I was looking at the Gorilla model here: viewtopic.php?f=20&p=10763#p10763
And trying to figure out why the piece in the upper right was messed up. Never figured it out, but what I found was the old phlatscript version 918 worked, while the new phlatscript v921 didn't. Maybe the beta testers can have a look at it. I just copied the part out to a new sketchup file, which is included below, and deleted all the phlatlines. You guy can use my dos batch files to quickly switch between phlatscript versions: viewtopic.php?f=98&t=1103 I have new batch files for use with the new sketchup version 7.1, which I upgraded to. If wanted, you can pm me.







Attached files Gorilla part test.skp (88.8 KB)Â

 

I confirmed the problem and wrote a ticket for it. The ticket number is 124. Thanks for pointing it out.

 

I see the problem, go to the view/hidden geometry and have a look, there are hidden lines that is messing with it. I deleted those lines and the Phlatscript worked fine

 

Jnida63, nice find with the hidden lines, but it doesn't fix the problem. I deleted all the hidden dashed lines in the slots, but still no go. I does look deceptively correct in sketchup, but the offset face is now completely missing instead of missing in some spots. So now the generated cnc cut file is basically empty. Nothing is plotted. Once again, I tried this new sketchup part with ps v918 and it works fine. One thing, I noticed was that v918 took only a second for the red outside cut line to finish processing and turn orange. v921 took longer, maybe 2 or 3 seconds.

Mark, Phlatten didn't help. I tried it both before and after deleting the hidden lines. But Flatten did do something strange. It made surfaces where the hidden lines were. At least if Phlatten ever does this, we know to check for hidden lines.

-Kwok

 

I've run into this exact same problem. It would do 95% of everything just fine and there would be one part that it would leave blank. The only way i was able to correct it was to use v918.

Was going to post something about it but you beat me to it.

 

Hi rayclark. If you got another sketch file with this problem, please post it.

 

I played with it some and I believe it is a problem with the drawing albeit I am not really sure what it could be. There are alot of lines that need welded or fixed in the drawing. Interesting how the older version works fine with it though.

 

Here is a quick explanation of why the old script works and the new one doesn't. Short story, it is definitely a bug that I introduced. Read on for the long version if you like.

Originally the script (918) would search for each specific type of cut in order and grouping was used to keep the travel times down while it completed all of each type. This meant the script asked SU for an array of everything in the model and then asked each of those items in the array if it was a specific type of cut (centerline, for example). When all of the first cut type was finished processing it would again ask SU for a list of everything and then query if each item was the next cut type. Centerlines, Fold Lines and Plunges were treated as freestanding edges while Inside and Outside cuts were expected to be on a closed loop (with a face). The trick to that version working is that it would ask SU one last time for everything and if any cut wasn't yet processed it was at that time. This picked up all the inside and outside cuts that didn't have a face.

Part of the rewrite was increasing the speed of the gcode output. I rewrote the section that loops through the items in the model so that it only does so once. I kept the idea that all inside and outside cuts should form a complete loop and neglected to add that last bit of cleanup that would have produced the gcode for the parts you are discussing.

 

gashmasher, while I don't follow completely, got a question for you. You're talking about the generation of the g-code, but I think the problem starts with the generation of the outside phlatline. When you say, "This picked up all the inside and outside cuts that didn't have a face." , I take it this means the face between the orginal line and the offset outside phlatline. The thing is the v918 phlatscript does generate the face, while the v921 has trouble with it. I do think it is something rare and probably has something to do with the original curves themselves.

-Kwok

 

I see your point and you are correct. I was talking about after all the lines are marked and gcode is being generated. I haven't had a chance to look over the file you posted but will take my foot out of my mouth and do so now

 

I looked at the file and found two issues to get it working. The first was to remove the hidden geometry Jnida63 pointed out. The second is there are 6 overlapping edges in the outside cut. I had a hard time finding the last one and it is on the upper left side. This is the same old bug we have had for awhile and the big fix for the next release.

I know it doesn't help much right now but my latest development version generated this gcode plot and I didn't have to use either of the above methods. Just selected the outside tool and clicked the part. Attached files

 

If it's already corrected in the new code, then that is great! I don't think it will effect most people.

Is that a new version of your gplot software showing the colors? I see color settings in my version, but my plot is mostly red.

 

That should be the same version of gplot you are using. I haven't had time to make any changes lately.

Using the default colors red is gcode G1 while teal and olive are the arc codes. If you used the weld tool on the part before doing the outside cut all the arc information was lost and your gcode just doesn't have the g2/3 codes.

Here is the gcode that was generated.
Attached files Gorilla part test.zip (5.1 KB)Â