Run From Line bug

I support another user's table & software, and recently updated LinuxCNC for him so we could take advantage of the fix to material handling as described in this post . Now the material selection works as expected when using Conversational or Single Cut, but it seems a Run From Line bug related to material has been introduced.

Behavior observed: when trying to Run From Line, the material that is set in the ngc file is ignored, and QtPlasmaC reverts to the default material. If Cycle Start is pressed, an error pops up: "Operator Error: Materials Error: Material #1000000 is not in the materials list." The gantry moves to the correct XY coordinates and follows the cut path, but the torch does not come down and probe, and of course the torch is not fired.

Run from line used to work great before the update.

QtPlasmaC version: 1.236.290
LinuxCNC version: 2.10.0-pre0-2069-g9de3c86d4

phillc54 wrote:

Oops, sorry about that. A fix has been pushed, thanks for the report.

 

Is 2.9 still accepting commits?
Andy last night was building installers for the final release...

Thanks so much for the quick fix, phillc!

I pulled source and built, and installed on the machine. Run from line works, thanks again! We depend on this functionality in production quite a bit. Just one little observation / nitpick -- the G-code is now uppercase when a file is loaded, which is nice. (I support this change) However, when in RFL mode, the g-code reverts to lowercase. Clicking Reload to exit RFL results in uppercase code again.

Gcode now stays caps in RFL, thanks. I think I may have uncovered another material handling issue related to conversational last evening. I had 4 parts to cut and was able to fit them on to a scrap piece of material, but only if I cut two at a time at a different place and rotation on the remnant piece. These are the steps I took:

1. Load g-code file, nest of 2 parts (normal post from SheetCAM)
2. Cut the parts -- all is well so far
3. Used Conversational to rotate my nest 90 degrees
4. Tried to set machine X/Y zero at another spot on the scrap I was cutting from

This resulted in the same "material #10000000 is not in the materials list" error. I hit Retry and the error went away, but after this I wasn't able to get the table to respond properly until I closed LinuxCNC and re-started. I ended up doing a new nest in Sheetcam at 90 degrees to the first one and using a fresh gcode file instead.