Date: Thu, 30 May 2002 16:43:50 -0700
From: Bob Hinden <[EMAIL PROTECTED]>
Message-ID: <[EMAIL PROTECTED]>
| It doesn't seem right to make them non-compliant (i.e., make RO a MUST).
Bob, that's completely bogus as an argument. I don't know enough about
the issues to comment on the substance, but if RO is something that the
WG feels is important for all nodes to implement, then of course MUST is
the right thing.
Implementations that already exist can't possibly pretend to conform to
a specification that doesn't yet exist, whatever is in that specification.
When the new spec does appear, if older implementations decide that they
want to be compliant with it, then they may need to issue software upgrades,
or something, to get them there.
There's nothing new in this, one day soon I'd hope that we want to make
IPv6 a MUST implement for the Internet (one of the base required standards
like IPb4 is today) - but if the argument is that we can't do that, or lots
of stuff that shipped 20 years ago would become non-conformant, then we'd
never get anywhere.
When we're inventing something new, it is correct and reasonable to
consider what effect that will have on the installed base which knows
nothing about it - and I assume (and hope) that the RO stuff has been
planned in a way such that the world won't break if some node or other
happens not to have implemented it (even assuming it is a MUST).
But we can not, now or ever, allow considerations of existing implementations
conformance state affect the decisions about what new implementations
should be required to implement.
kre
--------------------------------------------------------------------
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]
--------------------------------------------------------------------
