re:Re: Re: gcode comments

Ray,

(From a recent thread) If one control says G70 is a finishing pass
(lathe canned cycle) and another says G70 is programming units are
inches; are they offspring or cousins?

>Offspring as opposed to cousin might better express the
>relationship.

>I don't believe that it is appropriate to speak of MSdos and CPM as
>variants.

Some people believe the reason the digital research guy (don't
remember his name just now) who invented CPM shot himself was because
he felt "ripped off" by DOS, even tho MS put a middle "ripper-offer"
in their purchase lineage. Anyone looking at both CPM and DOS can see
the similarities (beyond obvious,IMO)

I was trying to point out that languages evolve from common
ancestors. At any point in time there will be BOTH offspring AND
cousins...

I'd agree that mazatrol is a cousin. EMC is an offspring (a rambling,
renegade, rebel in some areas where,IMO, it's just not necessary).
It's kind of like the kid who emulates his GRANDparents, so as not to
be to much like the PARENTS! OK, but a lot of times later, the same
kid realises the parents had some good points :-)

I just want EMC to respect its' parents AND peers... Gcode is a
place where dying your hair purple (to stand out) is not,IMO, a good
thing. I'd much rather that gcode looked like soldiers in uniform,
VERY much alike! But I do want those soldiers WELL trained in Gcode
tactics!

>IMO -- If a machine can't remember where it's home is, it has a
>problem. Why add codes to the part program language to correct a
>machine problem.<snip>

I agree; losing home IS a problem. BUT... in the real world machines
aren't infallible. And programmers DO push machines too hard (out of
both innocence and recklessness). What is wrong with allowing a
person (who wants to) a way of "checking" BEFORE that "critical cut"?
Just 'cause the code is available does not mean you have to use
it... On the other hand, just because you can afford to use the
G9000 Amazyx control (servos, really top of the line,NEVER gets OUT
OF WHACK) "shouldn't" mean you lobby to 'leave off' features useful
to others with lesser controls... or other desires.

>I wouldn't want that 50 ton machine, with the 150kva spindle turning
>a face mill the size of a tractor tire, trying to recover all on
>it's own.

This is a 'red herring' argument. BUT... As long as the size and
power of the machine were taken into account, why not!? LOTS of items
with this sort of "failure potential" have 'self-correction'
mechanisms in place... And I'd rather rely on a "recovery circuit"
than a (possibly) inattentive operator... Isn't that why we have anti-
lock brake systems on cars. Even with their faults, they GENERALLY
help!

>>>IMHO - Why muddle up relatively simple g-code language with all of
>>>the great ideas that we can think up.
>>We actually AGREE on this! I think the RS-274 standard IS entirely
>>adequate. BUT, it was written with the "holes" in place to create
>>new features, as discovered or required.

>In an age of multitasking computers it should be relatively easy to
>start up a new feature in parallel rather than insisting that that
>feature be a part of the motion language.

That's fine, if you're a programmer, and have access to the source
code. How about the people who have to rely on what Doug Y, or Art F,
or Otto E, or Richard H, etc. believe is important?? And they are
trying to protect their "market share" by NOT allowing anyone ELSE to
do 'just' the kind of parallel process ADD-in you mention? I have
tried to get access to various programs' "hooks" for years... There
are many NEARLY GREAT programs, and not enough TRULY GREAT programs.

The marketplace will accept (and reward) nearly great (even just
barely good enough<g>) programs, and since that is true, there is no
real "push" for TRULY GREAT!

Open source (like EMC) solves part of the problem at the expense of
the the other part. People WILL add to open source for reasons of
need and ego, and others will benefit... But the person who adds the
code is OFTEN NOT the beneficiary of his/her work.

>Yes we should add all kinds of helps and calculators and tests to
>assist programmers and operators but that stuff needs to be
>available from the GUI not in the part program code itself.
>Much of this kind of stuff doesn't need to be available at run time
>at all.

Agreed. In fact, I have before suggested that some of the error-
checking of EMC is along this line... It error checks me right out of
some useful programming practice, like the "/" code mentioned
previously...

>And last but not least with cross platform languages it should be
>possible to write this kind of assisting stuff so that they will run
>equally well on any system that we might want to use for programming
>or operating.

Again, you gotta have access to the code, in one way or another...

>Thanks for the stimulating conversation here. I mean no disrespect.
>Ray

No harm, No foul.

Hope this helps. (more than JUST EMC)

Ballendo