[keen]

The weirdest thing happened today.

I set up to re run a part I had run many times before - when I hit the start button nothing happened - no error alert, just nothing.

I tried other programmes and they started and ran fine, but not the one I wanted!.... But it had always run fine before.....??!!!

I tried everything I could think of including running the programme in my other 1100 - same problem!

I decided to go out and mull it over lunch. A well spent hour. When I got back I put my deduction plan into action and soon uncovered the issue. Very surprising and worth revealing as someone else may get stuck with this also.

I will post what it was...... but if any of you like a diagnostic quiz.... feel free to ask questions or guess!

Keen

[keen]

Nobody wants to know?.............where are all the enquiring minds ?

Keen

[Hirudin]

I like guessing, but I don't know where to start. Seems like a program that has been run before should continue to work. Or, if it's going to refuse to work, it should throw an error or something.

Um... a new Path Pilot update was loaded and it was expecting a certain character to be at the start or end of the program. Like, maybe your program was missing the final block/empty line.

Did you accidentally delete a ")", thus turning the whole program into one large comment?

How about a nested comment (a set of parentheses inside another set (like this))? That has caused me problems wit' Mach3.

[shred]

Same here. With so little information I didn't see much point, and I get enough 'twenty-questions' from people wanting tech support on vague and poorly-described issues as it is.
My only guess was the strange-characters in comments issue, which is an odd one that appears in new versions of PP but that throws an error message.

[keen]

Well good work you two - very close.

There was no change to the programme - that was working, and then stopped. The change was the Pathpilot update. The latest version is more sensitive to bad code. There was some numbers outside a nested intro block - this was tolerated with the previous version. But on upgrade it was no longer tolerated - but did not produce error message.

Good work! Keen

[kstrauss]

Could you post an example so that we'll all be wiser? What did Tormach say? How did you troubleshoot? Is this version 1.9.3 or ...

[keen]

The intro block was:
( content )
0100 (program number)
(prog name )

Version 1.9.2 tolerated this. Version 1.9.3 did not and no error message.

I have not contacted Tormach - feel free to do that.

Trouble shooted via deleting the number - it now runs.

cheers Keen

[inov8r]

Was this code you wrote or is it from a cam program? If a cam which one?

[mountaindew]

Was this code you wrote or is it from a cam program? If a cam which one?

Sprutcam spits out some code that PP chokes on now and then. Although it does generate errors or warnings something like start and end radius error .
I have run into this issue with hsm tool path option creating paths that generate these errors. The reason to mention this is the same code NOT ever run/executed only loaded, changed the operation of the mill in such a way it scared me! When I pushed the page down button the z axis took off and didn't stop until long after the button was released. This was same for x and y axis also. If I had probe in spindle and started to manually jog around to set a offset it would have damaged or destroyed probe. Was easy to fix by taking out the hsm ops in one area and no errors and then reload code and no crazy PP mill control behavior. Again all it took was to load code not ever run it! And all it took to fix problem was load code that does not generate the error!
Did not think this was possible!
Or is my understanding of program execution way off?

[keen]

Was this code you wrote or is it from a cam program? If a cam which one?

It was unusual code I pasted together for mill turning, also partly hand written. But it ran fine in 1.9.2 - so the latest PP must be more selective - but no error message.

keen

[vmax549]

Most gcode controllers will NOT allow anything but an Oword (o100) on a line . No comments allowed on that line. Try it that way put teh Oword back and remove teh comment.

(;-) TP

[JW_55Degrees]

I think this is the thing reported as issue #1092 see:
www.tormach.com/pathpilot_tracking.html#issues

Probably related to the enhancement in 1.9.3 allowing M98/M99 Fanuc style subroutines to exist with LinuxCNC style subroutine naming and calls. The M98/M99 capability is pretty handy for running code from other systems.

JW

[shred]

If it's the bad-code pre-processor (code won't load, no preview, etc), there will be a message about not being a proper file displayed on the screen when you try to open the file, but nothing in the status window.