> Ah, a bug in GDSException?

Most likely.

> So that should explain the difference in messages.

Yes, it seems to stop on the second parameter when parsing the status
vector. Third and fourth seems to be missing. The only consequence is that
message is not correctly formatted.

> But the cause of the problem itself - win2000 only - is still a
> mystery, and I think it could be related to message #6726?

Might be... Though without seeing your code I cannot really tell anything.

Roman