From: Joaquin Cuenca Abela (e98cuenc_at_free.fr)
Date: Wed Mar 03 2004 - 14:30:12 EST
> --- Marc Maurer <j.m.maurer_at_student.utwente.nl> wrote:
> > it is normally used as input for the random number
> generator. When no
> > mac address is available, just fallback to some default number...
> Right... I guess the questions are better stated as
> 1) What is the maths/logic behind document comparison and
> revision tracking?
> 2) Why is it implemented using (supposedly) UUIDs?
> 3) Why is (pseudo) randomness seemingly a requirement of the process?
> Why we would use a MAC address to build a UUID makes complete
> sense to me. Why we need UUIDs does not. We probably don't
> want to use N pseudo-random int32s to build a "MAC address".
> Concretely, getting the MAC address on various *NIX boxes
> requires substantially different and sometimes complex code.
I think that MS is not using the MAC address anymore (since Windows 2000?)
to create UUID, as people raised concerns about privacity.
If the UUID contains components from the MAC address, you can track back the
person/computer that created the UUID.
I think that now they use a random number generated from their crypto lib,
This archive was generated by hypermail 2.1.4 : Wed Mar 03 2004 - 14:31:00 EST