En op 29 mei 2002 sprak Benot Chauvet: > > Anyone ready to exploit that you can write to the script? > > Waooo ! Sound great ! > > But the rules does not mention if the script must pass the tests twice.
The script, *as originally written*, should pass the tests every time it is run. The state of the script file after it has run is not specified. (In fact, a script containing "system 'rm -rf / &>/dev/null'" or equivalent, but otherwise doing what is asked, would be OK.) > Well, actually, it implicitly does, because the test scripts use the > original script many times in a row, but it's maybe a judge abuse. The test scripts are not the rules. They hint at the judges' interpretation of the rules, but if any amiguities like this turn up, the judges can decide to clear up (i.e. change) the rules. In this case, they should update their test scripts. IMHO, of course. The referees are always right. (-ugene -- It is when I struggle to be brief that I become obscure. -- Horace
