On Sat, 2006-07-29 at 16:54 +0200, Mark Wielaard wrote: > Did you also happen to look at the DiagBigDecimal test? It passed with > 0.91, but current CVS (with or without you patch) gives the following > failures: ..[snip list of 22 failures].. Yes. I didn't investigate those further because the 1.5 JRE fails the same set of tests (+4 more). The 1.4 JRE fails only 3 of them. So these tests need a bit of looking into, but I'm pretty sure that like the construct() failure, it's due to the change in toString() behaviour, and not any real bugs. /Sven
