[boost] Re: Formal Review: Indexed Set

From: Joaquin M Lopez Munoz (joaquin_at_[hidden])
Date: 2004-03-22 16:07:47

Brian McNamara <lorgon <at> cc.gatech.edu> writes:
> 
> On Sun, Mar 21, 2004 at 02:38:23PM +0100, JOAQUIN LOPEZ MU?Z wrote:
> > De: Brian McNamara <lorgon <at> cc.gatech.edu>
[...]
> > 
> > Currently, the approach is that the first index with
> > a given tag is picked [...]
> > I'm not convinced the approach you suggest is more
> > convenient: [...]
> > A static *warning* would be ideal here :)
> 
> A warning is better than nothing.  I still prefer an error.
> 
> I don't buy the meta-programming argument.  I think it's far more
> important to automatically detect an error for "normal" users (who are
> probably the majority of users) than to ensure that we don't
> "inconvenience" metaprogrammers (the minority).  (Metaprogrammers are
> smart enough to be able to generate their own unique tags, anyway.)
> 

Well, I don't have a strong feeling about this. I
agree that the situation is in most cases an error,
so if nobody else comments on this, I'll add a
BOOST_STATIC_ERROR.

Joaquín M López Muñoz
Telefónica, Investigación y Desarrollo

PS. Have you had a chance to take a look at the review
notes? Comments on naming issues raised there are
most welcome.