Carmack's was very much a part-time venture. It was interesting to read their blog and watch their videos. It was clear that they were very good at certain things (code + welding) and not very good at others (electronics, fire-proofing, planning, testing, finite element analysis, any kind of simulation that Carmack didn't write).

An interesting fact is that they did very well as long as they could conduct cheap tests very often. As their rockets got bigger, they could no longer do that and then they floundered.