On 16-May-2007, at 22:11, Iljitsch van Beijnum wrote:
On 17-mei-2007, at 3:29, Joe Abley wrote:
There is an argument that the right approach to facilitate source
routing experiments is to deprecate RH0, and define a new type of
routing header which is, from the outset, disabled by default.
Please present this argument; it's not self-evident.
I agree that was not especially clear.
Suppose there are reasons to think that deprecation rather than
disable-by-default is the preferred approach.
A requirement to continue to provide source-routing functionality
akin to RH0 could be satisfied by either abandoning the deprecation
approach in favour of disable-by-default, or by defining new types of
Routing Header which will facilitate the desired functionality in
such a way that will not jeopardise the stability of the network as a
whole.
If there is actually no requirement to provide source-routing
functionality, then this is a non-issue.
The remaining assumption is that deprecation rather than disable-by-
default is the right thing to do. That is presumably what you are
questioning. My point was that a requirement to support RH0-like
functionality is not fundamentally incompatible with the deprecation
approach.
Joe
--------------------------------------------------------------------
IETF IPv6 working group mailing list
[email protected]
Administrative Requests: https://www1.ietf.org/mailman/listinfo/ipv6
--------------------------------------------------------------------
