|
|
Boost : |
From: Daniel Frey (daniel.frey_at_[hidden])
Date: 2003-02-05 11:01:33
Dave Abrahams wrote:
>
> On Wednesday, February 05, 2003 8:37 AM [GMT+1=CET],
> Daniel Frey <daniel.frey_at_[hidden]> wrote:
>
> > That given, a member-function-pointer should match is_class, shouldn't
> > it? In fact all pointers should IMHO match is_class with this
> > implementation. Maybe it's just me but the boost source is feeling more
> > and more unmaintainable given the extrem use of MACROs to workaround
> > each and every problem some compilers have. Am I the only one who feels
> > uncomfortable with it?
>
> Using MACROS to work around compiler bugs at the low level of type_traits is
> far preferable to littering high-level components with compiler specific
> workarounds.
OK, it's not just the MACROs. My question is: Looking at the current
implementation of is_class (the work-around version), how are pointers
not detected as classes? It seems to be some magic I'm missing, the code
is not doing what it seems to express. And I had this feeling lots of
time when looking at boost code.
Regards, Daniel
-- Daniel Frey aixigo AG - financial training, research and technology Schloß-Rahe-Straße 15, 52072 Aachen, Germany fon: +49 (0)241 936737-42, fax: +49 (0)241 936737-99 eMail: daniel.frey_at_[hidden], web: http://www.aixigo.de
Boost list run by bdawes at acm.org, gregod at cs.rpi.edu, cpdaniel at pacbell.net, john at johnmaddock.co.uk