Re: Using MySQL from ADA program

["Juergen Pfeifer" <juergen.pfeifer@gmx.net>]

"Pascal Obry" <p.obry@wanadoo.fr> schrieb im Newsbeitrag
news:u8zcndgtf.fsf@wanadoo.fr...
>
> "M. A. Alves" <maa@liacc.up.pt> writes:
>
> > On Fri, 30 Nov 2001, M. A. Alves wrote:
> > > On Thu, 29 Nov 2001, Preben Randhol wrote:
> > > > . . . http://gnade.sourceforge.net . . .
> > >
> > > Yes, GNADE seems very nice now. . .
> >
> > . . . but, unfortunately, it won't even build.  <<File "gnu.ads" not
> > found.>>
> >
> > I must confess I am still under the impression GNADE is not working
> > anywhere outside its foundry.
>
> Certainly not. I have been able to build many versions. I've not tried the
> latest one though.
>
> Pascal.
>

Well, from time to time I try to build GNADE on machines where I only have
GNAT and the required basic GNU tools and it usually builds without
problems. I do this of course to ensure that it builds outside its
foundry;-)

Please note that the current focus of GNADE is the implementation of an
embedded SQL precompiler that is reasonable close to the ISO92 specs. All
the native bindings for MySQL or PostgreSQL are more or less placeholders
for future efforts. The ESQL preprocessor is built on the ODBC binding and
this was a good decision because we were able to interface with a variety of
different databases on different platforms without changing a single line of
source code.

Nevertheless ESQL is not the most elegant way to integrate generic database
support with Ada95, although it is a reasonable approach for smaller apps.
And ODBC is fine as long as you don't have to scale for thousands of
simultaneous connections to your backend. But the GNADE project is open for
the discussion of better approaches.

Personally my favourite scenario is still to have a GNAT.NET compiler for
the .NET platform and then simply to reuse the very rich ADO.NET
architecture for data access from Ada. My 2c :-)

J�rgen