[factotvm]

Then give them a coding problem that takes a short amount of time but allows you to evaluate their skills. I relish these opportunities and respect the companies that ask me to do them.

To keep with your metaphor, ask the carpenter to make you a cabinet door.

Again, I agree with others here... if you're passionate about the startup you're working at (and you should be), that is your side project. I distrust people who work at startups that have side projects.

[eropple]

A coding problem is less illustrative than a decent--LOC is always suspect, but let's say 10,000 LOC--corpus of work. Something to show much more than a toy problem.

Better still, however, is collaboration with open source projects--you can see individual communication between people as well as the developer's ability to enter a foreign codebase and be productive, (hopefully) without introducing bugs due to ill-advised changes without understanding what's being changed.

(We don't look at side projects or open source at my current employer. I wish we did.)

[cmorrisrsg]

Any coding problem that could feasibly be completed in an acceptable amount of time would be too small to actually tell me what I want to know. I mostly don't want to know what algorithms you're aware of or how quickly you can type. I want to know how you structure and modularize code. I want to know how you debug things. I want to know if you can refactor well. In short, I want to know about how you handle building and maintaining real software. If I hired people to do coding problems all day, that would be another thing, but I don't.