Date: 2001-08-23 05:36:40
> Hmm, perhaps tie() should just be removed from utility?
I think this was the conclusion made during tuple review.
>Though I suppose
> this means updating #includes that currently point to utility...
> or should
> utility include tuple?
It is of course possible to add some #ifdef guards for the two
argument tie function (if utility.hpp has defined 2 argument tie, do
not do it again in tuple, and vice versa), but there are some
differences between the tuple tie and utility tie as
tuple tie returns a tuple, and utility tie a 'tied' object.
> fresl> PS. There may be some other problems with tuple library:
This was due to some missing typenames, and is now fixed.
Boost list run by bdawes at acm.org, gregod at cs.rpi.edu, cpdaniel at pacbell.net, john at johnmaddock.co.uk