[prev in list] [next in list] [prev in thread] [next in thread]
List: kde-core-devel
Subject: Re: 0 or 0L for empty pointers?
From: Frans Englich <englich () kde ! org>
Date: 2006-06-23 12:18:52
Message-ID: 200606231218.52415.englich () kde ! org
[Download RAW message or body]
On Friday 23 June 2006 11:55, Allan Sandfeld Jensen wrote:
> On Friday 23 June 2006 13:42, Frans Englich wrote:
> > On Friday 23 June 2006 10:33, Thiago Macieira wrote:
> > > Allan Sandfeld Jensen wrote:
> > > >> But Windows 64-bit isn't. It's P64: longs are still 32-bit. Worst of
> > > >> all: Microsoft brags about it :-)
> > > >
> > > >Yes. But this is compiler dependent. All sensible compilers on Windows
> > > > still treat long as sizeof(void*) or can be configured to do it.
> > >
> > > Yes, we all know that Microsoft compilers aren't sensible :-)
> > >
> > > >I doubt KDE could be compiled by MSVC anyway.
> > >
> > > In fact, until I broke with the D-BUS merge, it was being compiled with
> > > MSVC. Just not the pre-historic versions like MSVC 6.0, that don't
> > > support templates properly.
> >
> > And I appreciate that, because now it won't be me breaking it ;-)
> >
> > Is there any C++ features one can't use in KDE in order to continue
> > support the platforms KDE do? Yes, some features should be avoided in any
> > case like dynamic_cast, but perhaps there's some template features or the
> > like that should be avoided.
>
> You should avoid templates outside headers.
You mean separating the implementation? To put the implementation for a member
in a template class outside the class definition? That seems to work just
fine with a couple of gcc versions.
Cheers,
Frans
[prev in list] [next in list] [prev in thread] [next in thread]
Configure |
About |
News |
Add a list |
Sponsored by KoreLogic