<br><br><div class="gmail_quote">On Thu, Mar 20, 2008 at 7:16 PM, Joe Abley <<a href="mailto:jabley@ca.afilias.info">jabley@ca.afilias.info</a>> wrote:<br><blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;">
<div class="Ih2E3d"><br>
On 19 Mar 2008, at 13:07 , Steve Wilcox wrote:<br>
<br>
> There is no 6to4 at Google today, whether there should be one I'm<br>
> seeing arguments pro and con. But it seems somewhat orthogonal as<br>
> there is no guarantee that having a 6to4 here would be the one that<br>
> inbound packets are taking and providing we make sure that the<br>
> outbound is going through a reliable path (I noted earlier that the<br>
> <a href="http://your.org" target="_blank">your.org</a> 6to4 was selected in Amsterdam) what would hosting a relay<br>
> fix?<br>
<br>
</div>The problem you could hope to fix at Google is having a stable route<br>
for 2002::/16 that leads somewhere that you control. That would<br>
eliminate the dependence on a third party for encaps of your reply<br>
traffic back towards 2002::/16-numbered clients.<br>
<div class="Ih2E3d"><br>
> Perhaps more pressure on transit providers to find better (and<br>
> untunneled)<br>
> IPv6 paths would also be helpful... ;-)<br>
<br>
</div>Well, true, but that will have near-zero effect on vista clients<br>
connected to networks whose last mile isn't going to feature native v6<br>
any time soon.</blockquote><div><br>What is the advantage to take a v4 client accessing a v4 service and then go and drop an asymmetric mtu reduced v6 tunnel in the middle anyway? <br><br>Steve<br></div></div><br>-- <br>
Global Infrastructure<br>Google Inc.