|
|
|
|
|
|
by jrockway
2440 days ago
|
|
|
Sure, but now they are a small company that has to maintain two separate codebases. Even big companies like Google can't keep their Android and iOS apps at feature parity, so something has to give and it's the early adopters with old hardware that is just not going to get any new features. Or they are going to have twice the software engineering overhead. I get why you would start from scratch with an M4. I don't get why you would maintain two different codebases commercially, though. (As for the motion control aspect, at the feedrates of 3D printers, I don't think you get anything from a faster processor. Motion control is not what's limiting print speed or accuracy.) |
|
|
I mean sure, it's not going to be a perfectly smooth transition, but I think you're underestimating how bad it was on the Atmega. There is _no_ code space left on those chips, which makes adding features nearly impossible. They'll probably just maintain the old version for as long as they can, while switching new products over to the new one.
The codebase is also probably simpler than you realize. These microcontrollers don't have the code space for programs of any significant size. Porting the peripheral interfaces over to the new controller is not that hard.
> (As for the motion control aspect, at the feedrates of 3D printers, I don't think you get anything from a faster processor. Motion control is not what's limiting print speed or accuracy.)
This is actually a real problem. It's not the basic motion control, but doing things like intelligent look-ahead or vibration compensation.
I think the technology is further ahead than you realize. It's not an Atmega running your microwave we're talking about. Those chips _are_ seriously limited, in material ways, even if you haven't run into it yourself.