Why is Keith so desperately wedged on one particular means of achieving his objective? It is entirely possible to make peer to peer applications work well with NAT, it is entirely possible even to make a server application work well with NAT. We are running out of IPv4 addresses and it is clear that IPv6 is not going to deploy fast enough to allow people to dispense with IPv4 before the exhaustion point is reached. Unless someone happens to have a working time machine handy the only plausible means of getting two billion plus users to attach multiple devices to the IPv4 Internet is for some devices to share an address. That means some form of NAT. I don't see any reason to expect that my personal Internet needs should require more than an IPv6 /96 and an IPv4 /38. That is 256 ports worth of pooled IPv4 connectivity. What I want is a little more than a NAT box, I was going to call it an Internet 2 box until Thomas Roessler suggested that the name Internet 2.0 would be more current. The basic concept of an I2.0 box is that it either plugs into or is a part of the Internet modem device (cable, ADSL, WiMAX, whatever). It has a 'network side' and an 'Internet side'. The Internet side can run IPv4, IPv6 or dual stack, dual stack support may be a full IPv4 address or a share thereof. The network side can consist of any combination of IPv6, IPv4 devices in any combination whatsoever. IPv4 service is by default via NAT. The I2.0 supports some form of port request protocol (UPnP, NAT-PIMP, whatever) and has sufficient smarts to fix up for well known, well used legacy protocols where necessary (FTP, HTTP, SIP). It has a local DNS resolver and makes necessary service assignment information available to applications on the network side via DNS SRV, TXT records in the .local domain. New application protocols are required to be I2.0 compliant, that means using the DNS as their service discovery mechanism including advertising the IPv4/v6 transition support. The only administration a user is required to do on an I2.0 box is to tell it which machines are allowed to connect to the network and which applications running on those machines are allowed to access particular Internetwork resources. Once a rough draft of the I2.0 spec was available manufacturers could advertise their product as being I2.0 compliant provided only that they agree to make an upgrade patch available to support the additional functionality necessary before we reach the IPv4 address crunch. The killer application for the I2.0 box would be to make videoconferencing work reliably on a shared IPv4 connection. Today this is simply not possible without endless amounts of grief unless you have a technical expert at BOTH ends of the wire. Untold misery is the result of trying to persuade your grandmother to configure her NAT box to forward the correct IP ports, for added fun try doing this where said grandmother is using VOIP off the same box. > -----Original Message----- > From: Sam Hartman [mailto:hartmans@xxxxxxx] > Sent: Tuesday, August 21, 2007 10:55 AM > To: Keith Moore > Cc: RJ Atkinson; ietf@xxxxxxxx > Subject: Re: IPv6 addresses really are scarce after all > > >>>>> "Keith" == Keith Moore <moore@xxxxxxxxxx> writes: > > >> Fourth, lots of folks (me included) happen to find it > >> convenient to use NAT between my site/house/office and my > >> upstream provider. > Keith> do you also find it "convenient" that NAT has effectively > Keith> thwarted the deployment of huge numbers of new > Keith> applications, significantly raised the cost of deploying > Keith> others, and harmed the reliability of all applications? > > I find the tradeoffs work in favor of NAT; I expect this to > be true both for V4 and V6. > > _______________________________________________ > Ietf mailing list > Ietf@xxxxxxxx > https://www1.ietf.org/mailman/listinfo/ietf > _______________________________________________ Ietf@xxxxxxxx https://www1.ietf.org/mailman/listinfo/ietf