I fell into this trap recently. I had a pretty good PCB design that had been lying around for a month, which I started designing about four months ago. I kept waiting though. I wanted to think of something that would be a problem and solve it before I have to. I want to jump a generation of iterating. And I did this for about a month. I waited a month, when I could have spent $25 and gotten boards by the time I started working on it again. Real world feedback is unparalleled in its ability to let you see what you're doing. I'm still waiting on the boards, but they'll be here soon and then I start making it...

Since I sent it out, I have thought of a couple nice 'features' that would have reduced some of my work load, but only about 3 minutes a board. I waited a month to save three minutes........

I forget what author I was reading (one of the many posts on 'doing something') when I had a moment of clarity - JUST SEND THE DAMN THING OUT. I reviewed it a little bit, said whatever, sent it out. We'll see how it works when I get it.

It's easy to forget how nice it is do develop software. You get instant feedback as to whether something works on the surface level. It would be much worse to code on a system where you wrote 100 lines of code, had to send it out and get the compiled version back in a week. Syntax error on line 4, no feedback given for lines 5-100. Hardware is a much more elusive beast, but having something physical, something that you can bring with you, that isn't created from a plane of RGB lights, that you can hold, touch, and feel.

There's always a double edged sword, for every problem there is an opportunity, you lose a little bit of something for every convenience you take up in your life, the hardware software divide is one with the sense of touch and smell. Software currently doesn't feel or smell like anything. It doesn't weigh anything. It has no value outside of the tubes. A piece of hardware can reach across different boundaries and engage different senses.