On Nov 8, 2012, at 7:25 AM, Brian E Carpenter <[email protected]> wrote: > Without some fundamental surgery on the IPv6 specs, I fear that is true, > so does it have to become (gulp) a feature of the homenet architecture?
I think it does, or the architecture becomes irrelevant because no-one will be brave enough to ship boxes that implement it. And so we need to specify how it happens. You are probably already aware that I strongly favor a routed ULA in the homenet (as opposed to the "keep using the ISP prefix after it expires kludge), but that doesn't solve this problem without NPTv6, which I think is a terrible idea. Bridging preserves end-to-end; I think that's far more important than preserving subnetting. That said, it would be nice if devices flashed a big red light or something if the ISP-delegated prefix is a /64. I don't know of a way to specify that in the document, though. Aside from the disagreement about ULA, I think your suggested modification to the text is right—the text as it is now recommends the wrong workaround to brokenness. _______________________________________________ homenet mailing list [email protected] https://www.ietf.org/mailman/listinfo/homenet
