El 8/10/24 a las 1:16, Gregor Riepl escribió:
That doesn't look like a related bug to me, and I didn't encounter it when I prepared and tested the patch.Maybe some other dependency changed in the meantime?
Maybe. Build-dependencies are changing all the time and breaking things.
Anyway, this should be an easy fix, just add a #include <random> near the top of the failing unit test source. I'll have a go, but it could take a while.
Ok, I'll file a different bug. Thanks.