Why A New Language

A friend of mine, this weekend while examining a new Green Arrays dev board with 288 cores, asked me why I spent so much time abusing other languages and didn't just implement my own. Now part of the inspiration for this question as the board sitting in front of us. Chuck Moore has not only designed the programming language, wrote his own CAD system, he also designed the chips as well. The question of could a skilled practitioner of the art build the full thing himself? The answer was undeniably yes (assuming he had access to a fab facility and could afford the wafer runs).

My answer to his question was that I was more interested in establishing the techniques necessary to program the types of systems, so if I could bend any existing language to my will, it was a short jump to running my code anywhere. Moreover, each experiment would allow me to better understand the problem space, and remove barriers to thinking by forcing myself to overcome orher's implicit assumptions. That being said, I still have a good idea of what my preferred language would look like:

And that's about it.