Joshua A. Andler wrote:

Hey all,

https://bugs.edge.launchpad.net/inkscape/+bug/202540 appears to pick up on a handful of extensions that are now misbehaving due to a change in our internal handling of path data.

Is the latest patches approach the correct one as far as everyone thinks? I'm relying on people w/ more experience with our extensions (and changes in path handling) to give input on this.

I've been tracking this and this is already quite an old issue, and definitely not something to rush into right before a release. The current patch might be acceptable to put in this release if we were sure it didn't break any extensions (it very likely does), but I think I'd prefer to wait until 0.48 is out, put it in and then start thinking about a better fix (using the 2geom bindings for Python?).

BTW, it's not about a change in our internal handling of path data, in most extensions there simply is no concept of a closed path as the "CubicSuperPath" just converts each and every segment into a cubic bezier (including a z/Z segment).