On Thu, Dec 6, 2018 at 12:31 AM Joe Touch <touch@xxxxxxxxxxxxxx> wrote:
Additionally, packets don’t emerge from different mole endpoints or are IP processed in any way. The mold acts like a wire, which is fine. That can be done by IP tunnels too. But not routers that converge and diverge packets.
this is patently incorrect. but also not particularly important I suppose.
Again, if HBH headers are meaningless and not needed, then go through the proper process and remove them from IPv6. If not, stop trying to hobble this protocol to the point where we all realize why nobody wants to use it.
I think everyone here is actually happy to see v6 progress.
I think the option to ignore if you are an adult is a fine option, and actually helps deploy v6 further/faster/more.
I do agree that IF the end systems expect their core networking (across the wider internet, not in their campus deployment) to 'do something' to the packets aside from 'forward' ... they are going to be sad. I don't know how much processing (aside from 'forward') matters though on the wider internet. What L3 does with my packet EXCEPT deliver it .. is unimportant to me, please just deliver my packet(s).
note I'm not trying to be intentionally combative, just attempting to say: "the best answer for the user here is PROBABLY to just have the core ignore all the EH business entirely"
