[boost] Re: [review results] hash functions accepted

From: Joaquin M Lopez Munoz (joaquin_at_[hidden])
Date: 2005-03-23 11:06:09

Peter Dimov <pdimov <at> mmltd.net> writes:

> 
> Joaquín M LópezMuñoz wrote:
> > The problem with this approach is that it is
> > allowed that stdlib implementors use extra template parameters.
> 
> They are not, see:
> 
> http://open-std.org/jtc1/sc22/wg21/docs/lwg-closed.html#94 
> 

So much the better! Is there any obstacle, then,
to forward declare STL containers so as to define
their corresponding hash_value overloads, instead of
including them? In the linked page it is said [quote]

"I can't think of any way that this extension could
break a conforming program, considering that users
are not permitted to forward-declare standard
library components"

Is this really so? Why cannot I forward declare a
stdlib component?

Joaquín M López Muñoz
Telefónica, Investigación y Desarrollo