Re: Re: EMC, ISA and other Questions...

>    From: "alex" <telecomt@...>
> Subject: Re: EMC, ISA and other Questions...
>
> Ok, started to understand - better late then never.
> Interpreter reads the G-code and translates it into the commands to be
> executed by the
> machine. Commands are parsed  into  C routines which control axes of
> the machine.
> Please correct me if I'm wrong.

Very close, Alex.

The interpreter builds an internal world model based on the g-code that
it has read. From this world model, it figures out what needs to happen
next in order for the model to be changed to match the new block of code.
From those changes it generates a set of Neutral Message Language (NML)
words that it passes on to whatever might be listening.

In an average EMC, there are task, motion, IO, and interface routines
also running. These listen to the NML channels and react to those
commands from the interpreter. They can also issue commands to each
other and to the interpreter. It matters little whether all of these
routines are running on one computer or several.

Yes there are ms-dos specific parts left in the code from some versions
that were compiled for that OS. My guess is that you'd have to search
them out and bring them up to date with what's happening now, both with
dos and emc.

I would guess that a real-time linux kernel could do most anything that a
dos kernel could do, including running on a 386 chip set. I saw a small
stepper hexapod that used a stock EMC and was running with a 486. The
encoder reader that you have in mind here should be easily able to run as
well with Linux as it does with dos. You are no doubt correct that if
it read four encoders at 200k ppr, it wouldn't have much left for a gui.

Ray