CUTPOLY, CUTPOLYA, CUTFORM, in AC 9

[Anonymous]

I have several objects I made in AC 6.5 which I updated to AC 8.1
They worked in both versions. I had occasion to use them in AC 9
but they fail to process in the 3D window and I get two error
messages "internal error in generator function at line.... of file...
(system error -1) and the library part was not converted properly.
When I open the object in the library editing environment and
check the 3D script using "check script" there are no errors
reported but when I open the 3D view I get the same
internal error.

The problem is with a CUTPOLYA statement because if I comment
out this one statement the part proccess with out a problem.
I have tried changing the "d" value, rotating the command
180°, changing the command to a CUTPOLY command, and
changing the command to the CUTFORM command
but as long as the target element is within range or the cutter
I get the same error and the object fails to show in the 3D view.

I can't use group operations because I want the cut edges
to not show.

Has anyone run into this ?
Does anyone know of a fix or work-around ?
Thank you,
Peter Devlin

[Frank Beister]

Can you post some code to it or send me the object? Without at least the CUTPOLYA-command there is less to say.

I have mentioned a crash in AC9 and AC10 in a certain parameter constallation in the CUTFORM command. Can't say which, because it was a graphical changed parameter, but maybe it has the same source.

Have you tried the simple command in a naked object?
Have you print out all variable values, which are part of the command?

AC9 is much more sensible than earlier versions. Maybe it is a script error AC didn't harm in AC8-.

[Anonymous]

Hello Frank,
Thanks for posting back.
I have used cutpolya in other objects and those objects
seem to work in AC 9.
I am going to take you up on your offer and send you the object.
This is one of several objects that have the same problem.
Thank you,
Peter Devlin

[Frank Beister]

Hello Peter,

there are sometimes problems with the enhanced precision in AC9. GS has something changed there in 9. I have a bench, working in 8 and struggles in 9. I know why, but can't fix it. In my case the tube command does not work in this way I used it in 8.

Your object:
The CUTPOLYA is correct. The TUBE too. But working together, they create inconsistence model data. I have worked this out by transfering the parts to a new object, I sent to you. It's very useful, that the poly based commands are very similar. So you can change a CUTPOLYA very fast to a prism_ to see if the coordinates work.

I have replaced the TUBE-command to a block (by PRISM) to get the script run. I think you have created the tube by polyline or profiler. You should give it a try to script it native by the enhanced status codes (1000 e.a.). Then you can adjust it maybe by RESOL or TOLER.

HTH.
Frank

[Anonymous]

Hello Frank,
Thank you very much for having a look at my object.
You seem to be saying that it is the interaction of the
TUBE command with any of the three "cutting" commands
that causes a problem in AC 9.
I will definitely check this out by scripting some simple TUBE
commands and see if I can cut them with one or all of the
"cutting" commands.
The TUBE command I used in my script was made by Profiler
and the code was copied and pasted into my script.

As far as my object is concerned, I want to ease the outside
edge of the deck flange so I will probably try HPRISM_
and see if AC 9 has a problem cutting that.
Come to think of it, HPRISM_ has the edge hiding capabilities
of PRISM_ so I don't have to cut out the middle of it anyway.

Again, thank you very much Frank.
As usual, you have helped me understand GDL.
I believe it was you who, 6 or 7 years ago, explained to me
why I could not write something like,

if X then Y=something AND Z=something.
It is the "logical" AND.

Peter Devlin

[Frank Beister]

Peter, HPRISM_ seems to be a good idea. I think it is a special constellation of TUBE and CUTPOLYA in your case. Deleting the TUBE or the CUTPOLYA made the error gone. In the fixed object I replaced the TUBE just with a PRISM_ and it worked (without rounded flange).

I remember the IF-THEN question, but neither that it was you, who asked for, nor it is so long ago. 😉