Re: Transport level multihoming

[Ran Atkinson <rja@extremenetworks.com>]

At 14:00 03/04/01, Greg Maxwell wrote:

>The reasoning stated more clearly: Application compatibility for
>multihoming in IPv6 is not important because IPv6 itself breaks
>application compatability to a similar extent as multihoming changes
>would. 

        It is not clear that anyone could know whether
"IPv6 itself breaks application compatibility to a similar 
extent as multihoming changes would", particularly since the 
"multihoming changes" are currently not specified anyplace.

>IPv6 supporting application can be deployed with multihoming
>support thus making the marginal cost of multihoming support zero.

        See above.  This assertion is impossible to support
given that we have no spec for multihoming in IPv6 other
than the approach currently deployed with IPv4 (which is known
to have real problems).

>Additionaly, multihoming changes would likely signifantly lower backwards
>compatibility cost then IPv6 itself 

        Again, since "multihoming changes" are not defined
anyplace, such a comparison is impossible to make in a
scientific way today.

Ran