Having just moved to the interviewer side I appreciate the remarks about wanting the interviewee to succeed. I know how stressful it can be doing an interview and really do try to make it something where I can judge the ability to code, not just ability to interview. I usually try to start with something simple to just get some code going - like "We're going to need some random numbers. Make a little function that returns a list of 100 random numbers between 0 and 10." At that point I can start asking questions about why they do something a particular way and what conditions would make them do something different. Once that gets going the stress usually drops noticeably.<p>One other point I would add is to know the difference between being relentless and just making things up when you run up against the end of your competence. Same as mentioned in the article I will try to push out to the limit of the interviewee and a simple "I don't know for sure, maybe something like ....?" at least lets me know you know your limits. Hopefully it's not right after the first question but really don't just start making things up.
Yeah, the response to adversity is key. Being able to keep going and figuring things out even when you're stuck or out of the realm of your domain knowledge is so important.