aappleby

41 ms·

511.
by aappleby 8y ago
Also, there is nothing magic about his fibonacci constant. Any large odd constant with roughly half the bits set at random will do the same thing.
512.
by aappleby 8y ago
Don't do this. Use a real hash function that guarantees a highly random distribution, make your hash tables power-of-two sized, and map from hash value to table index using (hash & (size-1)). The fibonacci constant thing will help
513.
by aappleby 8y ago
"The Open Policy Agent (OPA) is an open source, general-purpose policy engine that enables unified, context-aware policy enforcement across the entire stack." To someone who doesn't already know what a "policy engine&quo
514.
by aappleby 9y ago
Also order-of-magnitude questions about cpu vs gpu perf (arithmetic, special functions, memory bandwidth and latency)
515.
by aappleby 9y ago
Should add "write me a pseudocode vertex shader and fragment shader", and a question or two about the order of magnitude costs for switching uniforms/buffers/framebuffers/shaders.
516.
by aappleby 9y ago
Broad statistical tests are good enough for most all non-crypto RNGs - once you pass DieHarder you're probably good enough for practially all common RNG uses. Breaking crypto RNGs is a subtle enough art that having a standardized set o
517.
by aappleby 9y ago
The wikipedia page might be a bit misleading - there's a 128-bit murmur3 that uses 32-bit math (works well on most every processor), and a 128-bit murmur3 that uses 64-bit math (much faster on 64-bit processors, much slower on 32-bit o
518.
by aappleby 9y ago
I was going to ask if he'd run it through SMHasher, but... yeah, clearly not. -Austin (author of murmurhash/smhasher)
519.
by aappleby 9y ago
Really interesting analysis, but constraining their keys to 64-bit integers and using 'hash functions' that only accept 64-bit integers as input skews the results away from real-world use cases. In practice I've found that a
520.
by aappleby 10y ago
Oops, it's not a handheld though...
521.
by aappleby 10y ago
Uzebox, though it's a tiny bit dated.
522.
by aappleby 10y ago
Wrote this a while back but didn't get much feedback - might be helpful if you know more Javascript than Scheme. https://medium.com/@tanjent/refactoring-out-a-y-combinator-9...
523.
by aappleby 10y ago
Take a look at the tests in SMHasher, they indirectly spell out a lot of what makes a hash "good" (though not all the tests are particularly readable).
524.
by aappleby 10y ago
Making a fast, good hash function isn't too hard now, so some sort of "provable key-independent collision resistance" is definitely the next thing that needs to be worked on. That said, I haven't really looked into the t
525.
by aappleby 10y ago
Negligible, a few tens of cycles of extra overhead.
526.
by aappleby 10y ago
Actually, just noticed a minor issue - since there's no intermixing between the four lanes and the diffuse() function is the same for all of them, if any of the IVs match then I can swap all the blocks in those lanes and get the same h
527.
by aappleby 10y ago
Looks like a pretty straightforward 64-bit block hash unrolled 4 times. I'd prefer a bit more assymetry in the diffuse() method, but since it passes SMHasher it's probably OK. I wonder how the Rust version compares with plain-jane
528.
by aappleby 10y ago
class CSectionOfSingleMaterialDrawableObjectBuilder { ... };
529.
Refactoring out a Y Combinator(medium.com)
1 points by aappleby 10y ago|0 comments
530.
by aappleby 10y ago
If you can only hash data a byte at a time, you have bigger problems. ;) I have no real suggestions, but it would be trivial to feed Murmur one byte per loop iteration and then ditch the 'tail' section.
531.
by aappleby 10y ago
It'd be interesting to see how the 128-bit Murmur versions compare. Something not mentioned in the "Quality" section is that FNV and djb2 produce may produce decent results on words/filenames/source code, but they c
532.
by aappleby 10y ago
Imagine that I visit a cloning facility that can flawlessly duplicate a human, memories and all. The techs there put me in a medical gown and sedate me, and a while later I wake up in an empty room. Without any additional information, I hav
533.
by aappleby 10y ago
To clarify, I tried to do something like this back in the 90's and it was _not_ worth the effort. Using something like Google's Protocol Buffers or a similar system where you generate a standard serialization backend for all the p
534.
by aappleby 10y ago
"is a bad idea".
535.
by aappleby 10y ago
Common ones from when I started writing video games in C in the 90's - * Write malloc() and free(). * Write a gzip decompressor. * Write a triangle rasterizer and use it to draw a spinning cube on the screen. Each of those is a decent
536.
by aappleby 10y ago
I don't see anything in the Indiegogo page that shows that he's designed a chip before, and the github link goes to someone else's repo. Googling his name turns up a blog here - http://blog.the-leviathan.ch/
537.
by aappleby 10y ago
This is a fabulous tool for exploring micro-optimizations across a range of CPUs. I've used it for investigating how some inner loops get pipelined across AVR, ARM+Thumb, ARM32, ARM64, PowerPC, x86-32, and x86-64 processors.
538.
by aappleby 10y ago
Hmm, that's probably also a good place to start an analogy about code maintenance. Weed by hand first, then weed using good tools, then automate your weeding if things get out of hand.
539.
by aappleby 10y ago
My wife has been teaching herself to program for the last year - one of her big challenges has been understanding when to hack on a problem until it's solved versus when to go looking for sharper tools. She found this analogy quite hel
540.
“The Code Lumberjacks” – a simple analogy about how effective programmers think(docs.google.com)
7 points by aappleby 10y ago|3 comments

More ›