Flutterby™! : Interviewing

Next unread comment / Catchup all unread comments User Account Info | Logout | XML/Pilot/etc versions | Long version (with comments) | Weblog archives | Site Map | | Browse Topics


2004-08-28 16:47:28.756177+00 by Dan Lyke 4 comments

After being so quiet about one job that my first mention of it on Flutterby was when I was leaving it, I figure I'll get this one in up-front: I did a phone interview with Amazon yesterday. I'm still interested in all of the possibilities I'm exploring, that's just one of the several on the list.

It was interesting, a lot of "how would you solve this problem" sort of questions, and as a follow-up they sent a "how would you process this data" email, asking for a response in a few hours. So I documented my assumptions about the data and wrote this massively over-engineered solution which had some issues and some strengths, but...

One of the questions, while I was dictating code over the phone (and how fun is it to try to visualize that?), was "why do you use the post-increment operator (ie: i++) rather than the pre-increment (ie: ++i)?" [Edit: I was the one using the i++ form, he was questioning] It was relatively quick to discover that the interviewer was looking for the "the compiler has to store a temporary in case it's going to use it in an expression" answer, but I went back to some notes I wrote for a presentation on the 1988 Conference on Computers in Physics Instruction[Wiki], and it turns out he's wrong: Early versions of Turbo C[Wiki] were an instruction more efficient for the "post" operators than the "pre" ones. And it must not have mattered through the various versions of Microsoft C[Wiki] up through the 16 bit versions of Microsoft Visual C[Wiki], because I kept using that construct, and in those days I was very concerned with saving bytes.

Anyway, it reminded me of being back in high school physics and making up some answer for the "why do AM radio signals travel further than FM radio signals" because I knew of (and was playing with at that point) several applications where AM and FM were used on the same frequency range and didn't think to tie those notions to the "standard audio AM broadcasts are on lower frequencies than standard audio FM broadcasts, and therefore bounce off the ionosphere" [edit: and diffracts more strongly] answer, one of those places where too much knowledge was a bad thing.

[ related topics: Dan's Life Microsoft Software Engineering ]

comments in ascending chronological order (reverse):

#Comment Re: made: 2004-08-31 23:25:28.343779+00 by: Dan Lyke

Flying to Seattle next week...

#Comment Re: made: 2004-09-01 04:03:53.4824+00 by: topspin

I'm noting that the cool jobs at Amazon are in the Seattle/Tacoma area. Hmmmmm....

#Comment Re: made: 2004-09-01 05:37:21.044131+00 by: Diane Reese

This is giving me the thought that I ought to kick-start the organization of that South Bay Interesting Persons Gathering p.d.q. or the presence of Dan may not be assured...

#Comment Re: made: 2004-09-01 23:44:59.711073+00 by: Dan Lyke

Yeah, although I may have started a turf war, in my talking with folks down on the peninsula today, one of 'em was from another division of Amazon.

But if I'm going to move anyway, I think I'd rather move out of the Bay Area. Lots of tough decisions floating around in my mind, and the week's only halfway over.