Re: [67ATTENDEES] Re: IETF67 network

--On Wednesday, 15 November, 2006 11:12 -0800 Bill Fenner
<fenner@research.att.com> wrote:

> The noc box just came back up (Thanks, Joel!) so I ran a quick
> database query for MAC addresses - unfortunately the vendor
> names are somewhat arbitrary (taken directly from IEEE's
> "oui.txt") so you get to add up the duplicate entries.  Of
> course, there's no way to tell whether an "Intel Corporate"
> MAC address represents Windows, FreeBSD, Linux, BeOS or Plan
> 9...

Bill,

I suspect it tells us even less than that.  I've been told
repeatedly that a huge percentage of the basic Wireless LAN
chipsets (802.11 A, B/G, pre-N, mixtures thereof and even WiMax)
come out of a very small number of companies, almost all of them
headquarted in, or operating out of, the same country, and
selling them on an OEM basis.  Whether the MAC address in a
given NIC reflects the assignment pool of the
manufacturer-packager-vendor or that of the original
manufacturer is presumably a matter of commercial negotiation;
the decision may tell us more about marketing and branding than
about chipset origins, much less the type of computer or
operating system with which the chipset is being used.

    john

_______________________________________________
67ATTENDEES mailing list
67ATTENDEES@ietf.org
https://www1.ietf.org/mailman/listinfo/67attendees