[RFC] libstdc++/5734

Levente Farkas lfarkas@mindmaker.hu
Wed Feb 20 08:43:00 GMT 2002


Phil Edwards wrote:
> 
> On Wed, Feb 20, 2002 at 05:13:33PM +0100, Paolo Carlini wrote:
> >
> > On the other hand, I'm not sure about which is the current stance of the
> > project wrt this kind of extensions: should they be sooner or later
> > sequestered somewhere or not? I agree that they should not be harmful
> > (Not Our Problem :-) but they are neither part of the ISO/ANSI standard...
> >
> > For instance, STLPort enables this push_back() extension with a macro
> > named _ENABLE_ANACHRONISMS, or something like that.
> 
> I've been pushing for some form of _GLIBCPP_ENABLE_EXTENSIONS for a while
> now; it just needs more thought and planning than what I've been able to
> give to it.  I'd like to be more granular than a simple all-or-nothing
> switch, though.

at the same time I'd prefer ONE compiler option or global macro or...
to be able to switch to 100% prue c++ standard compatible and to switch
to allow the "well-known-extensions" (mostly from sgi).
in this case everybody would be happy.

-- 
  Levente                               "Si vis pacem para bellum!"



More information about the Libstdc++ mailing list