It depends what you mean by "screens". If you're referring to the 
different pages your script is accessing and the scripts are all running 
on the same machine, you can join your tests into a single environment 
so that the error handler can encapsulate all of them. Look into using a 
Class or Module.

If you mean you're running individual scripts on different machines, 
then you have the option of placing a log file in a shared location 
which any failing routine can append to, so that you can trace errors 
after the tests have been run, or check the file before running the 
tests to see whether this iteration has already failed and can be 
skipped.

-- 
Posted via http://www.ruby-forum.com/.