I have no idea what you just wrote. I don't speak C, know basically nothing about it, I'm not in the business and I'm not ever going to do either, because I don't like it.
So I looked that up. You're talking about intrinsics. I have no idea about that, beyond "Why would people rather remember this instead of just learning the proper assembly code, which is far easier to read and remember anyway?"
I don't write code for compilers to do the work I am supposed to be doing, so I have no idea of the world you code in.
Your response doesn't really answer my question, but that's totally my fault, so ...
Hand me a box with a defined problem, some compiled code I can run to benchmark my own solution against, then wait for me to beat your solution within defined parameters.
My goal is to have some challenges that need optimization, but there needs to be a point to it. I can take any problem from the web, rethink the approach and rewrite it in assembly, but in 99.999% of all cases there's no point to it and I have nothing to compare it against on my own machine anyway.
I do not know what's generally useful to optimize and I certainly do not know who actually cares about that anymore.
Like, for example, how the pattern matching/string search function in freepascal is damn slow. I've researched the topic and noticed that all the solutions I've found are basically crap, so I wrote my own, beating it by some wide margin in terms of codesize and performance. I don't even know why these devs accept subpar solutions, but responses I got to that question were, basically, "It's fast enough" ... which is dumb.
Still, I don't actually have a way of fairly comparing my work.
Regardless, apparently the fastest fixed size pattern/string search algorithms I've found are all crap, but I really have no way of creating a fair comparison. Oh, on that note ... you don't, by chance, have a pattern matching benchmark for fixed sized needles I could compare my own against?
Sorry for being a mess. :D I really just wanna work on something that's actually a challenge AND something people actually need to run more efficiently, just like the QuickSort example.
So I looked that up. You're talking about intrinsics. I have no idea about that, beyond "Why would people rather remember this instead of just learning the proper assembly code, which is far easier to read and remember anyway?"
I don't write code for compilers to do the work I am supposed to be doing, so I have no idea of the world you code in.
Your response doesn't really answer my question, but that's totally my fault, so ...
Hand me a box with a defined problem, some compiled code I can run to benchmark my own solution against, then wait for me to beat your solution within defined parameters.
My goal is to have some challenges that need optimization, but there needs to be a point to it. I can take any problem from the web, rethink the approach and rewrite it in assembly, but in 99.999% of all cases there's no point to it and I have nothing to compare it against on my own machine anyway.
I do not know what's generally useful to optimize and I certainly do not know who actually cares about that anymore.
Like, for example, how the pattern matching/string search function in freepascal is damn slow. I've researched the topic and noticed that all the solutions I've found are basically crap, so I wrote my own, beating it by some wide margin in terms of codesize and performance. I don't even know why these devs accept subpar solutions, but responses I got to that question were, basically, "It's fast enough" ... which is dumb.
Still, I don't actually have a way of fairly comparing my work.
Regardless, apparently the fastest fixed size pattern/string search algorithms I've found are all crap, but I really have no way of creating a fair comparison. Oh, on that note ... you don't, by chance, have a pattern matching benchmark for fixed sized needles I could compare my own against?
Sorry for being a mess. :D I really just wanna work on something that's actually a challenge AND something people actually need to run more efficiently, just like the QuickSort example.