> -----Original Message----- > From: David Conrad [mailto:[EMAIL PROTECTED] > I thought "Site Locals" were deprecated because people > couldn't agree on what a "site" was. > > > Are these ULA-Cs simply taking their place? > > That's my impression, but then again, I haven't see the > revised ULA-C draft.
Site local were deprecated because they were re-creating overlapping address space a la RFC1918. This was considered a bad property for multi-party applications that couldn't tell if they could forward them as reference or not. Some people believed that some kind of 'private addresses' were still needed and thus needed something to replace site local. They invented ULA. ULA fixed the (overlapping) addressing issue but did not fix the routability issue as has now been demonstrated in multiple post. We are essentially back to where we were a couple years ago, this time not at the application level but at the router level, it is hard to know if those (local/private) addresses should be passed to the next hop or not. I'm of the opinion that the added operational complexity to define sites and partial inter-site routing overcome the benefit of ULA especially when PI space is available from RIRs and accomplish the same thing at no extra cost. However, I do respect people who have a different opinion and once again I repeat my call/plea for an open meeting of the IPv6 wg at next IETF to articulate those issue and decide if the wg should or not continue on the path of defining those ULA-c despite the (very) negative feedback from the operational community. - Alain. -------------------------------------------------------------------- IETF IPv6 working group mailing list ipv6@ietf.org Administrative Requests: https://www1.ietf.org/mailman/listinfo/ipv6 --------------------------------------------------------------------