Changing the coordinates of a path segment does not update the automatic control point for the next smooth cubic curveto segment.
Created attachment 21900 [details] A testcase We have a path with a move to and a smooth curveto segment. The move to is moved by a timed javascript, showing that the automatic control point for the smooth curveto segment is not updated.
Our behavior is not identical to FF. I can't get Opera to launch on my machine... :( From the test case it's not very easy to tell if something is wrong or not... but I do think we might have an issue here. A better test case would make it very clear if the behavior is correct or not (by showing only green on pass and red on fail, or showing some pass/fail text, etc.)
(In reply to comment #2) > Our behavior is not identical to FF. I can't get Opera to launch on my > machine... :( From the test case it's not very easy to tell if something is > wrong or not... but I do think we might have an issue here. A better test case > would make it very clear if the behavior is correct or not (by showing only > green on pass and red on fail, or showing some pass/fail text, etc.) Opera behaves the same as FF here. The problem is that we expose only moveto, lineto and cubic curveto in this interface, not the smooth one. Likely the right fix would be to do expose them (lazily), and internally use some quick format. I wonder if there already is a bug for this... Cheers, Rob.
Ok, a bug that is slightly related that I was hinting at is bug 10827. I am having a look at fixing that as it should solve this problem AFAICS. Cheers, Rob.
*** This bug has been marked as a duplicate of bug 49580 ***