Quo Vadis IPv6 - Was: Re: [governance] IPv4 - IPv6 incompatiblity (was Re: Towards Singapore)
JFC Morfin
jefsey at jefsey.com
Sat Jun 18 16:48:14 EDT 2011
At 12:42 18/06/2011, Paul Wilson wrote:
>Avri,
>I also agree that, if that ever existed at all, it will probably
>never return, because there will always be reasons (good and bad) to
>put something in the way of certain connections to certain networks or devices.
>
>An Internet in which end-to-end connectivity is effectively
>impossible, or reserved as the privilege of a few, would be a very
>different Internet, and one which would be impossible I think to
>rescue. You may disagree, but I don't believe we are anywhere near
>that stage yet.
>
>Paul.
Interesting thread. IMHO, from observation and from what the emerging
IUse community identified, the network terminology should be restored
to fit with the Internet architectural intent that was in place at
the time people introduced them for a purpose, which has been blurred
by time and a lack of specific identification for what came afterward.
1. end to end is the dumb Internet. The first building stone.
2. edge to edge is more intricate [MHL1] as this implies OPES that no
one really developed as specified but that many actually use.
3. now, with IDNA2008, we identified that the Internet was to be more
complex and smarter in order to support diversity and the
presentation layer. And that it could in fact do it. This means
intelligence at the end(s), i.e. as per all the internet literature,
at the fringe.
As a result, our (IUse) approach is to identify that the network and
user ends are not the two parts of the same plug. There are two
plugs, one on the network side (identified by its IPv6 address) and
one on the user side (to be identified by its IDv6 address as we call
it: a local IID). This has at least five basic consequences:
1. the part between the network plug (end) and the user plug is the
Internet Use Interface, and actually the Intelligent Use Interface
(IUI) because it may relate the IUser with other parts of the digital
ecosystem other than the Internet.
2. the IPv6 address of the network plug may change while the IDv6
stays internally stable (and the associated IPv6, e.g. Fred Baker's
NPTR
(<http://tools.ietf.org/html/draft-mrw-nat66-14>http://tools.ietf.org/html/draft-mrw-nat66-14)).
3. the DNS extends to the whole digital ecosystem (WDE) and covers IP
addresses among other addressing schemes (one of them being local
IDv6) and, to that end, it needs adequate equivalent names in the
possible different formats, scripts, orthotypography, meanings (such
as IDv6 support through n+1 level labels). This is what we call the
ML-DNS (multi-layer DNS).
4. the IUI area is to be analyzed to be reduced into a clear
networking layer extension. It seems that once the presentation layer
has been supported (as for example in IDNA through the class header
(e.g. "xn--") two networking layers on the intelligent user side are
needed: the layer 9 for IUse applications (like the ML-DNS acting as
a smart front-end for the network DNS, and the same for Fred Baker's
NPT functions), and a layer 8 "interapplication system".
5. the appeals that I made against IESG and to IAB for not warning
the community were essentially to obtain objections to this vision (I
didn't receive any) and to clarify whose responsibility the IUI area
study, experimentation, and documentation was. The response was clear
enough: it is not the IETF area. However, the IETF is interested in
the results.
That said, there are probably several ways of conceiving the IUI
architecture and its use. I am mainly interested in testing the
concept through one that I call "interplus", where PLUS stands for
"Plugged Layers on the User Side", i.e. some networked real or
virtual bundled OPES that are transparent to the legacy technology,
down to the bits, and able to dialog on a fringe to fringe basis in
order to provide the user with an intelligent network experience.
The danger of the IUI would be a proliferation of IUI standalone
initiatives that could severely harm the Internet until it is well
understood and accepted. In my appeal to IAB I introduced and
extended system theory to help the modelization of the Internet and
other very, very large systems. Showing hackers, investors, and
governments the limits of what they can target in working in its area.
One of the reasons for the language evolution is also a constant
adequacy to the evolution of the network size and its related
architectural principles as identified by:
- RFC 1958 (1996): targeted at a large system, with the perpetual
need to adapt (perpetual change principle + functions/recipes).
- RFC 3439 (2002): targeted at a very large system, and as such based
upon the need (principle) of simplicity
- IDNA2008 (2008-2010): targeted at a very, very large/universal
system and, as such, having to call upon the power of the principle
of subsidiarity.
jfc
>--On 18 June 2011 3:25:59 PM +0800 Avri Doria <avri at ella.com> wrote:
>
>>Hi,
>>
>>A pure notion of end to end has lost to technology such as: firewalls,
>>VoIP servers, VoD servers, 3G service boxes, DSLAMs, the much hatted
>>NATs (which will exist in v6 as much as v4) etc...
>>
>>e2e these days stands, at best, for Edge to Edge. To speak of End to End
>>is to speak of a network lost in a time long past. And the Edge is
>>wherever the service provider wishes to put it.
>>
>>a.
