-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1
- -----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1
Peter,
Neither of us has the required data to back this up but my impression
of the IETF attendance in the last 5+ years has been largely hardware/
software vendors, followed at a distance by service providers, researchers
and curiosity seekers. If you count only those who participate, the number
of vendors and perhaps researchers would far outweigh service providers.
As such, v6 isn't really seen by the service provider as 'my fight'. (To
take an even more cynical view, v6 is often deployed as a reaction to
IPv4 as a 'North American' phenomenon (rightly or wrongly and forgive
me in advance)).
I agree that NA ISP's don't 'need' to deploy v6. I agree that they have no
short term economic reason to do so. In fact, as many of them are fighting
to stay alive they have very good reasons not to deploy v6. To reference
an earlier analogy, just as there were very good reasons not to toss a
bunch of perfectly good class 5 line cards in the hopper and replace them
with ISDN cards despite all good intentions of standards bodies and
vendors... The difference being that ISDN had nothing to do with an
inadequate numbering plan for the future.
While it is certainly possible to deploy v6 without ISP's, 'the universal
deployment of IPv6' will not happen without them. It won't even be a
smooth deployment of IPv6, it will be ugly and it will take a long time.
With regard to dual stacks, I think that this is the best that we can hope
for (and the only path that will lead to universal deployment of IPv6).
If I cared about this subject (I do) I would be lobbying in this direction
with my friends in ISP's and content providers.
It isn't that hard, you and I used to work for an ISP that deployed a dual stack, albeit OSI and IP, internet. (yes, and I liked TUBA too).
jy
On Nov 18, 2004, at 2:12 PM, Peter Ford wrote:
IPv6 is already taking off, and IPv6 will even more once it is clear to
all that IPv6 deployment is not gated by ISPs deploying IPv6. A core
problem in the IETF is that its major source of people is down in the
"ISP space". ISPs don't need IPv6. It is the end system folks that
need IPv6, and they KNOW they need it, and are acting accordingly.
The really good news is that IPv6 CAN be deployed without the networks changing. The ISPs can "catch up" when 1) they need to and/or 2) when they want to for the purpose of capturing economic gains.
In terms of the concerns of dual stacking: if you were a cell phone guy,
would you put IPv4 on your phones? I think not.
Regards, peterf
-----Original Message----- From: ietf-bounces@xxxxxxxx [mailto:ietf-bounces@xxxxxxxx] On Behalf Of Noel Chiappa Sent: Tuesday, November 16, 2004 9:17 AM To: ietf@xxxxxxxx Cc: jnc@xxxxxxxxxxxxxxxxxxx Subject: Re: How the IPnG effort was started
From: Jon Allen Boone <ipmonger@xxxxxxxxxxxxxx
In my experience, if a technology hasn't been readily adopted within a decade of it's creation, it's not going to be. It appears that time is rapidly approaching for IPv6.
Ah, you need to adjust your clock, or calendar, or whatever. SIP (what
we
now call IPv6) was created in 1992 (it was presented to the ANTF meeting
in August '92), and was adopted as IPng at the 30th IETF in Toronto, in
July 1994. That's already more than 10 years.
Just to give everyone a sense of what that really means, here are some things to jog our memories. In 1994:
- The WWW had about 2,700 sites, total. - The current Microsoft operating system was Windows 3.1
Think about that for a minute.
It's pretty clear by now that IPv6 is just not going to reach its stated
goal - which is to ubiquitously replace IPv4. Even many IPv6 proponents
are now speaking of an essentially indefinite period of co-existence.
Which essentially voids the original basic argument *for* IPv6...
And don't give me any of that "oh, we really needed to have the X system
available, now we've got that it'll really take off next year". We've
been hearing this exact excuse for years - I have a whole file full of
them.
Yes, there is going to be some deployment of IPv6. (With the amount of money that's been spent on it, it'd be totally astonishing if there *weren't*. If I were a barn manufacturer, and had the kind of budget that's been spent on IPv6, half the airline passengers today would be flying around on jet-propelled barn doors.) It will see some use in discrete areas of the network, particular networks that utilize IPv6.
It may even find a certain amount of utility as an end-end naming layer (which is incredibly ironic, but that deserves a rant in itself); but again, that not the original goal - which was to be the ubiquitous packet layer.
Look, I really do understand Brian's point - that the current situation is not good.
But acting like IPv6 is going to magically save us - when we have year after year after year after year of actual experience that is telling us "no, it isn't" - is not the way to fundamentally improve the situation.
The IETF needs to seriously face the reality of the network that's really out there, not the network some of us wish were there.
To put it another way (and mangle a well-known phrase in the process),
if
life gives you lemons, you can either sit around with a sour look on
your
face, or make lemonade. NAT's make me look sour too, but I'd rather make
lemonade.
Noel
_______________________________________________ Ietf@xxxxxxxx https://www1.ietf.org/mailman/listinfo/ietf
_______________________________________________ Ietf@xxxxxxxx https://www1.ietf.org/mailman/listinfo/ietf
- -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.2.3 (Darwin)
iD8DBQFBnS8AtzMIqcCTiFQRAh8nAKCFm8s9fVg0yIKcMybHXpgZqMJLHQCdHmMY /dEBCaI/JVQgxyiP695QZJE= =pvlN - -----END PGP SIGNATURE-----
-----BEGIN PGP SIGNATURE----- Version: GnuPG v1.2.3 (Darwin)
iD8DBQFBnTLaLWqnmaznXfoRAml8AKCSFxGR1UEmrnkofob2PPSYnKb+QACZATO8 f00etEkeqIAhY7en0s/qJx0= =HXie -----END PGP SIGNATURE-----
_______________________________________________ Ietf@xxxxxxxx https://www1.ietf.org/mailman/listinfo/ietf