If this isn't meta, what is? For those who can't get enough discussion of MO, Anton was the guest on the SO podcast this week.

Listening to it now, I'm surprised by how slow I talk.

I emailed Jeff and Joel afterwards to tell them that another pain point with the software is the crazy bounty rules. It's something they plan to revisit, but the SO team is working hard on an API, so it's not clear how much time they'll have to clean up bounties.

I agree with fpqc -- the interviewers seemed to forget for several minutes at a time that Anton was there. They didn't seem to be especially interested in what he had to say.

Still, it was interesting. I enjoyed the plug of my personal best question "Which journals publish expository work?" although I don't think Jeff and Joel understood the subtext of it, i.e., "Why so few, and what should we do about it?"

Anton also brought up the double covering of S^n question, which I decided to answer but only in a comment. He argued nicely for why it was good for the community to record an answer to this question. I'm still not quite persuaded -- the argument seems to begin "Consider a research mathematician who has not had a single course in algebraic topology." Such people exist, but when a research mathematician is so far out of his element that he lacks the basic knowledge that we want undergraduates / first year graduate students to learn, I feel like he is not acting like a research mathematician by asking this particular question, whereas MO is for mathematical discourse at the level of research mathematics. (I.e., we unanimously want to take questions from people who are not research mathematicians but can communicate at approximately that level. I am saying that perhaps this cuts the other way as well.) But I'm not yet really decided on this...