Thread Links Date Links
Thread Prev Thread Next Thread Index Date Prev Date Next Date Index



Thanks for forwarding this interpretation question to me.
As a member of the IEEE RAC, I would like to clarify the
meaning of:
  The EUI-64 values shall be sold within electronically readable parts;
  no more than one EUI-64 value shall be contained within each component
that is manufactured.

A vendor (such as Dallas Semiconductor) could choose to sell EUI-64
identifiers as a commercial product, perhaps at Fry's
(my friend tells me they have, but I haven't confirmed),
then they each device (typically <$0.50) can have only one EUI-64.
The intent is to avoid subcontracting of further subassignments.

There is no intent to restrict the number of protocols that can be
supported by any device. A perfectly acceptable procedure is
to assign one EUI-64 to each protocol, while allowing any device to
select freely which of the protocols its supports, as your group

As the author of the EUI-64, I feel confident in this interpretation.
I have copied this to the IEEE/RAC, to remove any possible doubts.


David V. James
3180 South Ct
Palo Alto, CA 94306
Home: +1.650.494.0926
Cell: +1.650.954.6906
Fax:  +1.360.242.5508
-----Original Message-----
From: Glenn Parsons []
Sent: Tuesday, September 16, 2003 3:37 AM
To: 'David V James'
Cc: ''
Subject: EFM OAM

You are up early sending a flurry of email ....
Anyway, on the EUI comments we had on EFM OAM.  An issue raised by 802.3
management -- that can be justified by the 'distribution restrictions' in
the tutorial here:
-- is that an one device cannot support more than one EUI.  That is, you
could not support interpreting 'Cisco' specific OAM frames and 'Nortel'
specific OAM frames.
Do you know if this URL is the definitive definition of EUI?
This is critical.  If this restriction is true, I will want to remove EUI
from RPR.