>>
>>On 18 Jun 2011, at 15:12, Milton L Mueller wrote:
>>
>>>Parminder
>>>The fact of the matter is that it is collective processes that are
>>>taking us away from end to end (e2e) as much as your hated individual
>>>choice. Indeed, probably more the former than the latter. When govts or
>>>corporations install firewalls that filtering incoming and outgoing
>>>traffic for spam, malware, illegal content they are departing from e2e,
>>>usually in the name of collective values or legal requirements. So I am
>>>afraid your attempt to score a quick point against liberalism fails.
>>>
>>>--MM
>>>
>>>From: governance at lists.cpsr.org [mailto:governance at lists.cpsr.org] On
>>>Behalf Of parminder Sent: Friday, June 17, 2011 1:58 AM
>>>To: governance at lists.cpsr.org
>>>Subject: Re: Quo Vadis IPv6 - Was: Re: [governance] IPv4 - IPv6
>>>incompatiblity (was Re: Towards Singapore)
>>>
>>>Hi All
>>>
>>>Karl provide a concise description of what is happening and what went
>>>wrong with the internet. This analysis is best represented in the
>>>following paragraph
>>>
>>>(Quote starts)
>>>
>>>In addition users of the net no longer view the internet as a vehicle
>>>for the transport of packets from one IP address to another. Rather
>>>users today see the internet as a bag of applications. They don't care
>>>how the engines underneath work as long as the applications work. In
>>>other words, users don't care about the end-to-end principle.
>>>
>>>So we have to evolving forces:
>>>
>>> A) the desire of gov'ts and others to create and regulate choke points
>>> into/out-from their chunks of the net
>>>
>>> B) the the consumer-eye view of the net as a platform for applications
>>>
>>>These two forces combine to allow the net to evolve in a direction many
>>>of us do not like to think about - a kind of soft fragmentation that I
>>>call the "lumpy" internet. (quote from Karl's email ends)
>>>
>>>Apart of understanding what is happening, we are a political advocacy
>>>need to figure out 'what can and should be done about it'. And in this
>>>respect the following part of Karl's email is very instructive.
>>>
>>>"It would be sad indeed, from the point of civil liberties and
>>>expression, to kiss goodbye to the end-to-end principle. But that loss
>>>is as much due to users who view the network as applications as to any
>>>of the other forces - attractive toys often distract us from social
>>>values. "
>>>
>>>Is it not something new that 'individual users' are acting in this way,
>>>it is a way they or we always/ mostly behave. Not everything can be
>>>given the right direction and, when needed. corrected by individual
>>>users themselves acting independently (the techno-liberal view) or
>>>consumers voting through their dollars (the neo-liberal view). This also
>>>shows the strong overlaps of the techno-liberal and neo-liberal views in
>>>their practical outcome and impact, which in this case, for instance, is
>>>that we have nearly lost out on end-to-end principle, and the chances of
>>>building the Internet as really an egalitarian platform and force, which
>>>was the global society's hope for quite some time.
>>>
>>>We need collective/political processes, how much ever a techno-liberal,
>>>instinctively hates the very term, to guide our soceities in the
>>>direction we want it to go. The dream that the new technology paradigm
>>>will by itself do it for us is fast evaporating, and it is good time
>>>that we pulled our heads out of the proverbial sand. It is time that we,
>>>as a prime civil society group in the global IG arena, tries to come up
>>>with a sound political vision - both substantive and institutional - for
>>>how the Internet should serve the highest and most noble causes or
>>>social values that we espouse, or, in default, one will have to say,
>>>which we think we espouse.
>>>
>>>
>>>parminder
>>>
>>>
>>>On Friday 17 June 2011 04:32 AM, Karl Auerbach wrote:
>>>On 06/16/2011 02:30 AM, Izumi AIZU wrote:
>>>
>>>Or, what was the biggest reason/rationale not to make IPv6 compatible
>>>with IPv4....
>>>
>>>IPv6 had a somewhat difficult birth back in the early 1990's.
>>>
>>>There were actually several proposals - my own favorite was a thing
>>>called TUBA, which was an adaptation of the ISO/OSI connectionless
>>>network layer. There were several aspects that were interesting, and it
>>>had an address that was expansible up to 160bits. The hostility towards
>>>ISO/OSI is still strong today - much to the detriment of the internet -
>>>and was much stronger back then. So TUBA sank beneath the IETF's waves.
>>>
>>>It was recognized back then that there were several issues in play; the
>>>address size was recognized as but one issue among many.
>>>
>>>The format of the address was another - the variable size of the TUBA
>>>"NSAP" scared people who built routers because of the overhead of
>>>parsing a flexible address format.
>>>
>>>Which leads to the big issue that IPv6 never squarely faced - the issue
>>>of how routing information is created, aggregated, propagated, used, and
>>>withdrawn on the net. As a general rule the net's routing
>>>infrastructure needs to be able to propagate route information faster
>>>than the average rate of route change. And since those days we've
>>>learned to be a lot more skeptical about the authenticity of routing
>>>information.
