I’ve spent a lot of time thinking about what I want to work on next, and in particular what project has the potential to become a “product” as well.
Since I was a kid I’ve wanted to have a computer company, and over the years I’ve imagined this in many forms. For the most part the idea doesn’t go much further than that, because I’ve never seen myself as the kind of person who could enjoy running a company, and most of the computers I want to build don’t have an obvious audience.
However over the last few years I’ve received positive feedback for a number of computer designs or prototypes I’ve built, and the one that seems to strike a chord with others most often is OffGRiD. I’ve never built a working prototype of this machine because there’s a number of tricky problems to solve, and it seems like each time I solve one problem, the solution to another problem becomes undone. I think I can solve all the problems, but it’s hard to justify the time, energy and resources required to build a computer for myself. However, if the design could be re-used to benefit others, then it’s easier to invest what’s needed to get it done.
The other barrier has been the scope of the project. There’s a number of “unknowns” that have to be answered before the machine as-designed will be usable, and it’s possible that I could spend a year (or more) building a prototype only to find out that it doesn’t work the way I expected. I personally don’t have the resources to absorb a miss like this, which is another reason I haven’t pursued the work aggressively.
The solution to this of course is to break-down the development process into smaller efforts that have testable results, but how to do this wasn’t clear to me until very recently. Spring rolled around and once again I wished that I had spent some of the winter months working on a computer I could use outdoors. This time however I decided I’m just going to do something about it, even if I can’t realize all of the design goals of OffGRiD. So I tracked-down a Pixel Qi display (the screen originally selected for OffGRiD, but which has since gone out of production) and did some research to see what off-the-shelf computer I could cram it into (the results of this effort will be described in detail in a future post).
Once I accepted this compromise it became clear to me how I might break-down the development of OffGRiD into a series of experimental machines, each yielding a useful computer with capabilities increasingly like OffGRiD. Each experiment will give me a chance to test the features I’ve planned for OffGRiD in the field, and determine whether or not they are as valuable as I expect them to be. This process will also let me learn what I need to learn incrementally and there’s less chance of spending a couple years building something that in the end isn’t as valuable as I thought it would be.
There will certainly be “elbow” points along the way that will require significant investments of time or other resources to move forward, but hopefully the preceding experiments will have demonstrated sufficient value to justify these investments. In addition to my personal experience with each prototype, I will be sharing the results with others (potentially building & distributing additional machines if there is demand), which could provide further justification for increased investment.
I’m also going to use this process to experiment with the idea of “non-scalable systems”: systems and devices designed deliberately to work well at small scales and avoiding compromises that are typically made in the name of scalability.