Re: Re: Re: G92 was Re: another EMC Glitch?

Hi Jon

I think that we have had most of this discussion before. My compalint
was that the subject said "another EMC Glitch." I was trying to point
out that g92 going away when the emc ended a program using a normal last
line was not a glitch -- it is the writers intended function of the stop
code with the new interpreter.

The issue that you raise about not liking its doing that is a complaint
about the version of rs274 that Tom wrote and the way he interpreted the
sources that he used to prepare the logic of the interpreter. We can
complain all we want about that but it isn't a glitch. It is the product
of the fact that rs274 is only a recommended standard and when it was in
it's hayday every control manufacturer that used it tried to twist it to
their own advantage.

I'll try to respond to a couple of your issues in the snipped up post
below.

>    From: Jon Elson <elson@...>
> Subject: Re: Re: G92 was Re: another EMC Glitch?

<s>

> Well, G92 is pretty important.  Some companies have fixtures that sit
> on a machine for years without being moved.  When the machine operator
> comes in and powers the control on, he homes the machine.  The G92
> offset is read in from the emc.var file (equivalent of parameter
> storage on dedicated CNC controls) and the reference position of the
> fixture is automatically set as the offset from home.  So, the operator
> doesn't have to spend 15 minutes fiddling with dial test indicators,
> coax indicators, edge finders, or whatever.
>
> The G55-59  offsets can be used to hold the offsets to multiple parts
> held in a multi-part fixture, so the same program can be run on the
> multiple parts.

Actually they can be used for anything that you want in the way of
offsetsetting. There is nothing that says that you can't use them unless
you have a bunch of vices on your table. I've got a stop that I can bolt
to mine and set one offset for the stop on the right side and one for the
left.

> The G92 is cleared by an F6, so I am careful to never hit that.  It
> probably should ask before clearing the offsets, though.

Yes it will. Unless those g92 offsets have been written into the var
file and then they will persist through a reset.
<s>

> I simply DON'T CARE what any manual says!  AND, I will tell you that
> there is ONLY one way to change these on many controls.  Either the
> program or the user in MDI commands them to change.  I can
> categorically state that NOTHING, including powering off and back on,
> will clear the G92 offset in an Allen-Bradley 7320 control.  (Reloading
> the executive or powering off the memory backup battery will clear it,
> of course.)  You can change it explicitly, of course.
>
> Anyway, if this becomes standard behavior for EMC, I will either fix it
> myself, or stop using (new versions of) EMC, as it will make it
> completely impossible for me to do a 2nd part.

And you have a right to your opinions. And you know enough C to fix
emc/src/rs274ngc_new/rs274ngc_pre.cc so that it does not dump them when a
program ends normally. All you have to do is rename it rs274jon and
recompile. That's the beauty of open source. If this was a proprietary
system that you bought there would be little that you could do.

The other thing that we could do is agree to a majority rules and all
post a vote to emc@... and see which way g92 falls.

Peace.

Ray