This is my use case: the page I'm testing returns CSV file. SimpleTest tries to parse it and runs out of memory (btw, my memory limit is 1 GB :). It may be sound idea not to parse non-HTML (maybe XML too) content, just represent it as a large string, so text assertions will still work, since they may be useful still---for example I'm checking for PHP errors regexp, even in CSV output...
2008-06-09 04:45:00 UTC in SimpleTest
