[discuss] Anything specific? Was: Re: IPv6 Deployment and IG

JFC Morfin jefsey at jefsey.com
Mon Dec 30 18:03:15 UTC 2013


At 17:49 30/12/2013, Andrew Sullivan wrote:
>I confess I didn't understand your answer.  Also, I find it incredibly
>hard to read your email since it's always HTML only, so I have to work
>around all the formatting cruft in there.  You could send
>multipart/alternative to make it easier for those of us who don't use
>HTML mail readers; I would appreciate it.

Sorry. I apologize, I did not realiize that. I use quite an old 
Eudora mail agent and usually I must use the Gmail copy to read many 
mails. Mine were correctly back.

All I tried to tell you is:

1. read the documents I quoted in an open-minded way. I know this may 
be difficult to someone used to the IETF restricted culture (cf. why 
RFC 5895 was not accepted in spite of the AD and Vint's support), but 
this may result from reading them in sequence, and along the RFC 6852 
new spirit in spite of its uncompletedness. In telling you this I 
just did not want to influence you, and benefit from technical 
judgment I trust.

2. If you do not think there is any contradiction, then I will not be 
able to go any further explaining you the problem. If you find 
contradictions you will know as much as I am.


To respond your other questions you have to accept my premises, 
otherwise you cannot understand what I discuss. My premise is that 
the Internet was designed to support multiple technologies with a 
"loose" definition of the world "local". This is what Vint wrote, in 
full agreement with the French language "globalité" meaning that I 
attach to  Louis Pouzin's "catenet". This means *simultaneous* 
multiple technologies/formats on the local networks, hence on their 
gateways, are possible. However this is formally repudiated by Vint. 
This is what I call the Internet BUG bogged down to a uniform/single 
global access control.

This multiglobal capacity (which was intrinsic to the Tymnet 
technology,100% used by public international services in 1983) was 
also partly supported (at the interprocess level) by OSI layer six 
through formats (Internet only really supports the text format). The 
demand/need now is for implementing a full catenet (i.e. 
"multi-global") support at inter-process level. The leadership taken 
by/given to Fadi (from GS1/IBM) leads to think that it is on his 
agenda, or it is the on-going trend, to integrate EDI support. 
Depending on the way this happens there may be one other or more 
ICANN-like international bodies being created by trade bodies or governments.

It is therefore probable that the outcome of Sao Paulo will be two or 
more IETFs and ICANNs. Sponsored by WCIT majority members and/or 
others. If we want to avoid bad effects of what ISOC calls 
balkanization, and I call opening, we will need quite an IG consensus 
among these bodies, involving ITU, Telcos, OpenStand, IUsers, China, 
Russia, OECD, Europe, etc. We need to prepare that possible consensus 
in being ready and preventing conflicts. What ISOC, Russ, etc. are 
trying is this with probably the belief or the hope that talking of 
MSism and "globalization" could ease the situation.


If I spend time on this list, it is not to influence it or be trolled 
by Jorge, it is to try and get new ideas and test the water, to 
prepare my own INTLNET positions and budget plans in the coming 
multi-R&D governance and IANA debate or race.

I do not know when exactly this happens. The time of the dumb text 
datagram is over. We are back in smart-networking.

