> b) if there's a strong need for appropriate > ICMP responses to make IPv6 function well, then an RFC stating as much > could be published. like, for instance, a need for path mtu discovery to work reliably? (since v6 routers don't fragment packets) > >The vast majority of applications do not pick their source address, nor is > >there a compelling reason for them to do so. A large number of > >applications don't even handle multiple destination addresses properly, so > >expecting this additional intelligence for the source address is > >irrational. > > At which point they might as well just select their global address unless > the destination address for a service is site local. This decision could > (should?) be in the hands of the IP stack, unless the application > specifically asks for such control. I disagree that the IP stack can supply reasonable defaults in the face of multiple scopes. The criteria for choosing a source address varies widely from one application to another. Some applications need stable addresses, others need addresses usable from all of their potential peers, others need to choose the source address that results in the best performance (where there is more than one meaning of 'best performance'). The default address selection rules are at best a guess. We should let the network do routing, so that hosts and apps aren't expected to make routing decisions.