Started working on something very similar a few years back and gave up pretty soon for some stupid reason. Maybe I should try again, am getting better at getting stuff done.
You should. I've been doing this kind of stuff on the side for 25+ years, and nothing ever gets "done." But the educational value is much higher than you think. And the skills learned can eventually propagate out into interesting paid work.
I think of them as retirement projects before I retire. When I actually retire I'll maybe finish them.
(That said, I have in the past tried taking jobs that were adjacent to my "research" interests, and found the joy of building these things from scratch is much better than fiddling with the levers on the side of someone else's thing they built from scratch years ago. I like working on and improving production systems, but if they intersect too closely to my personal interests, it can be demoralizing.)
I know the feeling - this project, like most of my other side projects, got abandoned a couple of times. But I was really curious about implementing a GC and, for once, I managed to finish something. I'm glad I did! :-)
Started working on something very similar a few years back and gave up pretty soon for some stupid reason. Maybe I should try again, am getting better at getting stuff done.