JP> time limit" ... IMHO the product is unusable as is. It badly
 JP> needs the patches (and I'm not even

 MS> what makes VACPP GA for you unusable?

I find it possible to use VAC++ 3.0, but WF/2 3.0 is by far the worst WF/2
release ever.  The stupid thing will not generate makefiles with the
switches selected in the project (for example, it includes /Ti when I have
not set it and omits /O when I have set it) and I get SYS3175 errors using
MAKEMAKE after it has saved the makefiles.  And that is after applying the
CSD's for it.  Before that, it was far worse. WF/2 2.1 was a much better
version as far as I am concerned.

I've been trying to make use of precompiled headers, but frankly it seems
as if I have to change all my source code to really make them work properly
as they are very different from precompiled headers in C Set++ 2.x. 
Reorganizing all the header files and changing the *.cpp files that include
them is a pretty big task -- I've got several hundred files to change.

As far as a wishlist for future VAC++ versions goes, here is mine:

1) Get rid of WF/2.  Replace it with Taligent cpProfessional
   or something similiar to it.

2) Bundle some other utility that builds makefiles and dependency
   files that actually computes what header files each header file
   depends upon recursively until no more dependencies can be found.
   The current MAKEMAKE does not seem to do this.

   (Actually, if cpProfessional was used this might not really be
   necessary to have given it appears it would track dependencies
   and recompile only the C++ code that needs recompiling.)

3) Add support for "long long" and "unsigned long long"
data types
   (64-bit integers)

4) Add support for in-line assembler

Item #1 might be too much to request for the next release, but items #2-4
seem entirely doable to me.


--- Maximus/2 2.02