On 7/14/20 1:25 PM, Miles Fidelman wrote:
- If someone asks me to do an algorithm or coding question, I generally tell them to pound sand; that I generally use the language statement or a standard library, or look up hard stuff in Knuth - and then ask them if they'd like to discuss the specifics about how I might approach finding/developing specialized algorithms for the problems I'll be working on. (I refuse to be a code monkey on a string - and if they insist, I know that there's no way I'd want to work for them.) I'm reminded of a story an old-line DEC engineer told me - at his interview they asked him about converting an octal number to hex, or some such - he basically asked if they had an octal-hex calculator handy (remember the old paper ones?). After that, the interview went swimmingly - he thought that was kind of a test to see how he'd react (who really wants to hire someone who's going to start doing paper calculations of something silly).
I got rejected once because he wanted me to write strstr() on a whiteboard and it was insufficiently Knuth blessed, which I admitted and said in real life I'd just look it up, because you know, that's what resourceful engineers do. That wasn't good enough. Mike, I didn't even know it was programming interview so it took me aback even more