kre,

>> Tony and I are proposing schemes that are aggregatable and
>> that are not tied to a provider.

> kre wrote:
> Both those schemes are geographic based addresses - these aggregate
> if and only if one assumes that areas that are geographically close
> are also topologically close.

This is absolutely untrue. I will let Tony defend his own, but for the
two different methods that ipv6mh is working on, there is no relation
whatsoever between the network/fiber plant topology and the geographical
aggregation.


> They're provider independent only of there are multiple providers
> that have agreed to cooperate with each other and share the
> prefix (essentially routing packets to each other when they arrive
> at the "wrong" place).

This is another wrong assumption. We know that asking ISPs to advertise
a geographical aggregate that contains competitor's addresses is not
going to fly. I would not do it myself. The solutions that we are
developing do _not_ require transporting competitor's traffic.


> While that is entirely possible to assume might happen in some
> high density usage parts of the world, where multiple providers
> can all exist and make a profit, in much of the world, the simply
> aren't enough active users to support the infrastructure for
> multiple providers.

We do not require any local infrastructure.

> Anyone would always be free to connect to a different provider,
> by simply connecting to a more distant location of course - but
> then their address would not (could not if any aggregation at all
> is to be achieved) be based upon their geography, but the
> provider's instead.  That is, to change providers an address
> change is likely to be required for many (perhaps most) and for
> those it isn't, the provider choice will be limited to those who
> have managed to join the local provider club.

True, but what if joining the local provider club is free? You would
have a point if being present at some kind of IX was required, but it's
not.


> On the other issue ... anyone can (attempt to) pay any ISP, or set
> of ISPs to carry any address.   The rish is certainly no greater of
> that happening with a SL address with a site-id embedded, than it is
> for a global address allocated by a different provider, or a
> geographic address from some other region.

A valid point. Please note that we are not claiming that geographical
addresses alone will solve all of the problems at once.


> If anything, the risk is less with SL addresses, as they can be
> clearly labelled "for local use only", lowering the chances that
> people will ever decide they would like to interpret them as global
> addresses (all of these things are just numbers, so perceptions,
> and what the ISPs will agree to do are all that matters anyway).
> Global addresses are expected to be globally visible, and there's
> no reason at all to assume that people won't go to a competitor
> ISP and say "I will connect to you, and pay you all these $'s if
> you will agree to advertise the prefix I have already been
> allocated this other way"

True enough. However, you forgot two important things:

1. If we make SLs with site IDs globally unique, there might be no
assignement (or the assignement might be automatically granted by the
addressing architecture, say a 37-bot hash of the MAC address, or
whatever). With the geographic scheme we are designing, there would
still be the need to go to a RIR and register a block, so we might still
decide that someone with a dial-up 28.8k modem does not qualify for a
portable, globally unique /48 block.

2. SLs do would not have a supporting protocol if used globally (how
could there be one, if they are not supposed to be used that way). The
way I see the geographical addresses is a support of the multihoming
protocol. And if there is a protocol, there can be safeties and policy
enforcement features built in.


> (and even say to ISPs, "I will connect > to you, and you allocate me
> an address, but you agree that once allocated you can never reclaim
> the address, no matter whether I stop paying you or not").

Not very realistic, IMHO. First, no ISP would put that in writing. The,
what happens if the ISP bellies up?

Michel.


--------------------------------------------------------------------
IETF IPng Working Group Mailing List
IPng Home Page:                      http://playground.sun.com/ipng
FTP archive:                      ftp://playground.sun.com/pub/ipng
Direct all administrative requests to [EMAIL PROTECTED]
--------------------------------------------------------------------

Reply via email to