[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[atlarge-discuss] Re: [ga] Vint Cerf opposed to users being redirected to search engines



On 04:33 06/09/03, L. Gallegos said:
ICANN chair Vint Cerf says he is adamantly opposed
to the practice, however, arguing that modifies DNS servers
to things they're not designed to do.
http://online.wsj.com/article/0,,SB1062710102434100,00.html
Dear Leah,
I am not to pay to see what Vint thinks on the matter. This can be expressed fore free at the GA or at the IETF. We are not at the ITU. Now the DNS purity is the business ICANN.

There is a multilayer name space. It has been initiated by Bob Tréhin and Joe Rinde as a two layers system for International relations in 1977 and we unlimitedly extended it in 1978 (INTL).

In 1983/84 this INTL name space semantic was completed in two ways: by the support of the X.121 ITU scheme and by the Internet DNS semantic, defined not conflict with these two leading semantics (INTL: http://intlnet.org/intl.htm - X.121 http://intlnet.org/d_dcc.htm ) and to match other networks flat naming semantic while internetting them. This is the 1984 RFC 920 which is the basis of the ICANN legitimacy (cf. ICP-3). The 1983 born Internet is defined by its IP addresses and the DNS. It is normal that ICANN protects the integrity of the DNS legacy within its boundaries. And did it up to now, even using the RFC 920 multiorganization permission in 2000. The only breaches are to some extent ".info" which is a gTLD and ".biz" which a deliberate conflict with an existing foreign name space.

The initial general naming scheme (before the INTL scheme) in every technology was either flat names (most of the networks), numeric addresses (Telenet I understand?) , or a mix of them (Tymnet). In the 90s the keyword approach developed (Real Names, AOL) introducing multiple flat name, however not in full disorder. This is today supported by the IETF/W3C "go:" scheme by Nicolas Popp from Verisign (from his experience as Real Names CTO) . However to my knowledge no protocol/solution supports it yet.

If the DNS semantic is better formed than the INTL semantic and much more versatile than X.121, it misses the unlimited expansion of INTL, the structured easiness of X.121, the simplicity of the "go:" scheme. Actually, it misses the intuitiveness, which implies a certain possible disorder, (typing, linguistic, etc) flexibility and (geographical, logic, thematic) abbreviations in the name structuring. There is a word to define all that which is "vernacular". And a system to work with : "brainware" (what people want, understand, think and react upon).

The way people welcomed and used that different solutions (brainware is brainware, don't expect it to be proactive in hardware and software areas) show that people want vernacular names. Want in brainware wording means that people collectively tend to organize the currently proposed solutions in given direction.

Also, most pros want to get rid of the old/odd domain concept which very often today is only a way to say "internet NTIA root based names sold under the odd and blocking NSI/ICANN economical model". They also want a cross networks generalization (GNS) and cross systems and services universal names (uninames). They actually want to control their own management of their permanent or dynamic virtual network resources, a user network operating system (a "netix").

Our problem is the migration from the old DNS Internet limited system to the NGN (new generation network) GNS as part of a universal naming and sorting versatile multilingual and localized (vernacular) system.

This calls for an encapsulation of the current 20 years old solutions we cannot get rid of and which shown their robustness; into some higher sophisticated service layers. For example we need to extensively use the notion of virtual zones in multilingual and thematic DNS files generation. The same we need to understand how to concert and cooperate for the management of a multiple source single authoritative matrix root system, etc.

Basically we have to come back to the origin: to listen to the user demand and not to impose the ideas of R&D or politics people. IETF is only an engineering task force not a vision builder and Congress is not the ideal R&D replacement for Bell Labs (to keep with US images). Today, for a datacomms Frenchman the Internet is a very simple thing to understand: a world Minitel II without a built-in economical model. For having been, a time, the one in the field to buy the Minitel I for the USA and tested the first and only international USA Minitel access, I know that the most important feature of the Minitel was, and still is, actually its economical model (and its built-in no spam architecture). A model in a way copied by SMS, which makes its success.

jfc





































---------------------------------------------------------------------
To unsubscribe, e-mail: atlarge-discuss-unsubscribe@lists.fitug.de
For additional commands, e-mail: atlarge-discuss-help@lists.fitug.de