Six weeks is a sweet spot when you are hopeful about the next possible steps but also feeling the growing backlog. This is the time where if the project hasn't died yet then just the right mixture of impatience and excitement might drive you to jump to acceleration. To double down on one of these possible next steps and go as fast as possible.

But you know what? I'm actually not so clear on which next step to take. They all look so damn good. If I was unclear but I hadn't been working on new features for this project the previous six weeks then I'd say that the best action would be to just pick one -- even at random -- and give it a go without pressure on the outcome. But given that I've been building for these past six weeks I'm actually going to take the next two weeks refactoring the code and writing more tests. 

This was really hard to commit to. I want to emphasize how difficult this is because I don't want anyone reading this -- especially in the future -- to think this was some easy decision. 

Currently excitement still is my main driver in this early project. I can't wait for the new user interactions I want to build. But honest reflection on how much sloggier implementing some of the Tribes interaction has been relative to earlier features and likewise features unrelated to the existing codebase has made me accept that I ought to actually do maintenance work if I'm in this for the long haul.

So no new features for the next two weeks, except for any minor ones that might naturally come as a consequence of refactoring itself. 

When you can't wait to do something, reflect... is it something new? Then go wild.
When you can't wait to do something and you've been there for awhile. Maybe try the waiting. Because here the waiting is more 'doing' than doing.