thank you for your comment even if it does not discuss the "internet national survival kit". I am afraid it continues a qui pro quo where we often say the same thing but from different points of view (not vision). Where you look from inside your technology, and me from a user's point of view, from outside and global (English/French meaning) putting Internet in competition with other alternatives (existing or possible).
I respond here. Because some who read could keep a strange idea of my positions (I did not even alluded to). Also because John is always interesting. Even when wrong about me, the topics are of interest. I will try to keep it technical, but it responds John who did not addressed the thread.
jfc
<- John and interested Members only - Thx ->
Dear John,
I appreciate much that you know to stay technical while caring about societal and political aspects (this is why you were IMHO a very good choice for the BoD).
But this is NOT the matter here.
At 17:47 29/11/03, John C Klensin wrote:
You should also entertain the hypothesis that no one has commented on those issues/suggestions because they are have been discussed too many times before and are inconsistent with the visions that drive the Internet. Some of them have even been the subject of fairly careful evaluation and associated statements, e.g., RFC 2826 on the unique DNS root issue (often summarized as "which part of 'unique' are you having trouble understanding?").
Please ... this is not the matter but I will respond as usual "the part you make it to play wrong". BTW your unique root is two different accesses which can be hacked differently. First risk.
The question is not the suggestions but "are there better reponses? would they they reduce or increase the death toll". In the cases described by the White House, the comparable cases in other countries contexts, the cases created by the White House strategy and the cases created by the other countries comparable strategies?
I understand no one wants to commit. We are not discussing development, architecture, etc. But technical aspects of Best Practices for national survival (economy, administration, transport, health, schools, banks, airlines, power, etc. etc.) in front of a large Internet failure resulting from an internal collapse, a international situation, an external catastrophe. Real - not intellectual - world stuff.
I think, as we have discussed in the past, that your vision of the Internet and its future differs from mine and that of many, probably most, of the people on this list. I would characterize your picture, I hope not too inaccurately, as one in which connectivity and the flow of information are driven (not unlike the PSTN) by bilateral agreements between countries. In that sort of world, different countries may reasonably establish different views of the DNS and different address spaces, with inter-country communications occurring through gateways that, among other things, can keep those views and address spaces separate.
This description is totally inaccurate. Even opposite (even it should support that scheme as a possible scheme and that I accept every consistent scheme). As you know my model is a cylinder where the Internet 4 layers are just a slice, as the OSI is and others.
You quote the DNS, please refer to the Universal Naming System (uniname/universal name). That is information society, computers, IP rights, culture, dictionaries, automatic translation, etc.). This is where national/religious/community, etc. etc. cultures belong and are to be defended. Not the IETF cup of tea.
As part of the UNS, the brainware (people) and the software (applications, encryption, etc.) identify network related names. This is the GNS (egenalized) which must eventually permit a harmonized communication support. When I send a fax or an MMS or a mail or access a web site I want the name to be and behave the same. Even IPv6 wants that (HIP)! I give the example of local root ".sos". It will become the unique referent for all the alarms, on all the systems, in all the languages - even it may translate in many different vernacular ways of scripting (uninames), sounds or icons.
As part of the GNS there is the Internet related restriction named the "legacy DNS" from the root names allocated in 84 and added in 2000.
All these are different layers and in each layers/languages, ways of use/systems. What the user wants is to see them to look the same. The part of unique you do not understand is that the world is NOT uniquely Internet. If you want to keep thinking hierarchical (which is very restrictive way), you should think http://domain.name.dns.gns.uns . If name in "dns.gns.uns" are the same and permit DNS usages to look unique (what IPC fights for as famous names). But that does not prevent that many other uninames to related to a dns name. This is what Keith Moore and Paul Hoffmann are actually trying to word out at imaa, from an Internet technology perspective.
I said hierarchy was a very very restrictive way. You must support local occurrences, you must support abbreviations. Do you put "USA" at the of each letter you send to US people. Do you put your street address on the envelope when you drop a letter in the room of your kids. When you write to your "beloved wife" do you confuse her with mine because we both use the same subscription? All these are however very precise designations of unique entities we happen to have used for millenaries. We do not want to create confusion in being forced into an "EITF voted reality" to comply with a 20 years old program.
