From: Warren Young
Date: November 11 2009 5:58pm
Subject: Re: MySQL++ with MinGW GCC 4
List-Archive: http://lists.mysql.com/plusplus/8766
Message-Id: <1D57C31D-5F67-4B48-9751-85F6A86FCCE2@etr-usa.com>
MIME-Version: 1.0 (Apple Message framework v1076)
Content-Type: text/plain; charset=windows-1252; format=flowed; delsp=yes
Content-Transfer-Encoding: quoted-printable

On Nov 11, 2009, at 6:51 PM, Wahyu Arm wrote:
> Yes. It does require at least GCC 3.4.5, however. Older versions =20
> don=92t build MySQL++ correctly. Also, as of this writing, GCC 4 is =20=

> still considered experimental on MinGW. We=92ve had reports of GCC 4.x =
=20
> not handling MySQL++ exceptions properly on MinGW, for instance."

Oopsie.  I searched for "GCC4", no space.

> I am confused by the sentence, whether there is a problem with mysql =20=

> + + library or MinGW.

The statement was true when it was written, but that was back when GCC =20=

4 was still experimental for MinGW.  As I said, that changed about a =20
couple of months ago, so it's possible it's still the case that it =20
doesn't work, but I would hope they'd figured out the exceptions =20
across DLL boundaries issue before pushing it to release status.

I wouldn't take your cross-compile tests as conclusive.  Build on a =20
real Windows box, test on a real Windows box.  If MySQL++'s own =20
examples fail in that situation, then I'll believe they still have a =20
problem.  In that case, you'd want to take the problem up on the MinGW =20=

mailing list.

It might also be educational to try this on Cygwin.  Those two teams =20
share development resources, and are working together on the GCC 4 for =20=

Windows effort.  I find that the Cygwin side gets things deployed =20
faster, however, so maybe it works under Cygwin, and all that's =20
wanting is a release push from the MinGW folk of the same basic code.=