[lee]

Is he doing some kind of rubber duck debugging? http://en.wikipedia.org/wiki/Rubber_duck_debugging

I don't talk to myself, but sometimes by simply talking aloud about my problems to my peers I can flesh out a solution to a hard problem.

I have seen other programmers talk to themselves to find a solution when they're stuck, and it seems to work quite well.

[brandnewlow]

What tends to happen with me is I'll hit a snag that I can't seem to figure out or debug myself. So I fire up gmail and write a note to one of the contractors who does odds and ends stuff for me.

I describe the problem in decent detail and ask when he/she is available to fix it. Send.

Then a few minutes later, the solution appears in my head, I go solve the problem myself, and send an apology note to my developer.

[evilduck]

I do this too, but I've found that I don't usually need to hit send, I just save it as a draft then go take a walk. The answer usually comes to me when I come back, but if it doesn't I'll send it.

The act of explaining it in enough detail so another person could understand the problem usually increases my understanding of the problem.

[asdfghjkjnmhb]

>The process is to meticulously explain code to an inanimate object

Isn't that what students are for?

[mattjung]

Nice link!