There were a couple of scripts in _tests.txt which were commented out for this reason, I moved them to the new wontfix folder too. I think it's cleaner and easier mto maintain to only have tests in the "tests" folder which actually work and are meant to succeed.
(I actually spent half an hour fiddling with my first interruption script, because the tutorial on the blog doesn't mention the any command to end interrupt sequences. Then I checked the "Counterspell" script, was happy to find the "endinterrupt" keyword, and fiddled for another 30 minutes with it before I found out that this keyword is actually wrong and the parser doesn't understand it at all, it just so happens that the Counterspell script (and two others) work nevertheless because in these scripts the interruption doesn't need to end. I wholeheartedly agree with the passage in wololo's mail in which he says that we could use a "real" test suite tutorial / reference.
Also added two scripts that don't demonstrate bugs (apparently the bug reports that I used to write the scripts were either false or outdated). nevertheless it doesn't hurt to have these scripts in the test suite, it prevents us from breaking something for the respective cards in the future and not noticing it.
The script fear_i147.txt has been moved into the "generic" folder and added to the regular test sequence for the same reason.
Also fixed the script generic/tokens.txt. It was referencing The Hive once via name and once via ID. This caused problems when (unofficial) Beta edition _cards.dat was installed.
