Python COM and distribution issues

Robin Becker robin at jessikat.fsnet.co.uk
Tue May 15 03:37:52 EDT 2001


In article <fqe1gt4vlh8437kdgmt1gjd2deni1b9ube at 4ax.com>, Tim Roberts
<timr at probo.com> writes
>hungjunglu at yahoo.com wrote:
....
>The original GUID scheme included the 6-byte MAC ID of the generating
>machine, a 4-byte date and time, and an 6-byte random number.  Assuming
>that the generating machines always have an Ethernet card, it is impossible
>for such a scheme to produce a duplicate.
many best laid plans etc. This assumes the MAC's to be distinct, but
there are many examples where supposedly unique identifier schemes have
failed. In Britain a national newspaper refused to pay out on a
competition when every copy had the winning sequence. Ethernet chips
probably aren't worth faking, but there's no guarantee that the
manufacturing process always steps the number.
>- Tim Roberts, timr at probo.com
>  Providenza & Boekelheide, Inc.

-- 
Robin Becker



More information about the Python-list mailing list