There are problems with arc_to in the OpenGL backend (it segfaults), but more generally it would be good to have a benchmark and/or tests that ensure that backends behave sensibly for a range of inputs and/or handle ill-defined cases gracefully (eg. all three points the same).
There are problems with
arc_toin the OpenGL backend (it segfaults), but more generally it would be good to have a benchmark and/or tests that ensure that backends behave sensibly for a range of inputs and/or handle ill-defined cases gracefully (eg. all three points the same).