Hi Francesco,

found the problem: is -TP flag used in config/win/msvc.cf at line 15.
>
> I have removed it changing from
> CPPFLAGS = -nologo -W4 -wd4127 -Gs -I. -I$(HB_INC_COMPILE) -TP
> to
> CPPFLAGS = -nologo -W4 -wd4127 -Gs -I. -I$(HB_INC_COMPILE)
>
> and now ole works again with msvc.
> Now I'm testing other parts because actually I did only a test using
> contrib/hbwin/tests/testole.prg
> contrib/hbole/tests/testole.prg
> Both didn't work before and now both work again.

Many thanks. IMO we should try to find what gets broken in C++ mode,
rather than disabling it for this lib. Normally all our code should work in
C++ mode, and AFAICR it did last summer.

I suspect some of the compiler patches might be the culprit here,
as OLE code is normally C++ and we're doing hacks to access it
from plain C. This hack needs to be done differently from compiler
to compiler and maybe it went wrong. You're right that other
contribs may also be affected (GTWVG, hbole).

There are also a lot of warnings inside build log file about some wrong
> conversions, like (sorry if they are in Italian, but the warning should be
> clear):
> ../../expropt2.c(557) : warning C4244: '=': conversione da 'HB_LONG' a
> 'LONG'. Possibile perdita di dati.


Yes, I know about them, and I think we should deal with these when
switching to new Harbour types, I didn't want to start any cleanup
before that anyway.

PS: I will test also you patch ASAP to see if it also solves the problem.


Thanks a lot!

Brgds,
Viktor
_______________________________________________
Harbour mailing list
Harbour@harbour-project.org
http://lists.harbour-project.org/mailman/listinfo/harbour

Reply via email to