Also go look at the source address selection stuff where we now have kernel
and user
space hacking dns abstract types back and forth. Another case of excessive
engineering insanity and architecture without consideration for performance.
Its good work as a spec but I have my doubts about what it will do to the
performance of end system **production** running code. Or now we use IPv4
anycast in ngtrans which has never been deployed or even well tested in an
evironment like the core interoperability events IPv6 has worked on.
Or the most ridiculous thing of late is this privacy stuff of temporary and
public
addresses cause some adhoc group is paranoid about my MAC address. Then
people take all this and make some kind of grand computer science cosmos
wonder out of it and slow down everything else around IPv6 or put an
interrupt in it for real implementors. Oh yeah go look at the scoping.
Sooner or later the market will just barf on all this and force the vendors
to just deliver basic IPv6 as IPv4 so we can solve the real problem lack of
address space and NATism. But I truly hope your mail at least enlightens my
colleagues to fix these A6 records. It will not work.
Not feeling politically correct and want IPv6 to be deployed with warts if
necessary,
/jim
-
--------------------------------------------------------------------
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]
--------------------------------------------------------------------
