Link Details

I've been interviewing programmers for 2+ years now and I can sympathize with how hard it is to find decent talent. Most are pretty bad.

I guess the problem is that the good ones don't really move about.

I'd say you're right bloid. I usually try to go after people that are currently employed. They seem to work out best.

Good people usually have their own agenda. You can poke around, but if your company isn't attractive to them, you won't get them from a jobs site.

I don't think the good ones ever hit a jobs site... It's all based on word of mouth and friends of friends

Hm, even good programmers sometimes need to hit job sites to look for industries they might be more interested to design solutions for than only those industries their friends are working for. Personally I don't know anybody doing development for train traffic control systems or am aware they know somebody - but I bet it would be interesting to design such a system.

That's mostly true - but the occasional search on a site at least keeps you aware of the market: what's hot, what the pay is doing and whether it might just be worth putting out the feelers.

Oh yeah, that's what Friday afternoons are for ;-)

I've been getting the reputation as the "Chaos Interviewer" lately, about a year ago I changed my approach from the standard "Sit and go over major topics in a nice orderly fashion" way of doing interviews to one that keeps people on their toes.

Firstly I do most of the general topic material on the walk from the lobby to the conference room, and before they can sit down and get comfortable I throw them a marker and have them do pseudo-code of some set of functions, let them get almost done and change some requirement they covered early in the code.
Next I put a small set of books related to the topic of the design that I give them, a REALLY loose design and a horrible requirements list, and leave the room quickly, and watch them on the conference room webcam and make sure I'm gone for a good 10 minutes before I return.
Then I give them a "perfect world" scenario design, touting that it IS a perfect design, while it has obvious flaws, VERY obvious flaws.

This is all before we get into technical ability, the walk covers people who can juggle multiple fast paced tasks, the 1st design covers people who can think on their feet, the 2nd design I'm actually looking for people who will try to come and find me or better yet crack open the reference materials on the end of the table and find the answer themselves, 3rd I'm looking for people who will actually argue with me when something I give them is wrong, I dislike "Yes-men".

Failing one of these isn't fatal, but failing 2 or more is.

I can work with single tasking, fast thinking, researching, free thinkers.
But single tasking, slow thinking, needy, yes-men belong in the government.

*makes a note to pack a spare pair of pants when going for an interview with dgary*

;-)

Of note I might add, is any developer who has gotten past me and gotten hired has been worth their weight in gold, but those who failed me and still got hired (I'm not HR, and nepotism exists everywhere), have been worthless piles of crap, cost the company twice what their paycheck was (if not more) and put several projects back more than the amount of time they spent working on them.

You'd probably pass at least 3 of my tests Bloid, well, maybe 2 =D