Dinsdag 17 mei 2005, 19:09
"(...) an e-mail in which a test is cancelled is more likely to reach all concerned than one announcing it (...)"