mwkaufma

15 ms·

361.
by mwkaufma 2y ago
Yeah, I hate that JS can do it too. Website bloat is a problem we should be fixing, not accepting. Have you tried to browse the web on a "budget laptop" lately?
362.
by mwkaufma 2y ago
Let figma ask me for explicit permission first, then, as opposed to making Memory64 support implicitly available.
363.
by mwkaufma 2y ago
The Memory64 proposal is for >4gb space -- and specifically to use it, not just reserve it. Learn context before commenting.
364.
by mwkaufma 2y ago
The website may want to, but I don't.
365.
by mwkaufma 2y ago
Love to visit untrusted websites that feel entitled to over 4gb ram without asking.
366.
by mwkaufma 2y ago
Tedious moral panic.
367.
by mwkaufma 2y ago
Another article immediately skipped for leading with GenAI image slop.
368.
by mwkaufma 2y ago
Ask a podcaster to name intellectuals, and they just name other podcasters.
369.
by mwkaufma 2y ago
Making a mountain out of a molehill.
370.
by mwkaufma 2y ago
That's section 10.2 of the linked article, yes.
371.
by mwkaufma 2y ago
How do we discriminate when a response is correct, vs. when it's "hallucinating" an accurate fact, by coincidence? Are all responses hallucinations, independent of correspondence to ground-truth?
372.
by mwkaufma 2y ago
Spyware-As-A-Service
373.
by mwkaufma 2y ago
Author here AMA
374.
Tech Breakdown: AI with Finite State Machines(blog.littlepolygon.com)
2 points by mwkaufma 2y ago|1 comments
375.
by mwkaufma 2y ago
(2003)
376.
by mwkaufma 2y ago
"they just float around and help people" heaven forbid
377.
by mwkaufma 2y ago
Will we see a return of the High C compiler extensions? https://duriansoftware.com/joe/the-lost-language-extensions-...
378.
by mwkaufma 2y ago
The "next" field in the lead code sample is supposed to be a pointer, right?
379.
by mwkaufma 2y ago
AI generated banner image: immediate skip.
380.
by mwkaufma 2y ago
What's the point of profiling for performance without optimizations if it leads you where the OP winds up hand rolling the same transformations?
381.
by mwkaufma 2y ago
I don't understand why half of these aren't optimized by the compiler automatically. (x - y).norm_squared()? Why is f32::clone() not just an inline mov? Begging a lot of questions.
382.
by mwkaufma 2y ago
Yeah exactly, that's why refusing to simply spec the bytecode at the time was exactly the sabotage that everyone called it out as.
383.
by mwkaufma 2y ago
SDL3 offers a spir-v-cross reference impl as an optional auxiliary library.
384.
by mwkaufma 2y ago
WGSL is defined as a "bijection" to a "subset" of spir-v -- they could have simply specced that subset.
385.
by mwkaufma 2y ago
CVS at my first job was configured to reject commits containing f-bombs after a spicy debug message sneaked into production and showed up in a children's game.
386.
by mwkaufma 2y ago
"For dollar figures I find a prefixed dollar symbol and convert the number following it into a float." Bloombug red flag!!
387.
by mwkaufma 2y ago
The optimization is the linear memory layout of the nodes -- value speculation is decoration.
388.
by mwkaufma 2y ago
Why are braces-for-both-blocks-and-literals/etc a deal breaker here, but not in JS?
389.
by mwkaufma 2y ago
>> If you look at medicine, one of the slightly strange things about it is that they subtract the placebo effect. Now, given that the placebo effect can contribute to a cure, or to the efficacy of a treatment, you’d think people would
390.
by mwkaufma 2y ago
That's consistent with STL's move to use "explicit" and other narrowing annotations, as described in the article.

More ›