Robert Elz wrote:
> 
>     Date:        Sat, 07 Jun 2003 11:02:38 -0400
>     From:        Margaret Wasserman <[EMAIL PROTECTED]>
>     Message-ID:  <[EMAIL PROTECTED]>
> 
>   | The problem with using FECO::/10 for these addresses is that there
>   | are implementations that include special semantics to handle
>   | the ambiguity of FECO::/10 addresses, such as requiring a Zone ID
>   | to disambiguate FECO::/10 addresses, setting up separate routing
>   | tables for separate FECO::/10 zones, etc.
> 
> All that is OK, there's no implementation that I know of that requires
> that SL addresses exist in multiple scopes.   An implementation as you
> describe it will work just fine, making use only of one scope for all
> of these addresses, until perhaps some later version of the system
> stops using the scope identifiers altogether.
> 
> Nothing is going to break.
> 
>   | In Bob's proposal, addresses are globally unique
> 
> Once again, no, they're not.   The difference between them and the SL
> that we have had is that we are going to ignore the non-uniqueness,
> rather than explicitly handle it.

To be more precise, we are going to define them in a way that makes
them extremely likely to be unique, and then treat them as if they
are mathematically unique. To me, that's good enough.

I don't mind too much which prefix we use. A clean separation from
FEC0 might be safer.

   Brian


--------------------------------------------------------------------
IETF IPng Working Group Mailing List
IPng Home Page:                      http://playground.sun.com/ipng
FTP archive:                      ftp://playground.sun.com/pub/ipng
Direct all administrative requests to [EMAIL PROTECTED]
--------------------------------------------------------------------

Reply via email to