begin quoting Tracy R Reed as of Sat, May 20, 2006 at 04:45:14PM -0700: > James G. Sack (jim) wrote: > >..then, what size range of IPs could an end-user expect to get? > > Who knows. It is really at the discretion of the provider. But I suspect > users will be able to get as many as they want because there are so many > available. They are there for the asking.
I don't see an economic incentive for an ISP to give 'em away; you might get a half-dozen where you now get one.... but the consumers aren't going to choose an ISP based on how many IPs they get. They'll still choose according to how slick the commercials are and how fast they can download web-pages. Do you really think AOL is going to hand out IP addresses by the handful? > >Hmmm, is IPv6 routing qualitatively any different than IPv4? > > No, it is pretty much the same. Address tables get larger. > >I suppose that IPs might be allocated to end-users in fairly sizable > >blocks, without significantly fragmenting the IP space. But beancounters > >will doubtless charge by the pound (er.. by the K, or maybe 8K -- how > >many do you need TR?, SS?). > > They would like to be able to charge but when one ISP has the policy of > giving out as many IP's as their customers want because their IANA > allocated IPv4 class C has changed from 256 IP's to 1099511627776 > (2^32*256) IPv6 IP's everyone else will follow. That's awfully pie-in-the-sky. Remember, most customers won't know why more IP addresses are useful. They won't care. > >Will DHCP go away? > > No. Although IPv6 hosts can self assign IP's based on MAC they will > still need to be told what nameserver and gateway etc. to use. DHCP is > still very useful and has very little downside as far as I can see so I > am not rooting for it to disappear any time soon. Misconfigured/feature-poor DHCP servers have been a far greater problem for me than any misconfigured/feature-poor NAT boxen. [snip] > >What'l (how _does_ one spell that?) it take to get the US (and Europe) > >into the IPv6 boat? > > Internet in the US is a relatively old technology with lots of s/old/mature/ > entrenched old standards like IPv4. It is always easier to roll out IPv6 > when you are starting from scratch like much of the world is. For the US > to really change it is going to take some sort of economic hardship. > IPv4 addresses will have to get scarce enough and expensive enough that > we tolerate the pain of changing. Unfortunately, NAT is only prolonging s/Unfortunately,// > our misery. But NAT is really only detrimental to the end-users and does ...who use software written by incompetent and/or lazy programmers... > not hurt the ISP's themselves in any way (it only helps them) so there > is no incentive for them to implement IPv6 on their networks. Another Ya gotta make it sexy to upgrade. Like it is to upgrade PCs. Obviously we need an IPv8. > factor is that someday a lot of good information that people really want > will be found only on IPv6 as the rest of the world passes us by. That > might be an incentive to change also. Chicken-and-egg... most of the information is available on IPv4. So IPv6 will, by that argument, die. -- _ |\_ \| -- [email protected] http://www.kernel-panic.org/cgi-bin/mailman/listinfo/kplug-list
