Margaret Wasserman wrote: > It is my personal opinion that we should leave site-local > addresses as-is and strictly limit their use to completely > disconnected, isolated networks.
1) It is not possible for the IETF to restrict their use to completely isolated networks. 2) It is not a requirement that a connected network not use them. The only reasonable argument here is that the IETF may choose to leave the details of such use undocumented, and not spend time addressing the DNS issues. > > I also think that we should work on a way to offer provider- > independent IPv6 addressing, but I think that these should be > global addresses, not site-locals. While it is possible that > ISPs will not advertise these addresses in global routing > tables, I do not think that they should be limited to a > "site", or treated as scoped addresses (with zone IDs, etc. > as in the scoped addressing architecture). I think that they > should be global addresses that may be filtered at > administrative boundaries (and _will_ be filtered at some > boundaries, if we can't find an aggregable way to allocate them). I agree with the need for PI, and for those who would like to provide comments, the current state of the updates to my approach can be found at: http://www.tndh.net/~tony/ietf/ipv6piaddressformat-04.txt & http://www.tndh.net/~tony/ietf/ipv6piaddressusage-04.txt I found a way to change the grid sizes and origin so now all of Europe is in one logical grouping. > > I do not, personally, think that we gain much by trying to > improve current site-locals to be more unique. We might provide an informational document that provides guidance to sites to randomize the upper 38 bits to avoid potential problems with mergers later. We don't need to specify a standard mechansim, but might suggest a couple of reasonable stratigies. Tony > > Margaret > > > At 02:47 PM 12/5/2002 -0800, Michel Py wrote: > > > Margaret Wasserman wrote: > > > I am not sure that we have general agreement that there should be > > > two courses. I, for one, do not think that we should create two > > > different types of addresses... > > > >Now you got me lost. If you say site-locals should not be used for > >external connectivity, how are you going to do GUPIs without > creating > >another type of address? > > > >Michel. > > > -------------------------------------------------------------------- > 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] > -------------------------------------------------------------------- > -------------------------------------------------------------------- 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] --------------------------------------------------------------------
