I wouldn't; CPython is an infamously basic bitch implementation that leaves all kinds of low-hanging fruit behind. On the flip side, this means that % vs & barely makes a difference because both operators will likely go through bytecode dispatch or indirect calls anyway. Hell, even in C the difference would only matter in a hot/tight loop.
You're correct. Though, this case refers more specifically to bitmasks.
val & 0xFFFF0000 to isolate the first 2 bytes of a 32-bit int. You'll see this most often in little-endian architectures, since the least significant bits are placed first.
Interesting difference between GCC and Clang here - GCC always does the optimization, even with no -O flags.
One day I'm gonna fine-tune my own LLM by scraping all the best blogs/dev resources and it'll be like using Google circa 2004
ChatGPT is know to hallucinate sources as well. Article about it.
No it doesnt
While I respectfully disagree with you about whether or not LLM is a useful educational resource, I can definitely agree with you here. The LLM usually will run out of new information past the surface level of a topic, at which point I'll ask for book recommendations. I guess at the end of the day, the LLM replaces the librarian instead of the library.
Common Python L.
Psst
We have discussed textbooks some in the Math Thread. Linked is a post about a textbook that's good for applied calculus. As for other fields of math, I can suggest books, but I need the subject you're looking for.
endianness is only relevant when byte-swapping, the masking expressions work the same on all platforms otherwise the code wouldn't be portable.
Endianness is relevant whenever you're converting between word sizes because there is no canonical way to convert between e.g. a series of 32-bit words and a series of 8-bit words without making some choice. It's simply ambiguous to say "these four 8-bit numbers represent a 32-bit number", that's why network/disk/etc formats tell you the endianness. If you find yourself swapping bytes, you are almost certainly doing something wrong.
Thats a fairly ridiculous statement
You see it most in compiled binaries if you throw them into a debugger, which is why I mention endianness as a bit of a footnote for beginners to think about. And if you're reading/writing a file in binary mode, the ordering of the bytes does matter quite significantly.
The actual practice recommended by that guy is sensible, but everything about how he phrases it is oblique and/or misleading.
I guess if you were trying to associate accounts you'd try to build up a 'profile' to see if visits and visit prevalence matched between accounts. Even just domains in common would presumably help narrow things down?
Byte swapping is very common, so much so it is implemented absolutely everywhere in system headers, standard libraries, and compiler/platform intrinsics (C++20 added
std::endian, C++23 adds std::byteswap 
). Choose one and don't think about it.Yes, there's nothing actually wrong with it. Usually it's written more explicitly.
I've seen it countless times in crappy C codebases. Keep in mind the post is from 2012 and by Rob Pike, so it's had some time to influence people.
Common and sensible are different things, and it's not like C++ is above standardizing stupid or niche things.
- GDB trying to print Hello, world!