One of my pull requests is experiencing autotester failures on a particular platform, but I've been unable to reproduce the problem (it works on my environment, plus I don't have ready access to said platform).

Is there a simple way of reproducing an autotester failure in a way that can be used for debugging? Short of adding writeln's to unittest blocks and pushing it to github, that is.


T

-- 
If Java had true garbage collection, most programs would delete themselves upon execution. -- Robert Sewell