Yes! That totally handled the whole thing! All I had to do after including
AWCPP.h in all my .cpp files was figure out where all the ostream&
ambiguities were coming from. As it turned out, I had included <iostream.h>
(and referenced ostream&) in one or more compilation units in which the
declarations in the standard library had been made global by my utilization
of the following statement:
using namespace std;
Very cool. Back on track.
Tom Russell
[View Quote]Edward Sumerfield wrote in message <37AAF7F8.A0A93E0 at poboxes.com>...
>The way I eventually simplified this include business is to put all the
AWCPP
>include files into one AWCPP.hpp file. By doing this I loose a little
compile
>performance, including files that aren't always necessary, but I am assured
that
>I have all the correct include files in the appropriate order. There is
nothing
>more tedious, to my mind, than chasing down include file dependency
problems.
>
>I am not sure which version of AWCPP you are using Tom but the latest
version,
>0.4, has AWCPP.hpp included into AvatarMe.C. The AWCPP.hpp file contains
the
>appropriate Avatar.hpp and AvatarMe.hpp include files.
>
>If I had a perfect object model it should tell me which classes were
dependent
>on which other classes. It just seems like so much work to go through all
that
>when one can put them all in one file. It is also a simplification for
people
>who wish to use the AWCPP.lib. They just have one include to add to their
code
>to allow them to use any of the classes.
>
>In thirteen years of programming I have never been happy with my include
file
>strategies.
>
>Walter Knupe wrote:
>
although
It
able
other
and
have
just a
is
the
a
your
layout.
before
gonna
"Remove.cpp"
>
|
