--pgp-sign-Multipart_Mon_Dec_11_12:17:27_2006-1
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: quoted-printable

At Sun, 10 Dec 2006 13:51:11 -0500,
D'Arcy J.M. Cain wrote:
>=20
> Maybe this time the change will happen.  Here is a proposed change.
> Can we find out what the concensus is?

Personally I'd much rather see the opposite, i.e. that the default be
such that if a _system_ header file requires some other _system_ header
file, and especially if this is documented as a dependency in the system
API, and even more so if some API standard documents this dependency as
required, then the compiler should halt with an error stating that the
required #include is missing.  It should most definitly not silently
include the documented dependent header!

While some programmers would like the system to take care of its own
dependencies, I would much rather that the programmer be forced to
understand these dependencies and the resulting impact on the final
code.

--=20
						Greg A. Woods

H:+1 416 218-0098 W:+1 416 489-5852 x122 VE3TCP RoboHack <woods@robohack.ca>
Planix, Inc. <woods@planix.com>       Secrets of the Weird <woods@weird.com>

--pgp-sign-Multipart_Mon_Dec_11_12:17:27_2006-1
Content-Type: application/pgp-signature
Content-Transfer-Encoding: 7bit

-----BEGIN PGP SIGNATURE-----
Version: PGPfreeware 5.0i for non-commercial use
MessageID: pCiNNz9bhmYlzCCNYm3kEI0W9Hh7wkOh

iQA/AwUBRX2SyGJ7XxTCWceFEQL9egCfepsCP1FUkVNTkWByq4kpJeGfMUEAoM4g
5hT91FgyrASPPIjZWjc8+59d
=YvOS
-----END PGP SIGNATURE-----

--pgp-sign-Multipart_Mon_Dec_11_12:17:27_2006-1--