jfc











> > - saying that there is no need for a technical governance can only means
>
>I don't think anyone is saying that, though.  If one takes the meaning
>of "governance" broadly, then we have things like the RIRs and the ARO
>and the ICANN procedures and so on, and what I have been asking is
>what _more_ is needed there.  Your message here (and your previous one
>I didn't understand) offers me no direction for the answer to that
>question.  (If we take the meaning of "governance" to be outside those
>activities like what RIRs do and so on, then we have a semantic
>question, and we still have open the question of what more needs to be
>done.)
>
> > the parameters, but the interrelations of the authorities which establish
>܈\ÙH[Kœœ€ ¤'F†­æ²F†Bw2â­çFW&W7F­àg question, but one that is potentially
>descriptive rather than presciptive; and I think "governance" is
>prescriptive.
>
> > (1.1.) you will not be told any particular things that people 
> <u>want</uˆÈÙYHÚ[™ÙY‚‚•\È\ÈH™\@uirement for a cure without a 
> statement even of symptoms
>of a disease.  See below.
>
> > (1.2.) you will never have <u>proposals</u›Üˆ[™ÜÈÈ™HÚ[™ÙYˆ[ÝBˆÚ[]™@:<br>
> > - either analysis (like mine) to tell you where analysers think the
> > people reported problems come from (and you will most probably have
>Y™™\™[[˜[\Ù\ÈÈÀompare).
>
>In this case, I would like to have an analysis I can understand.
>
>H܈[™›Ü›X][ÛˆÛˆÛÜšÜÈ[™ØYÙY€y &quot;lead users&quot; following
>ÛÛYH
>܈HÞ[\Ú\À) of these analyses.
>
>Please, a pointer, then.
>
> > Where we need mutual governance it is to prevent confusion.
>
>[
]
>
>
>ÛÝ[Ý[™H[˜ÛÛ\]Y[ˆ^H[šÚ[™ËˆH[Herefore obliged to
> > convince those who will help me and use my deliverables, showing them
> > that my vision is correct. We are back to &quot;running code&quot;. This
> > obliges to a perpetual enhancement.
>
>But what "governance", _more than_ what we already have, is necessary
>for this?  It seems to me your example is possibly evidence the answer
>is "none": you can get the code points you need, and you can get the
>co-ordination you need by publishing Internet-Drafts, and in a
>smart-edge system that's everything you need.
>
>Many of the apparent fans of "more governance" don't actually seem to
>want that sort of freedom.  They seem instead to want to ossify the
>network to ensure that it can't evolve more, because a static target
>is one that is easier to control.
>
>]\È\È[ÝKÝ\È
>XÚšXØ[ - political solvers) to translate it in
>[™ÜÈÈÛÜœ€ect and to propose solutions.<brœ‚‚]]\È™\ÜÀterous if, 
>when I ask for more details, nobody will
>answer.  Except in deficient polities, we do not make laws for the
>sake of it.  We make laws to attempt to address specific issues.  When
>someone proposes a new measure, the question is quite correctly what
>that measure is supposed to address.
>
>If I go to the doctor and I say, "It hurts," the doctor asks me for
>specifics: how much does it hurt, where, how long has it been going
>on, and so onH[H›Àt asking people to diagnose themselves, but to
>say _something_ about what their problem is beyond, "We need this
>solution."  Indeed, if I went to the doctor and said, "Give me a
>prescription for Lipitor," the doctor would be quite correct in asking
>about things that made me think I needed Lipitor.  Similarly if I
>asked for a prescription for morphine.  We do not expect to go to the
>doctor and say, "This is what I need."  We expect to describe
>symptoms.
>
>Why should this case be different‚‚¶#æ'æRâF†RÆ6° of users' 
>post-Snowden trust in the technology, hence
> > in its engineering.and therefore in its governance's capacity to provoque
>H™XÙ\ÜØ\žH™\ÙX\˜Ú›Ü›X[^˜][Û‹ development, validation and
>\Þ[Y[Ý˜]YÚY\ˈ‚"@aybe not.  If the issue is "lack of trust" then 
>what we need are
>measures that will cause the technology to be trustworthy _somehow_.
>Governance may be one way to achieve that.  On the other hand,
>governance mechanisms are as easily abused as anything else: Arthur
>Andersen was a key part of the governance of Enron.
>
>Best regards,
>
>A
>
>--
>Andrew Sullivan
>ajs at anvilwalrusden.com
>
>_______________________________________________
>discuss mailing list
>discuss at 1net.org
>http://1net.org/mailman/listinfo/discuss




More information about the discuss mailing list