Hi Przemek,
Il 23/03/2009 23.55, Przemyslaw Czerpak ha scritto:
No. It was not even my intetion to make any fixes. I just simply
corrected casting to pacify MinGW build warnings. Nothing more.
This whole code is waiting for someone who will rewrite it from
scratch. There are many things inside which has been never working
correctly. It's also not MT safe and should not be used in MT
programs.
I hope that some windows developers will work on it.
Sorry but I do not know anything about OLE and using it. I do not
even use Windows so I cannot easy make basic tests.
BTW does this code work with current BCC and MinGW builds?
The problem is related to -TP flag of MSVC (in config/msvc.cf) that
force to build all in cpp mode. Removing it ole works as before.
Tested with BCC and it works. But AFAIR there is no changes in BCC flags.
My question was if you have in mind what can change in code, probably
only in CreateOleObject() or GetOleObject() functions, between c and c++
mode that results in error and I am thinking that is related to
different type handling.
Is this problem local to MSVS?
It's seems so.
Does anyone test it with OpenWatcom?
best regards,
Przemek
ps. I see that you added:
hb_itemPutTD( pItem, pVariant->n1.n2.n3.dblVal + ( double ) 2415019 );
to store VT_DATE values. It means that all VT_DATE values read from OLE
structure will be converted to Harbour timestamp items.
Is it intentional?
Actually yes, also if I'm still doing tests.
Best regards,
Francesco
_______________________________________________
Harbour mailing list
[email protected]
http://lists.harbour-project.org/mailman/listinfo/harbour
