> The current basic socket enhancements draft > (draft-ietf-ipngwg-2553bis-05.txt) specifies that this 32-bit integer > "identifies a set of interfaces". More specifically, a "interface index" > for a link-local scope sin6_addr,
I think it should say "link-local scope id" for link local address, to be logical (which is not necessarily same as interface index). The idea of having a scope id which is of different type than than the address was rejected. (Which makes me wonder why does the scoped architecture draft talk about having the 4 bits in ID assigned to scope type: I think this is internal implementation issue, and not required, if system works without such trick). However, I still don't have strong opininion about this, e.g. whether scope identifiers type can differ from type of address. Currently, I'm writing a version where identifiers scope type is *always* determined from the address -- you cannot have isolated scope id. Seems to give fairly clean result.. so far... -------------------------------------------------------------------- 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] --------------------------------------------------------------------
