143196Re: [mach1mach2cnc] Mach3 trajectory testing (from a linuxcnc guy)
- Feb 12, 2014I dont really know the details and dont have time to properly look at
the code, but in some instances linuxcnc and mach3 treat "wrong" code
There was a thread short while ago that demonstrated this very well.
Maybe 2-3 months ago..
Corner rounding was involved as I recall ...
and the basic problem was that the code had impossible rapids when
looked at from the possibility of the machine in question.
Both linuxcnc and mach3 will run such code, but the results are different.
The original premise from the OP was that linux did this "better" and
mach3 did this worse.
When the actual case seemed to be, very clearly, that the code itself
was wrong and both just treated it "some way".
That thread could have something do do with guitar bodies, iirc.
In my opinion, when the code itself is in error, there is no better or
worse involved, necessarily.
But I do understand that if whatever cam system spits this out, and this
runs better (through wrong) on another system (still wrong, but
different) it seems very tempting to call one worse and one better.
-hanermo (cnc designs)
- << Previous post in topic Next post in topic >>