Hi Pekka/Brian, I was thinking of enable/disable flags for separate prefixes which override the global settings.
Let's say you want privacy addresses for everything but ULA you would have the following settings Global -> Enabled fc00::/7 -> Disabled Let's say Brian just wants to enable them for 2001::/16 and 2002::/16 Global -> Disabled 2001::/16 -> Enabled 2002::/16 -> Enabled I think that should address both your concern and Brian's concern. Thanks Suresh On Thu, 21 Oct 2004, Pekka Savola wrote: >On Thu, 21 Oct 2004, Suresh Krishnan wrote: >> Hi Brian, >> That sounds fair to me. I will come up with text with SHOULD language >> for per-prefix enabling of privacy addresses. I just have to figure out >> how it will interact/override with the global enable/disable option. >> >> Pekka, >> If I make this change, would you still like me to add specific defaults >> for ULAs? > >I can live with 2001::/16 + 2002::/16, but I think that's a bad choice >for multiple reasons. What if we invent 6to4v2 which uses 2005::/16 >and we'd like to automatically apply these semantics to it? What if >we run out of 2001::/16 for native allocations? -- actually we've >already 1/3 used it up. > >Thus being generic and excluding just those that we _know_ aren't >really, really global might seem as a better approach -- one that we >might not need to tweak e.g., 2-3 years down the road.. > > -------------------------------------------------------------------- IETF IPv6 working group mailing list [EMAIL PROTECTED] Administrative Requests: https://www1.ietf.org/mailman/listinfo/ipv6 --------------------------------------------------------------------
