11–20 of 220 posts
There’s another algorithm that doesn’t depend on knowing which value is larger, the U.S. patent for which expired in 2016: unsigned average(unsigned a, unsigned b) { return (a / 2) + (b / 2) + (a & b & 1); } There's no way that should be patentable.
Wow! Just wow...
There’s another algorithm that doesn’t depend on knowing which value is larger, the U.S. patent for which expired in 2016: unsigned average(unsigned a, unsigned b) { return (a / 2) + (b / 2) + (a & b & 1); } There's no way that should be patentable.
There’s another algorithm that doesn’t depend on knowing which value is larger, the U.S. patent for which expired in 2016: unsigned average(unsigned a, unsigned b) { return (a / 2) + (b / 2) + (a & b & 1); } There's no way that should be patentable.
There’s another algorithm that doesn’t depend on knowing which value is larger, the U.S. patent for which expired in 2016: unsigned average(unsigned a, unsigned b) { return (a / 2) + (b / 2) + (a & b & 1); } There's no way that should be patentable.
There’s another algorithm that doesn’t depend on knowing which value is larger, the U.S. patent for which expired in 2016: unsigned average(unsigned a, unsigned b) { return (a / 2) + (b / 2) + (a & b & 1); } There's no way that should be patentable.
There’s another algorithm that doesn’t depend on knowing which value is larger, the U.S. patent for which expired in 2016: unsigned average(unsigned a, unsigned b) { return (a / 2) + (b / 2) + (a & b & 1); } There's no way that should be patentable.
That's utterly hilarious. I've never come across this problem before, I read the headline and that solution came into my head immediately before I'd even clicked. I don't think I'm clever, surely half of HN feels the same way. Software patents are comical.
Some unreal solutions here that show how amazing mathematics can be. Especially that Google patented method that only just recently expired. Props for including the assembler breakdown for every major CPU architecture.
I had to lol when I saw there was a patent for that. Divide both operands by 2 was my first idea before loading the page. (I like to try that sometimes before reading the articles.) I didn't think about the carry bit, but it seems like that would be a logical solution after 5 minutes of extra thinking. I'm not sure how that's patentable. That's insane to me. But maybe there is more too it. I didn't read the patent it…
The patent is for a circuit design to perform that algorithm in a single cycle. The algorithm was never patented, nor could it be.
There’s another algorithm that doesn’t depend on knowing which value is larger, the U.S. patent for which expired in 2016: unsigned average(unsigned a, unsigned b) { return (a / 2) + (b / 2) + (a & b & 1); } There's no way that should be patentable.
Exactly. I saw the title, thought "I wonder what other way there is to do this than the obvious one of pre-dividing by 2" and then opened the article and saw that the trivial way to do it was covered by a patent. Wow! Just wow...
There’s another algorithm that doesn’t depend on knowing which value is larger, the U.S. patent for which expired in 2016: unsigned average(unsigned a, unsigned b) { return (a / 2) + (b / 2) + (a & b & 1); } There's no way that should be patentable.
There’s another algorithm that doesn’t depend on knowing which value is larger, the U.S. patent for which expired in 2016: unsigned average(unsigned a, unsigned b) { return (a / 2) + (b / 2) + (a & b & 1); } There's no way that should be patentable.
Do bitwise shifts instead of dividing by 2. (BRB going to the patent office.)