From: John Maddock (john_at_[hidden])
Date: 2006-06-04 11:12:58
> Several TR1 tests are failing due to an old workaround that took swap
> out of namespace std::tr1.
> The latest compiler is much more compatibile the the overloads for
> complex math, so updated that workaround too.
> There were a few occurrences of the problem where BCB does not allow a
> leading :: following the typename keyword. These seem to be the
> exception, rather than the rule, in the files so I simply removed the
>>> for all compilers, rather than insert another borland workaround.
Yep, I usually remember not to do that, but sometimes they slip in.
> Diffs follow against 1_34_0 branch. OK to commit?
> [mainline, then merge to branch]
Almost certainly, but I don't see the diffs attached?
Boost list run by bdawes at acm.org, david.abrahams at rcn.com, gregod at cs.rpi.edu, cpdaniel at pacbell.net, john at johnmaddock.co.uk