[insoluble]

My biggest failure in 2015 was probably that I was too reactionary. I lacked the autonomy that I often had in the more distant past. Instead of making and following long-term plans, I found myself doing small projects and waiting for people's reactions before proceeding to do something else. What I would do next was too often based on the responses I received. There have been times in the distant past where I was the opposite -- too much autonomy for too long -- leading to building a large project that nobody really wanted. In 2015 I saw the opposite end of the spectrum. Recently recognising this phenomenon has helped to remind me that both ends -- heavy autonomy and heavy interaction -- are bad for productivity. I am now thinking that perhaps autonomous spurts of 3 to 9 weeks at a time are best in the long run. Waiting for a reaction after just a few days of work is too short a time, and building alone for 9 months without validation is too dangerous.