Features aren't that important; what's important is capabilities.
40.
by mgaunard 3mo ago
The cost of code actually increased; code debt is being accumulated faster than we can clean it up.
41.
by mgaunard 3mo ago
You're missing the fact that "multiplying a digit with a number" is not a single operation.
42.
by mgaunard 3mo ago
I feel like people use this "remember the Vasa" idea every 6 months. Yes, we all remember the Vasa, and understand how it applies to any untested overengineered idea.
43.
by mgaunard 3mo ago
how is it obfuscated? It's literally written as plain black monospace text on a white background. Pretty sure any AI can solve it in 20 seconds.
44.
by mgaunard 3mo ago
99% of all loose ties are due to people doing a granny knot instead of a square knot.
45.
by mgaunard 3mo ago
There is no shared lineage here. Some of the people who worked on sender/receivers indeed also worked at Nvidia on Thrust, which never had anything to do with coroutines.
46.
by mgaunard 3mo ago
most of those are just bindings, trivial work.
47.
by mgaunard 3mo ago
An actor does not have a queue; that's again the whole point I already made: it's event-driven, things are decoupled and the actor is not aware of how control flow happens. Regardless, when the queue which exists and is per-thread
48.
by mgaunard 3mo ago
Why is the AI only able to reach 17%? Surely it can just keep iterating until it implements the full test suite?
49.
by mgaunard 3mo ago
Well in mine for example, threads are pinned to a core and always spin, and actors are pinned to a given thread (you can have an arbitrary number of actors per thread, they're just objects), so "waiting" is certainly not some
50.
by mgaunard 3mo ago
That sounds like a misunderstanding of how the actor model works. An actor doesn't wait. It's an event-driven system, it doesn't get to own and decide when messages get fed to it.
51.
by mgaunard 3mo ago
not really, they yield to other event sources, which is the opposite of waiting (which blocks forward progress and is a violation of real-time guarantees)
52.
by mgaunard 3mo ago
yield to the event loop
53.
by mgaunard 3mo ago
Coroutines aren't great, even C++26 is explicitly avoiding them for its foundational concurrency model. In any case a good concurrency library doesn't need more than C++14.
54.
by mgaunard 4mo ago
essentially this is a mini-ISPC?
55.
by mgaunard 4mo ago
boost unordered flat map didn't exist in 2016 (nor 2019).
56.
by mgaunard 4mo ago
How does it compare to boost unordered flat map? Looks like the benchmarks were last updated in 2019.
57.
by mgaunard 4mo ago
You're suggesting pre-filling future data based on extrapolating current rules (up to an arbitratily unknown high date) instead of using a constant. I don't really see how that is a good approach in any capacity.
58.
by mgaunard 4mo ago
Never heard of DST? The authoritative time is constant in the local time zone, but needs to change in UTC twice a year. This is the exact reason people store time in local time zones. Also remember the date/time where DST switching occ
59.
by mgaunard 4mo ago
A whole lot of nonsense from a web guy. Please, keep using JWTs, they do their job well: giving you an access or ID token that you can pass between applications and trust based on cryptographic signatures from an identity provider.
60.
by mgaunard 4mo ago
Film is 24fps which is lower than PAL. The 30fps of NTSC is actually associated with "cheap-looking" video.

More ›