But, as far as I know, IPv8 doesn't give a solution for the QoS issue. IPv6 has
the potential. I think the killer applications of IPv6, or whichever the next
generation IP protocol, are very likely what demand high QoS and natively
peer-to-peer, such as video interactive game. If an application doesn't demand
high QoS it can somehow be implemented with the inexpensive NAT middleboxes
and/or message relays with ripe experience.
Jason.
----- Original Message -----
From: "Jim Fleming" <JimFleming(_at_)ameritech(_dot_)net>
To: "Fred Baker" <fred(_at_)cisco(_dot_)com>; "Jason Gao"
<jag(_at_)kinet(_dot_)com(_dot_)cn>
Cc: <steinle(_at_)smartvia(_dot_)de>; "Richard J. Sexton"
<richard(_at_)vrx(_dot_)net>; <karl(_at_)cavebear(_dot_)com>; "Joe Baptista"
<baptista(_at_)dot-god(_dot_)com>; <jefsey(_at_)jefsey(_dot_)com>;
<andy(_at_)ccc(_dot_)de>; "@quasar Internet Solutions, Inc."
<shore(_at_)quasar(_dot_)net>; "Vittorio Bertola"
<vb(_at_)vitaminic(_dot_)net>; <sotiris(_at_)hermesnetwork(_dot_)com>;
"Richard Henderson" <richardhenderson(_at_)ntlworld(_dot_)com>;
<k(_at_)widgital(_dot_)com>; <jwkckid1(_at_)ix(_dot_)netcom(_dot_)com>;
"Judith Oppenheimer" <joppenheimer(_at_)icbtollfree(_dot_)com>; "Joanna Lane"
<jo-uk(_at_)rcn(_dot_)com>; <eric(_at_)hi-tek(_dot_)com>;
<DannyYounger(_at_)cs(_dot_)com>; "Bruce Young"
<Bruce(_at_)barelyadequate(_dot_)info>
Sent: Tuesday, September 10, 2002 8:28 PM
Subject: Re: Fuzzy-layering and its suggestion - Towards better QoS solution
in the IPv6 network
----- Original Message -----
From: "Jason Gao" <jag(_at_)kinet(_dot_)com(_dot_)cn>
To: "Fred Baker" <fred(_at_)cisco(_dot_)com>
You don't get to pull a bit out for your private use.
It is still not a requirement that every node in the Internet must be
DS-compliant.
Actually, it is just the opposite. The edge of the global, public, 32-bit,
IPv4++ Internet
is defined as the collection of private companies and people, connected
together, that
pass 160-bit headers with the 8-bit TOS field unchanged. Those 8-bits can
be used
to expand the addressing at the edges of that network, 4-bits are needed in
each direction.
That expands the address space by adding 15 more Internets, as large as the
legacy net.
You can easily obtain code to do this, or help write more code to do it.
http://www.netfilter.org/
http://netfilter.samba.org
http://www.google.com/search?q=TOS+routing
http://www.google.com/search?q=packet+mangling
As for IPv6, one can not assume that the 128-bit DNS implies IPv6. As an
example,
if you use a 2002 model year address, such as 2002:[IPv4]:* in the 128-bit
DNS AAAA
records, your computer will likely generate IPv4 packet headers. A6 DNS
records are
for IPv6. AAAA records work for IPv4++. Private companies and private
individuals
are working together to define what goes in the AAAA records. You are FREE
to
participate. Some private companies and closed societies may try to tell
you that you are
not free to participate, but that is not the case. Even with tens of
millions of dollars from
domain name sales and address space leasing, they will not be able to pull
together enough
money to buy the freedom of all people on planet Earth. They may try...but,
most people's
minds are not for sale.
Jim Fleming
2002:[IPv4]:000X:03DB:...IPv8 is closer than you think...IPv16 is even
closer...
http://ipv8.dyndns.tv
http://ipv8.yi.org
http://ipv8.dyns.cx
http://ipv8.no-ip.com
http://ipv8.no-ip.org
http://ipv8.no-ip.biz
http://ipv8.no-ip.info
http://ipv8.myip.us
http://ipv8.dyn.ee
http://ipv8.community.net.au
http://www.iana.org/assignments/ipv4-address-space
http://www.ntia.doc.gov/ntiahome/domainname/130dftmail/unir.txt