I spent years building: 1. an 'abstract visual debugger' 2. a new kind of program editor —and I made the same critical mistake in both cases: I didn't treat

I spent years building: 1. an 'abstract visual debugger' 2. a new kind of program editor —and I made the same critical mistake in both cases: I didn't treat the programs as sufficiently exploratory. I reached a point in each where I thought the idea was good enough that with my limited resources I should jump directly to building the real/production version. Both were far too big/complex for me to build production versions of alone, though I could've iterated on design issues through prototyping and probably attracted investors to build for real. Looking back I think part of what underlay this poor decision was partly a form of perfectionism, and partly that I didn't know good tech for prototyping at the time. [1] http://symbolflux.com/projects/avd http://symbolflux.com/projects/avd [2] http://symbolflux.com/projects/tiledtext http://symbolflux.com/projects/tiledtext