I work on personal projects a few days or weeks at a time and then after solving the first set of big problems I generally let it slip and lose interest. The source code lives on my hard drive and then one weekend next year or a few years later I remember what I had been writing and I come to revisit it. Maybe I rewrite the program or extend it up to the next level, or just look at it and think "That's where I was, mentally, in 2010; I think I've made progress since, I'm not the same programmer anymore."
I place no intrinsic value on finishing the projects itself because then they would start to resemble work and I want my hobby projects to be fun.
At work I procrastinate too but that's mostly doing the more interesting tasks or tasks with the highest impact first and the less interesting tasks only when really, really needed. So I don't basically procrastinate, I just prioritize.
I think procrastination is prioritizing, really.
And sometimes you prioritize complete idling over doing anything at all which I think just tells that you were trying to do the wrong thing because if you weren't you would have forgotten to eat while staying up doing it.