> An implementation note which identifies the need for any multi-party > apps to have a scope determination mechanism before using SL is > appropriate.
no, I'm sorry. It's not. it's insane. look, it's a separation of function argument. the network's job is to do "best effort delivery" SO THE APPLICATION DOESN'T HAVE TO. if you expect applications to do routing and address selection and to know about network topology not only on the local network but from the point of view of distant peers just in order to get the packets delivered, why don't we just dispense with routing protocols entirely? > Claiming SL is something that applications can't > effectively use is a bit over the top. For a simple 2 party app (like > sending a file to my printer), SL is a very appropriate addressing > mechanism. If I don't want the world to connect to my printers, it is > much easier to filter FEC0::/16 at the border than it is to list every > printer in an access control list. that's the same kind of bogus claim that's being made by the zeroconf folks for v4 LL addressing. > A multi-party app developer should be happy there is a specific range of > addresses set aside for SL. Without that it becomes a guessing game as > to which addresses might work or not. it's a guessing game anyway if sites think that SL addresses can be used as a substitute for routable addresses. Keith -------------------------------------------------------------------- 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] --------------------------------------------------------------------
