The thing that helps me work through this is to be very aggressive about a small list of features for "version 1". Focus on getting version 1 online and working

The thing that helps me work through this is to be very aggressive about a small list of features for "version 1". Focus on getting version 1 online and working, and put all your shiny ideas on the TODO list for version 2. Once version 1 is out the door, the momentum of people using it will carry you forward. I've learned the hard way that for me, by far the biggest risk in a side project is me losing interest before I have something up and working. So now I just keep that in mind as I'm planning and pruning features for the initial version. The risk isn't "what if I don't have feature X and people get mad?", it's "what if this gets too complex and messy to implement and I get bored and quit before finishing?"