ethan_

But I really want to be an astronaut ;-). If I fail my astronaut exam, perhaps I'll temporarily make french fries, only to be back for the next astronaut exam.

Getting a job right now is like preparing for the astronaut exam, not taking it, and deciding to make french fries instead. While making french fries I'll constantly be wondering whether I would have passed the exam and become an astronaut.

At the moment, the pressure put on myself to become independent and self-sustainable is less than the pressure to finish this project.

I admit it's irrational to view getting a job as failing. However irrational it may be though, at the moment it is still my viewpoint. In the future this may change.

I completely agree that getting a job can be a valuable experience and I could certainly use some perspective.

What I am concerned about is whether I can realistically manage to develop for 8 hours a day at work, come home, and develop for 1-4 more hours. Even if I did manage to do it, how much time does that leave me for other activities? Not to mention, the feedback loop would be absurdly long for anything but the most trivial weekend projects.

All this said, if my parents decide to cut me off tomorrow, I'm immediately looking for a job. I want independence. I want self-sustainability. I know a job can provide them. But before I start looking for a job, I want to finish my project and see if it's as promising as I anticipate. If it flops, at least I tried.

Is there anything wrong with this? I suppose it depends on how long it takes to get things done. I do not want to be in this position 6 months from now, being 90% done instead of 80% done.

What would the consequences be of not getting my stuff done at a crappy job?

Good analogy. I can definitely relate to this feeling.

I like the idea of launching a minimum viable product and I have recently been keeping it in mind during my design/development cycles. The idea of focusing on maintaining velocity seems more appropriate for web apps. My project has a long learning curve and an inherently long feedback loop. The amount of boilerplate code required for something minimally usable is extremely high compared to that of a webapp.

I should clarify something. My project involves software that places automated wagers based on statistical analysis and machine learning. It is initially intended for private use. When I "launch it" I will get feedback in a matter of weeks regarding its profitability, after adjusting for variance. I already make money part time wagering money, but it's not enough to fully support myself. Automating the process allows for greatly increased profit potential. Some of my limited savings will be used as a starting bankroll.