Short version:

I am about halfway though chapter 2 of SICP and doing the majority of the problems, and have been distracted by small things more related to development. Should I focus on wrapping up SICP or should I focus on designing a rough product, putting it on the web, and selling it?

Long version: I am currently working as a mechanical engineer and am sometimes doing coding projects for work, and often doing small coding projects part-time. The coding projects I have worked on are mainly things I worked on while on company time for specific projects, and thus I do not believe I own the rights to them.

I would love to launch a project with the potential to sell it online.

My coding background:

I coded for the first time when I was less than 10 years old, and this mainly consisted of BASIC code. In my teens, I made a website with html that I hand-coded (frames and all) for a game that I was playing at the time. The only programming course that I took while in college was c++. I have been using Matlab/Octave on a continual basis since entering college about 8 years ago.

More specifically, I have created the following: In Java - I was in a PhD program and I worked on some Finite Element Analysis code that my professor created in Java. Before I dropped the program, I had added a module for a specific, relatively complicated composite element.

In C++ - I created a mechanical linkage simulator that would compute mechanical advantages, stresses, and deflections given design parameters (mainly geometry). More recently, I created a program to determine a system response to a power spectral density. (Basically, I have a mechanical structure and I shake it with a simulated rocket launch to determine the accelerations that the system responds at)

In Python - I ported the mechanical linkage simulator to Python because it was much easier to make a simple GUI in Python. I then modified the program enough to animate the opening/closing motion and to change geometry. (it was a simple outline and line drawing of the linkage)

In Octave - Most recently, I created a program that simulates combustion, gas flow, and heat transfer in a tightly coupled system that could really help my current employer.

In Scheme - No projects other than those problems in SICP.

Most of my projects have been text-based because I have been one of the only users of the program, except for that ported program in Python. I'm somewhat worried about making the most recent simulation program into an outside-of-work project of mine (generalizing it, giving it a real GUI and some graphics) due to intellectual property issues. I have come up with a few ideas not related to my work aside from this, but I'm concerned about the small potential market of specialised engineering programs.

As for what has distracted me from SICP, I decided to set up an old desktop I had laying around as an Ubuntu server, putting LAMP on it, and starting to work through another book on the basics of web design. I have also figured out a way to not only SSH into my server from my cellphone, but also access Emacs and I've written a few hello world programs in Scheme/C++ using my phone and ran or compiled them remotely in the same manner.

I've started to re-evaluate my focus and desires due to reading "Creating" by Robert Fritz, and I suppose that I don't have a specific path to go down. I wouldn't mind working on a project that has potential (either an idea I conjure up or someone else's) part time, but my current job is in a relatively remote/non-tech region. It's been somewhat difficult for me to get over the clean slate in front of me to create something on my own unrelated to work, even though I know that I have the capability to create just about any tool I would use at work, given enough time.

Any advice would be appreciated.