Dear Experts,

I'm attempting to get a "concertina" effect like this (if my ASCII art works):

+------+
.\........\
..\........\
../......../
./......../
.\........\
..\........\
...+------+

i.e. I already have a fairly complex "flat" image, and now I want to make it look like it has folds in it by shearing bands alternately left and right.

I think I should be able to do this by using "object to pattern" on my original flat image, then drawing three rectangles, setting the pattern as the rectangle fill and adjusting the pattern origin handles so that they line up, and then transforming the rectangles to do the shear.

Unfortunately, when I try this I see a few patches where the pattern is visible and correct but most of the rectangles are empty. Which areas are visible changes fairly randomly as I adjust things.

My suspicion is that it may be something to do with the complexity of the content that I am using as the pattern. In particular, I have used object clipping to give a custom outline to the content. I suspect that the clipping is doing the wrong thing after it has been through the use-as-pattern machinery and is clipping more than it should. But that's just an uneducated guess. Maybe someone could make some sense of this assertion that I see: "File display/nr-arena-item.cpp line 308 (?): Assertion item->state & NR_ARENA_ITEM_STATE_BBOX failed". Any ideas?

This is with the Debian package of Inkscape 0.46. Any suggestions as to what I might be able to do to work around this would be much appreciated.

Thanks, Phil.

endecotp wrote:Maybe someone could make some sense of this assertion that I see: "File display/nr-arena-item.cpp line 308 (?): Assertion item->state & NR_ARENA_ITEM_STATE_BBOX failed". Any ideas?

a google search reveals a confirmed bug ;-)

I just opened the example attached to that bug report with a most recent build of Inkscape from SVN and get the same console messages about NR_ARENA_ITEM_STATE_BBOX. This means the bug is still present in the prerelease version of Inkscape. The bug report itself hasn't seen any recent activity since 2008-01-09 - so I doubt there will be a fix in Inkscape 0.47.

If you have any new insight what causes this failure you might consider to add a comment there - 'bump' it up/back to maybe get some attention by whomever recently worked on 'display/nr-arena-item.cpp' ;-)

hth, ~suv

Thanks for that. My search just found the patch where that assert was added.

It looks like the bug is just in the display; I can save and view using rsvg-view and it's OK.

I'm now making clones and clipping them individually, which seems to work.