Menu

#8 [Java] Bug in outline code - related to svn revision #620

open-rejected
None
5
2007-06-15
2007-06-07
Anonymous
No

There seems to be a bug in the preview pane (and maybe in the actual RepRapping as well, but I cannot check that). It sometimes forgets to draw a segment from the outline of a part. When you produce the part again, the missing segments are somewhere else again. But the hatching is always spot on. There is an example of what I mean here: http://blerrik.googlepages.com/missing-segments.JPG. It is a picture from the test piece, but the same problem happens with other pieces as well, and only on the outlines. Oh, one more thing, the missing segments appear random, and change places when you reprint the piece, but they appear on exactly the same spot when you restart the software. So it seems to be related to Java's random function with a fixed randseed at startup.

Jonathan confirmed the issue and figured out it is related to svn revision #620, by Adrian. The svn log says:

> ------------------------------------------------------------------------
> r620 | adrian-bowyer | 2007-06-04 12:59:28 -0700 (Mon, 04 Jun 2007) | 5 lines
>
> Polygon outlining changed to start each layer at a random vertex
> of the polygon(s). That way the start and stop point don't
> build up splongy bits in the same place later after layer
> on the boundary.

--Blerik (blerrik@gmail.com)

Discussion

  • Jonathan Marsden

    • assigned_to: nobody --> adrian-bowyer
     
  • Adrian Bowyer

    Adrian Bowyer - 2007-06-15

    Logged In: YES
    user_id=1278673
    Originator: NO

    That ain't no bug! That's a feature!

    A build parameter that you can alter is the distance before the end of a profile that the extruder stops being driven. The extrudate continues to emerge, and fills the last few mm "by inertia". The simulation window mirrors this by stopping plotting as soon as the head is turned off, which gives you a lot more information about what's going on.

    I guess we could plot the last bit in a different colour, rather than leaving it blank.

     
  • Adrian Bowyer

    Adrian Bowyer - 2007-06-15
    • status: open --> open-rejected
     

Log in to post a comment.