Julia,
I made all the patches to let it compile clean on MS VC 5.0
I reworked some things with the order of the wonderful Microsoft header files so that the Q_LOG can remain
compiled in yet the driver compiles clean.
I'll put the latest stuff on the web site still until I hear otherwise.
Regards,
Byron
Julia A.Case wrote:
> Quoting Byron Nikolaidis (byronn@insightdist.com):
> > Julia,
> >
> > I would like to make it compile on 5.0 also, I just have a couple of questions about this patch:
> >
> > > - result = SQLGetData(hstmt, 1, SQL_C_SLONG, &self->lobj_type, sizeof(self->lobj_type), NULL);
> > > + result = SQLGetData(hstmt, 1, SQL_C_LONG, &self->lobj_type, sizeof(self->lobj_type), NULL);
> >
> > SQL_C_SLONG is the ODBC 2.0 type to use and should be in sql.h; SQL_C_LONG is a ODBC 1.0 type.
> >
> Perhaps there was another missing <sqlext.h> ???
>
> > > -#define Q_LOG
> > > +// #define Q_LOG
> > >
> > >
> >
> > Why would this cause a problem? This is to compile in the communications log feature which is quite
> > useful and was wanted for a long time. If it is commented out like you have shown it, it could not be
> > activated from the setup dialog boxes.
> >
> > Do these things cause warnings or errors?
> >
> the va_args macros cause errors... and to be honest I don't know
> enough to fix them, so it seemed simpler to comment them out. Will that
> many END users be using Q_LOG? I'd think that would mostly be a developer
> thing... and if so, a developer could uncomment it if they wanted it.
>
> Julie
>
> --
> [ Julia Anne Case ] [ Ships are safe inside the harbor, ]
> [Programmer at large] [ but is that what ships are really for. ]
> [ Admining Linux ] [ To thine own self be true. ]
> [ Windows/WindowsNT ] [ Fair is where you take your cows to be judged. ]