https://bugs.documentfoundation.org/show_bug.cgi?id=173555
--- Comment #3 from [email protected] --- Hey there! Yeah, I peered at the code and that this is meant to be a bilinear warp (based on the formulas in the code). So in order to elaborate, I'm going to attach the exact same test case in LibreOffice as well as in Affinity, which contains the same warp as an option, correctly implemented. So in the new attachment you can see the result of one "warp" step in which the top left corner is brought to the bottom left ("multiplying the curve by x") successfully deforming the horizontal line to a 45 degree line in both Draw/Impress and Affinity. However, the second iteration of this process yields visibly different results. As I said before, this is because the transform does not, in LibreOffice, transform the points on the curve but rather the Cubic Bezier's control points, constructing a new curve from them. However this is not equivalent to warping the points on the curve itself. I'm sorry, I'm bad at explaining this and I hope the image explains it slightly better. -- You are receiving this mail because: You are the assignee for the bug.
