--IJpNTDwzlM2Ie8A6
Content-Type: text/plain; charset=iso-8859-1
Content-Disposition: inline
Content-Transfer-Encoding: quoted-printable

Salut,

On Wed, Jul 20, 2005 at 09:36:10PM +0100, David Laight wrote:
> The chance of missing an error with CRC32 is (about) 1 in 2^32.
> A single bit error is guaranteed to give an error (indeed I think any
> single burst error block of upto 32bits will give an error).
>=20
> For a 'safety' algorith a modulo 2^32-1 sum is probably enough, and
> much, much faster to calculate.

That's what I keep saying here..

Use CRC32, it's ok for this purpose.

				Tonnerre

--IJpNTDwzlM2Ie8A6
Content-Type: application/pgp-signature
Content-Disposition: inline

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.1 (NetBSD)

iD8DBQFC3uOuXUVlAbfmNMIRAtL+AJ410AzyrJ4sbyL17nRNV0s73uLgfQCgo/HO
ardgz26VlDLSLYNbgWj1m6s=
=szAb
-----END PGP SIGNATURE-----

--IJpNTDwzlM2Ie8A6--