From: Pete Becker (petebecker_at_[hidden])
Date: 2002-06-13 07:47:12
At 01:41 PM 6/13/2002 +0300, Peter Dimov wrote:
>From: "Pete Becker" <petebecker_at_[hidden]>
> > Standards are about stability. The only thing that could be done along
> > line would be to deprecate it in the next version of the standard, thus
> > warning people that it might go away in the subsequent version of the
> > standard.
>From: "Howard Hinnant" <hinnant_at_[hidden]>
> > Yes, I know of Metrowerks customers who use it, and depend on it being a
> > packed bit representation.
>The real question is: are there customers that use vector<T> in a generic
>algorithm and depend on the fact that, when T = bool, they get something
>that is radically different from the other vector instantiations?
vector<bool> is specified in the standard. There are procedures for
removing things from the standard. They take a long time, and with good
reason. If you want to propose a replacement for vector<bool> that retains
all of its current semantics you're talking about something that can be
done much more quickly. But I suspect that that won't do what you want.
But this is not the appropriate place for this discussion.
Boost list run by bdawes at acm.org, gregod at cs.rpi.edu, cpdaniel at pacbell.net, john at johnmaddock.co.uk