Material is deleted when cloning

Dan Rathbun

Anyway. save the name of the material, and perhaps also it's texture filepath, if it is not yet IN the model.

Test any material reference for validity before using it. If there is no current material nil may be returned, which evaluates as false.

if matl && matl.is_a?(Sketchup;;Material)
  # go ahead and use the matl reference in code
  matl_set = Sketchup.active_model.materials
  if matl_set.include?(matl)
    # do something
  end
end

Dan Rathbun

The Ruby methods clone and dup, do NOT yet work upon SketchUp API classes.

We discussed this in previous topics:

Marc477

Thank you, I fixed my problem by processing only materials used in the model:


aMat = Sketchup.active_model.materials.current
if aMat != nil and Sketchup.active_model.materials.include?(aMat)
   ...
end

tt_su

@dan rathbun said:

The Ruby methods clone and dup, do NOT yet work upon SketchUp API classes.

We should improve this. But they would not make sense for Entity based classes as instances are persistent and have to be created via entity.add_*.

tt_su

@dan rathbun said:

I just did the following in SU2014 and got a SPLAT!

Choose a material that is not in the model.
Open console.
test = Sketchup.active_model.materials.current test.name
Choose a material that IS in the model.
test2 = Sketchup.active_model.materials.current test2.name
Everything is fine.
Now,.. try to get the name of the previous material.
test.name
-- > SPLAT

I cannot reproduce this. Did you submit the splat? Did you include any info I can use to look it up?
Do you see this consistently? (Can you do a video recording?)

Dan Rathbun

I do think I sent in the Splat Report.

I do not have video capture software installed (unless it comes with Windows 7?)

This was 2 weeks ago... I haven't tried since.

tt_su

Did you enter any details into the BugSplat? (So we can search for it.)

Dan Rathbun

I entered the steps that caused the splat, and then copied from the splat dialog, into the above SCF report.

tt_su

SU2014 Win?

tt_su

Ah! Found the crash report. And I realized I'd overlooked the last step in your instructions - accessing the name of the first material after it was not current any more. It's another bug where Ruby is allowed to get a handle to a transient material. I'm logging a new issue.

Dan Rathbun

@tt_su said:

Ah! Found the crash report. ... I'm logging a new issue.