I don't have a lot of structure. Highest level, I try to work on things with the greatest marginal benefit. Micro opportunities fit into this framework because

I don't have a lot of structure. Highest level, I try to work on things with the greatest marginal benefit. Micro opportunities fit into this framework because for a brief period they have a really high peaks of marginal benefit. The trick is to be able to see which ones are really worthwhile and which are not. No matter what you'll have misses, but they're fun so I still don't feel bad about that. But over time I've gotten better at estimating the size of the opportunity and the scope of the work required to address it. To answer your question more directly, I keep a number of projects open and move forward on them incrementally. That's the main branch and my default position. I don't set deadlines or goals beyond MVP. Then I have a background thread that monitors these things (HN, feedback emails and the like). If I see something worthwhile/fun/whatever, I just drop everything and do it.