🚀 go-pugleaf

In article <CRAIGM.93Jan23140053@egnext1.SLAC.Stanford.EDU> craigm@egnext1.SLAC.Stanford.EDU (Craig Markwardt) writes:
>
>In article <1993Jan22.012720.28932@sygnus.uucp> tpurves@sygnus.uucp (Timothy Purves) writes:
>
>   In article <CRAIGM.93Jan19132529@egnext1.SLAC.Stanford.EDU> craigm@egnext1.SLAC.Stanford.EDU (Craig Markwardt) writes:
>   >
>   >In article <C13v5G.1xn@panix.com> rryan@panix.com (Rob Ryan) writes:
>   >
>   >BC++ also has specialized Windows features which do not appear in any
>   >other version of C++ that I know of.  These are incompatible with all
>   >other C++ compilers.  Microsoft, in their MFC package, has used an 
>   >underhanded loophole in the C++ spec to accomplish their goals, which 
>   >BC++ later incorporated as well, but there are no other compilers out
>
>   What "underhanded loophole". Please explain.
>
>Sorry, I don't have the original news article any more, but it was my
>understanding that Microsoft used a rather hazily undocumented
>"feature" of C++ in their MFC package, which allows messaging to be
>done.  Borland had to modify their compiler to take advantage of this
>"feature" and use MFC.
>

Let me add some clarification.  MFC does rely on an interpretation of
an ambigous issue in the ANSI spec.  It doesn't confer any special
advantage, it just allows our message maps to have a simpler and
cleaner syntax.  Since this issue was ambiguous in the language spec,
Borland added a switch to allow their compiler to use the same
interepration.  It is in no way an "underhanded loophole."  

OWL relies of Borland's proprietary DDVTs.  The syntax of DDVTs
explicity violates the ANSI-draft standard and is therefore a 
language extension.  No other compilers currently support this extension.

The bottom line is that MFC can be used on a wide variety of compilers,
including (but not limited to) MSC and BCC.  OWL can only be used with
BCC.


-Greg
-- 
Greg DeMichillie          Microsoft AFX         
gregde@microsoft.com      All opinions are mine, not Microsoft's