--===============1534075491==
Content-Type: multipart/signed;
  boundary="nextPart1332607.ih9ejSQzuz";
  protocol="application/pgp-signature";
  micalg=pgp-sha1
Content-Transfer-Encoding: 7bit

--nextPart1332607.ih9ejSQzuz
Content-Type: text/plain;
  charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable
Content-Disposition: inline

On Monday 23 May 2005 18:19, Paulo Moura Guedes wrote:
> On Monday 23 May 2005 16:02, Jens Herden wrote:
> > On Monday 23 May 2005 21:08, Paulo Moura Guedes wrote:
> > > On Friday 20 May 2005 19:57, Andras Mantia wrote:
> > > > Paulo,
> > > >
> > > > =A0please, please DO NOT USE "assert" in Quanta!
> > >
> > > You know i have a diferent view but ok, you are the boss :)
> > >
> > > What I don't understand is why your request is related to this
> > > bug. An include was missing...
> >
> > Wasn't the include assert.h ? Why would one miss this include?
>
> Andras doesn't like std::assert because it crashes the application
> (although one can use NDEBUG).

Right and because of the bug I posted. I suppose it compiled for you=20
without the include, but does not compile of others. This is why it is=20
better to avoid STD and other things not supported on every=20
platform/compiler. What is good about Q_ASSERT and Qt is that it is=20
guaranteed to work on all supported platforms under the same condition=20
(unless it is documented otherwise, but the exceptions are for=20
Windows).

> But in this case, an include was=20
> missing. If one use Q_ASSERT and doesn't #include <qapplication.h>,
> it will have the same problem :) It is not "assert" specific.

Ok, so I can guess two things:
=2D you committed untested /not compiled code, which is not good
=2D you compiled, committed but it still misses the include, which proves=20
what I said above.

> I don't know why but I can always compile without #include <cassert>,
> so I keep forgetting :)

Because you have a different compiler. Even g++ !=3D g++ as distributions=20
tend to backport patches from g++ CVS HEAD and not always the same one.

And of course, crashing the application is not nice. If you really want=20
to use assert, I also request to add -DNDEBUG or whatever to the=20
Makefile.am. You can enable the crash for yourself. ;-)

Andras

=2D-=20
Quanta Plus developer - http://quanta.kdewebdev.org
K Desktop Environment - http://www.kde.org

--nextPart1332607.ih9ejSQzuz
Content-Type: application/pgp-signature

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.5 (GNU/Linux)

iD8DBQBCkhrMTQdfac6L/08RAqPOAJ42qpq3sF4j306KjjzUJtBZuK9uvgCghA+2
VTHBbQH8cvOq4cOx7XEfGvc=
=BZri
-----END PGP SIGNATURE-----

--nextPart1332607.ih9ejSQzuz--

--===============1534075491==
Content-Type: text/plain; charset="us-ascii"
MIME-Version: 1.0
Content-Transfer-Encoding: 7bit
Content-Disposition: inline

_______________________________________________
quanta-devel mailing list
quanta-devel@kde.org
https://mail.kde.org/mailman/listinfo/quanta-devel

--===============1534075491==--