Interesting: I like you piercing spirit. But, I am afraid you are too much
legacy intoxicated :-) what I think surprising. I suppose we agree but you
have odd ways of seeing it.
At 18:58 02/08/2005, Hallam-Baker, Phillip wrote:
> From: JFC (Jefsey) Morfin [mailto:jefsey@xxxxxxxxxx]
> Except if you can grab a BCP. I am not sure you are actually
> right. You certainly know a few cases.
The lack of an IETF endorsed spec from MARID did not stop Microsoft from
holding an industry gala two weeks ago in NYC. Nobody commented or
appeared to care that the spec was not ratified.
Think of it as a recess appointment.
> The problem with IETF is there is no architectural common vision.
No, that is its strength. The Web was not part of the IETF common
vision. SSL was diametrically opposed to the IETF security vision.
I am afraid you speak of details. These are applications. There is no
common vision of the reality of the digital ecosystem nature. IETF has fun
over layer 8 and 9. Layer 8 to 12 have a precise meaning, as has layer 0.
Sharing this meaning would help a common analysis and avoid confusing
Multilingual Internet with a bunch of typewritters, using typographer ISO
tables to document it.
> IETF and IANA have a defacto monopoly on the architecture.
No they don't. W3C and OASIS are both more influential as standards
bodies at this point, particularly once we get above the session layer.
Again you speak of details. Of applications. I am speaking of the network
architecture. The evolution of the architecture is very very slow.
What I say is that in the real world, users are not interested in all that.
This is for application providers and applications are decided by the
users. Who known the W3C and SGML 10 years ago. Will we still know them 10
years from now?
The URI identifier architecture introduced in PICS and since adopted in
XML eliminates the need for fixed registries like the IANA. That was the
whole point, to eliminate the control point. I did not want a central
registry of PICS censorship schemes. Of course other people did, mostly
the people who used euphemisms like 'content selection' rather than
censorship.
Agree. But IMHO this is a way to introduce at application level the very
basic "root name" principle introduced by Robert Tréhin and Joe Rinde in
1977 which founded the International Network, in part the OSI and defaulted
to "root" with the Internet defaulting its architectural parameter to
"mono" from "multiple" in Tymnet and from "separated" in OSI. This is what
we have to correct now.
I would phrase it another way. The IANA is one of the many roots in the
International Network forest. But that trunc of that root hidden the
forest. The Multilingual Internet is probably the best application to force
and fund the necessary effort to look at the forests. But some suspect that
the resulting user-centric architecture (each one having its many roots)
has a different economical model.
And status quo is the best target for dominant one. At ICANN they use to
call the "stakeholders" ....
> For example the whole IPv6 issue is that they did not understand that
> their current deployement (2001) is disposable.
The failure to get the deployment stakeholders round the table to ask
the question 'what will it take to make this happen' is in my view the
root cause.
I do not. Because what will make it to happen is the disappearance of
"stakeholders". Let understand, the current network is made of people who
want to organise, to sell, etc. it. The future stable IPv6 network by
nature (it would not be an evolution otherwise) will be made of people
wanting just want to use it. And the first thing they want to get rid of is
the artificial limitations of the stakeholders.
Take care. I am quite interested in your security factor in relations. Did
you work on that (I did not recall exactly how you termed it: we called
"delta sec", and people grab the idea quick). I suppose that network
security could be discussed in a similar way to network value?
jfc
_______________________________________________
Ietf@xxxxxxxx
https://www1.ietf.org/mailman/listinfo/ietf