Earlier, Eliot Lear wrote:
% Put simply, you are over-specifying the RID
% and derive no benefit from doing so.
There is benefit to implementers in having a specific
example that is known to meet the requirements of
this specification. This benefit also accrues to users,
as implementers are more likely to correctly implement
when giving a specific example that they may reuse.
% Also, the following text in section 3 Page 7 is contorted:
%
%> This means that this document does not formally obsolete or
%> deprecate any of the existing algorithms to generate Interface IDs
%> (e.g. such as that specified in [RFC2464]). However, those IPv6
%> implementations that employ this specification must generate all
%> of their "stable" addresses as specified in this document.
%
% My suggestion is to simplify remove it as it is self-evident.
It would reduce consensus within the WG and the IETF as a whole
to remove that text -- as it clarifies that other mechanisms for
generating such IDs are not affected by this specification.
Text to this effect was specifically requested by various WG
participants, precisely because it is not "self-evident" and
instead was confusing.
% Finally, this algorithm requires that the resultant host
% portion be 64 bits. Is that necessary?
Yes, it is required, for technical reasons relating to multiple
IPv6-related standards-track RFCs.
I oppose Eliot's proposed edits on grounds that they would
reduce the clarity of the specification and also would
reduce IETF and WG consensus about this specification.
Yours,
Ran