* Leif Johansson <[EMAIL PROTECTED]> 20030325.0913: > > Long experience shows that private addresses in any > form are a > bad idea on the Internet.
But what I understand of Saywell's requirement is that this isn't the Internet (big 'I'). It is _an_ internet, whose topology changes shape, albeit not too frequently. Sometimes, some parts of this internet have links to the Internet, and so will see globally routable address prefixes that may be propagated to all the nodes in the 'site', meaning that every participant node in Saywell's network will (for that time at least) have 'n' globally routable addresses from the 'n' nodes that have links up to the Big-I. However, that level of connectivity is ephemeral at best, for the target audience, as I perceive it, includes people having nodes on dial-up lines (therefore potentially getting a different global prefix on each connection), DSL lines, and vary rarely any form of fixed provider (its a community networking activity, right?). Therefore, it is *not* clear that the nodes will have any form of known/fixed/reliable address that they can communicate with each other internally and, given that this is a multi-link network, there is a desire for something that looks a lot like site local addressing. (Or have I misinterpreted the application?) Mark/ -- "There are no stupid questions, just stupid people" - Mr. Garrison, South Park, CO -------------------------------------------------------------------- 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] --------------------------------------------------------------------