>>>
>>>Early on there was much talk and though about IPv6 transition - how
>>>things might co-exist, even with intermediated interoperation of IPv4
>>>and IPv6 devices. But over time the energy to have a smooth transition
>>>withered and left us more with a conversion from IPv4 to IPv6 rather
>>>than a transition - the difference is subtle, conversion tends to be a
>>>more painful hurdle to leap than a transition.
>>>
>>>My own personal feeling is that IPv6 is too little and too late, that it
>>>will hit with about the same force as ISO/OSI - which like IPv6 had the
>>>backing of governments (GOSIP) and large companies (MAP - General
>>>Motors, TOP - Boeing).
>>>
>>>We are here talking on a mailing list in which many of the discussions
>>>are based on a recognition of the increasing desire of governments,
>>>intellectual property protectors, corporations, and others to stake out
>>>territories for them to control.
>>>
>>>In other words, we here are quite familiar with the fact that there are
>>>many forces that want to carve the internet up into fiefdoms and draw
>>>paywalls or tariff-walls or censorship lines around their dominions.
>>>
>>>In addition users of the net no longer view the internet as a vehicle
>>>for the transport of packets from one IP address to another. Rather
>>>users today see the internet as a bag of applications. They don't care
>>>how the engines underneath work as long as the applications work. In
>>>other words, users don't care about the end-to-end principle.
>>>
>>>So we have to evolving forces:
>>>
>>> A) the desire of gov'ts and others to create and regulate choke points
>>> into/out-from their chunks of the net
>>>
>>> B) the the consumer-eye view of the net as a platform for applications
>>>
>>>These two forces combine to allow the net to evolve in a direction many
>>>of us do not like to think about - a kind of soft fragmentation that I
>>>call the "lumpy" internet.
>>>
>>>Such a lumpy internet would be composed of distinct, but each fully
>>>formed, IPv4 (or IPv6) address spaces. Each lump would have its own
>>>routing infrastructure, own hierarchy, etc. If someone, like China or
>>>Comcast, needed more addresses than IPv4 could provide, they could
>>>create more lumps for themselves, each with a full 32-bit address space.
>>>
>>>These lumps would be connected by Application Level Gateways - things
>>>like web proxies. These would act as relays between the lumps.
>>>End-to-end addressing is by names, such as URIs or twitter tags or
>>>whatever seems appropriate.
>>>
>>>This may seem far fetched, but it is not unlike the way that mobile
>>>phone networks interconnect applications (voice being one application,
>>>texting be another) between competing, even hostile providers such as
>>>AT&T and Verizon.
>>>
>>>(These ALGs are much like a concept I proposed back in the 1980 and that
>>>Cisco revived a couple of years back - they are essentially the
>>>application layer analog to layer 3 IP routers.)
>>>
>>>Domain names would become contextual - their meaning would depend on the
>>>lump in which they were uttered. However, people don't like surprises
>>>and there would be a natural pressure for the DNS naming systems of
>>>different lumps to construct mechanisms or clearinghouses to assure a
>>>reasonable, but probably not perfect, degree of consistency, while
>>>allowing local/per-lump variations and extensions. Application level
>>>gateways might find that one of their jobs is mapping out
>>>inconsistencies of names between lumps.
>>>
>>>Internet lumps have some attractive properties, at least in the eyes of
>>>some:
>>>
>>> - They are "owned" so that the owner, whether that be a country or a
>>> corporation or a religious group, can open contact with the rest of
>>> the world only through guarded portals (i.e. their set of application
>>> gateways.)
>>>
>>> - Those portals can be taxed, censored, data-mined as desired. And
>>> since application level gateways pull user-data up to the application
>>> layer, there is no need for deep packet inspection technologies.
>>>
>>> - Since each lump is in itself a complete IPv4 space, there is no need
>>> for transition to IPv6. Each lump could give itself the entire 32-bit
>>> IPv4 address space, just as today we each re-use the same chunks of
>>> IPv4 private address space behind the NAT's in our homes.
>>>
>>> - Application level gateways between lumps do not require super-NATs,
>>> so the 64K limit on TCP/UDP port number issues do not arise.
>>>
>>>This not necessarily an attractive view of the future, but it is
>>>possible and, I believe, likely.
>>>
>>>It would be sad indeed, from the point of civil liberties and
>>>expression, to kiss goodbye to the end-to-end principle. But that loss
>>>is as much due to users who view the network as applications as to any
>>>of the other forces - attractive toys often distract us from social
>>>values.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.igcaucus.org/pipermail/governance/attachments/20110618/1121ed2f/attachment.htm>
-------------- next part --------------
____________________________________________________________
You received this message as a subscriber on the list:
governance at lists.cpsr.org
To be removed from the list, visit:
http://www.igcaucus.org/unsubscribing
For all other list information and functions, see:
http://lists.cpsr.org/lists/info/governance
To edit your profile and to find the IGC's charter, see:
http://www.igcaucus.org/
Translate this email: http://translate.google.com/translate_t
More information about the Governance
mailing list