PF>
  > But metaware relies on them, as opposed to command-line switches, too
  > much.
PF>

  This is your own fault for not getting hold of the documentation.

PF>
  > I'd much rather use "/j" for "default char unsigned",
  > then to have to use #pragma's.
PF>

  In fact, you don't usually use #pragma for that.  The usual means is ...

  

  ... a command line switch :

    hc -Hon=Char_Default_Unsigned

  So you can see why I find your criticism somewhat suspect.  (-:

  ( Yes, you can also use `pragma On(Char_Default_Unsigned)' in source
    code.  But since Borland's compilers also provide the means of
    altering command-line switches within a source file, this isn't
    anything special.  )

  Returning to the point at issue, I still maintain that people who use
  #pragma (for the things that it is used in VisualAge C++ -- namely
  calling convention specifiers) should *expect* to have to work to port
  their code.  They wrote it in a non-portable style in the first place.

  They *certainly* shouldn't complain when other C++ compilers don't
  recognise or act upon the #pragmas in their source.

  > JdeBP <
___
 X MegaMail 2.10 #0:
--- Maximus/2 3.01