Re: Irrelevant Gcode?!? wasRe: Interface. Gauging interest.

I'm with the other guy here, but from a different direction. Gcode is
a constraint that keeps the machine from doing everything that it
could do rather than an enabler. It was designed in a time when the
controller boxes had much less compute power than the Silicon Graphics
CAD stations where the parts were designed, so it simplifies basically
everything.

Its only staying power is inertia of installed machinery and CAD/CAM
software. But that translates to money, which I do respect.

Anyway, a machine controller should be able to take multiple CAD file
formats directly along with input about ambient temperature, cutter
dimensions and type, and stock alloy. Then scan the work surface for
workpiece placement, establish fiducials for multiple operations, then
map the finished part into the block of steel. Like a sculptor sees
the finished head of Caesar in a block of marble, the controller sees
the part inside the steel. THEN it carves the part in the most
efficient manner possible. When finished, it then picks up a precision
grinder, and in conjunction with the same laser scanner that saw the
block proceeds to give it a high polish to micron tolerance (or
whatever finish you tell it to give).

And it should be able to do all this with the processor in my PDA.

So much like the Visual BASIC macros in an an Excel spreadsheet
thinking about G-code as anything other than a necessary evil is a
mistake.

Seriously, no offense to people making their living off G-code. Feel
free to say bad things about the TRS-80 if it makes you feel better.

On 5/17/06, ballendo <ballendo@...> wrote:
> Phil,
>
> You're off the mark on this one IMO.
>
> Those "irrelevant" items you mention are QUITE relevant in today's
> modern machine shop. Even one with 50 thousand buck CADCAM...
>
> AND, also in a one man shop with cheap or NO cadcam. In fact
> those "irrelevant things like alternate COsys's are what allow that
> one man shop to compete with the big guys...
>
> Grab Peter Smid's book on Gcode programming, and anything by Mike
> Lynch. CNCconcepts.com
>
> Ballendo
>
> --- In CAD_CAM_EDM_DRO@yahoogroups.com, "Phil Mattison"
> <mattison20@...> wrote:
> >
> > Hi Chris,
> >
> > I think I said I didn't expect it change any time soon. I said the RS274NGC
> > spec is a piece of crap. It is obviously from a time when most CNC code was
> > written by hand, like the days when my dad wrote financial programs in
> > assembly language. It is full of constructs supporting things that are
> > largely irrelevant now, like alternate coordinate systems for array
> > machining, automatic tool radius compensation, etc. G-Code as a convention
> > (i.e. G0 G1, etc.) is as good as any equivalent, but the collection of
> > capabilities is mostly obsolete because most of what was created to simplify
> > manual G-Code writing is no longer necessary with all the automatic code
> > generators already available. Whoever produces the desired results in the
> > cheapest, easiest manner will ultimately win, tradition notwithstanding.
> > --
> > Phil Mattison
> > http://www.ohmikron.com/
> >
> > ----- Original Message -----
> > From: lcdpublishing <lcdpublishing@...>
> >
> > > Hi Phil,
> > >
> > > Not sure what you are basing this opinion on, but, don't expect RS274
> > > (or simply put, G-Code programming) to go away. The most complex
> > > shapes that are machined today are done with G-codes on ordinary CNC
> > > machines all over the world.
> > >
> > > The G-Code method of programming has been around since the beginning
> > > of CNC, and will remain there, probably forever. There really isn't
> > > anything lacking with it, nor is there anything that can't be machined
> > > with it.
> > >
> > > There have been a number of attempts over the years to "change the
> > > world", most have failed, some have done okay. For example, at one
> > > point in time, it was thought that programming on the shop floor was
> > > the best way to get things done. During that phase, control mfgs. came
> > > out with various "Conversational" controls which were somewhat of a
> > > CAM system on top of the basic G-Code interpreting control.
> > >
> > > While this method of programming is still around and still applicable
> > > in short-run shops, production shops frown on it as being very
> > > inefficient. I mean really, you have a $150,000.00+ machine tool
> > > being used as a $5,000.00 CAM system.
> > >
> > > As to the various flavors of G-Code programming, there have been many
> > > attempts to change and it seems as though every control mfg puts their
> > > own little spin on it. But it's still the basic G-Code and cartesian
> > > coordinate system that gets the work done on machine tools.
> > >
> > > So, while you may think it is "Crap", the leaders in the machine tool
> > > industry put their money on G-Code programming as THE method of
> > > programming CNC machines.
> > >
> > > Don't expect it to change anytime soon.
> > >
> > > Chris
> > >
> > > > After a careful reading of the various flavors of RS274 it looks to me like
> > > > RS274NGC is pretty much an outdated piece of crap when it comes to machine
> > > > control, surviving only by virtue of familiarity and inertia. Not that I
> > > > think it is going away any time soon, but it certainly doesn't look like
> > > > golden handcuffs if you're interested in the general motion-control market.
> > > > I suspect it may not really have much of a strangle-hold on the hobby CNC
> > > > market either, since that will have many new-comers if it is in fact growing
> > > > rapidly.