It is not simple. But who said life was simple :-)
In my world, direct, end-to-end global connectivity, interoperability, and integrity of DNS and URI references are very important -- for interpersonal communication, for commerce, for intellectual development and dissemination, and perhaps even as a religious principle. That belief has caused me to spend most of my time in the last few years on internationalization issues, not to empower governments, but to permit better communication among people (and, indeed, to reduce the belief by governments that they need to "solve" the problem, probably in some xenophobic way).
Spent much time, pains and 100s of thousands of miles at the begining, explaining Govs it would be nice that they put they network data and information in my unique file. Took years but eventually worked and was copied. The part about that uniqueness is it can be copied and stay unique.
What you fail to see right now is that the system you support is Babel. Please reread the Babel tower story. What men were confronted to was to build the tower with stones of sand and tar. What figured they were prisonners from a unique language and thinking. God never punished them !! He freed them - through many languages - and like sand of the sea, the sand of the people could flow. Read Church Fathers about that. Very interesting to understand the DNS and networks. I am sure that you are one of those who may really be interested.
Communication among people is NOT to make them use an unique language, with a unique thinking and an unique vision and an unique gov. It is to find acceptable ways to bridge there differences (languages, cultures, history...) without reducing them for them to inter-relate. Far, far more complex.
A reasonable accepted way is to use numbers. For years I support that we may use numbers from 0 to Z and "-" is a convenient thing. We added "@" recently. I suppose we should also use "#". And this is what is being done in many technologies now (and in URLs). But (see above) this is DNS or GNS. Some acceptable common trade-off to be able to relate together.
With regard to ICANN and its processes, I don't much like the way a good deal of that has turned out, even while I believe that things are gradually getting better. I lament the set of decisions that led to the US Govt deciding that it needed to be actively involved and to some of the risks, delays, and socially undesirable statements that situation has created. At the same time, all of the alternatives continue to strike me as much worse, including moving the technical/administrative issues into forums in which variations on the theme of "we don't like reality, so we will vote it to be different, regardless of what they might do to the Internet or human communications in general".
This is not technical. But I am in partial agreement.
Partial because the nature of ICANN as a de facto US agency and an international coordinating body will never permit to be a multinational concertation (documented that in a previous mail) hub it has to be.
So, while ICANN, IMO, continues to need careful watching -- most importantly to be sure that it does not expand into "governance" issues that are outside its rational scope-- I don't see "give it to XXX" or "everyone runs off in his own direction" as viable alternatives.
Fundamental difference in our use of the word "governance". In my local idiom "governance" (since XIIIth century) means more or less what - you - you would like it to be. But the ICANN people are tied by their common native understanding of the word, which is much more directive. This is technical. Nothing political. Human languages are like computer languages. You depend on the language you use to think and work together. If your common references are not what you think they should be you meet problems. A brainware bug.
On the other hand, one of the nice things about the network as it is now constituted is that anyone has the option of opting-out: disconnecting, setting up a private DNS and a private addressing system, and communicating, if at all, through a restrictive, address-and-protocol-translating gateway.
This is absurd. Only ICANN does that. Voting its own reality. What is unique is reality. Also what is unique is the incredible way ICANN wants to decide about reality. I had for long a slogan for ICANN Esther used for a lecture or a book "let get real".
We even know how to run IP over X.25 and X.75, and that option is available as well.
Happily, in here the Customs and part of the Army runs X.25 under IP. Protects us.
The question of who will miss anyone who takes that opt-out option is an interesting one sociologically, but the Internet has sufficient critical mass at this point, and is sufficiently important commercially in most of the world, that "opportunity to shoot yourself in the foot" might figure into such an analysis.
This is an incredible thing to say when we are precisely confronted to Govs representing 3/4 of the humanity saying "No! you will not appropriate our common network!". Or have you already opted-out 80 of the mankind?
Anyway, this NOT the matter, John. The mater is to help these 80% survive the Internet until the world fixes a new generation network fulfilling your hopes better, more securely, etc.
If you are convinced of the viability of your ideas, by all means go off and try them: just be sure that your namespaces and addresses don't leak into the real network.
Which ideas???
I have listed a survival kit. From different propositions. By different groups. For different purposes. That I certainly find acceptable and coherent. And I say "please say what is wrong with them before having them enforced or describe a replacement". I doubt that this was discussed beofre orplase povide the URL.
This follows a study of an ICP-3 conformant test bed. BTW I note there is only one thing I disagree with your ICANN there, is when it says this may lead to the unique authoritative root concept to be deprecacted. Absurd, as saying that in the future there could be several colors to the sun . But which sun are we talking about.
Take care. jfc