In my test program, which draws a perfectly horizontal path, and then prints stroke and fill extents: fill: x1=32767,000000 y1=32767,000000 x2=-32768,000000 y2=-32768,000000 stroke: x1=-10,000000 y1=170,000000 x2=760,000000 y2=190,000000 It gives bogus results for the fill extents, as you can see. If the path is not perfectly horizontal then the fill extents get sane values.
Created attachment 6656 [details] test program
Gustavo, Can you put your test case under a license like GPL please? Just say so and I'll take care of the rest.
And also can you confirm that this is fixed in 1.3.14?
This was fixed in: commit a8ca155f83098c02fb8d3acc57b0492d5b753d54 Author: Robert O'Callahan <robert@ocallahan.org> Date: Mon Sep 25 23:16:54 2006 -0700 Fix stroke/fill extents bounding boxes Correctly return the transformed bounding box for stroke/fill extents, instead of just transforming the two corners separately. When you license your test case, I'll add it to cairo's test suite to make sure we never regress. Thanks.
Sure, make my test case GPL or whatever you want :)
Actually I found out about get-path-extents test and just added another case to that. Thanks.
Use of freedesktop.org services, including Bugzilla, is subject to our Code of Conduct. How we collect and use information is described in our Privacy Policy.