>>On the other hand, real-time control on a cheap microcontroller is
>>not as easy as it seems.
>
> That's the least of the issues. The major bottleneck is in the CNC
> software. Most of it is written for open loop stepper setups using the
> tradional I/O control port of choice. This therefore influences
> hardware design. Even though there is USB ready CNC software out there,
> why build in a USB interface when most are using the parallel port. Let
> the user get an adapter instead.
>
> What desktop CNC needs is an IBM of sorts to create a standard that
> both the hardware and software guys buy into. Somehow, a computer in
> every home (or room) is easier to visualize than a CNC setup so we're
> probably a long way from seeing anything like that.
Go look at what Microsoft is doing. Are they IBM-ish enough for you?
They call it robotics, since more people understand that than CNC. Sure,
the 'usual crowd' will complain about M$, but since they haven't managed
to come up with a standard...
Despite that, I don't really have a problem with the status quo, junked
PCs controlling machines thru the parallel port. In 5 years time I'll
think about USB, and in 10 actually do something about it. By that time
PCs will have USB5 and there will be USB2 PIC chips that actually work and
we can actually program easily.
Of course, by that stage Windows might be dead, and we'll all be using web
-based GoogleOS. Then we're screwed.
Tony