I'd say yes. Grab a microcontroller and do something fun with it, or setup a constrained machine and write some real software using it. I don't necessarily su

I'd say yes. Grab a microcontroller and do something fun with it, or setup a constrained machine and write some real software using it. I don't necessarily suggest using old architectures that are no longer relevant, but instead a constrained resource machine. My reasoning is it forces you to really make better code choices, algorithms etc. Simply because many times choosing the wrong algorithm etc on newer hardware has little consequence until you get a larger data set that is hard to test with. Hopefully that all makes sense.