[arin-ppml] inevitability of NAT?

[Owen DeLong]

On Feb 9, 2011, at 2:53 PM, Scott Helms wrote:

> On 2/9/2011 5:46 PM, Owen DeLong wrote:
>> 
>> End users want to be able to do their hosted games on their Playstation. End users want to be able to get to their home networks through services like Go2mypc or whatever it's called (I don't need a service, I don't use NAT, so, I don't pay a lot of attention to the services available).
> 
> I can't name a mass market service today that doesn't work through NAT and the ones you point out are certainly do.
> 
They work through CONSUMER NAT because consumer NAT has nat traversal facilities like uPNP.

LSN/CGNAT is a whole different ballgame.

> Playstation Network works, by design, flawlessly through NAT and is _not_ a peer to peer application.  All traffic goes to the PSN servers and then to clients.  Go2MyPC is designed for NAT and in fact one could argue that it wouldn't exist if NAT weren't wide spread.

Yes. You entirely missed the point of my statement. These things work today because CONSUMER NAT
provides NAT Traversal technologies (NAT-PMP, STUN, uPNP, etc.) that make things work through NAT
and which these applications depend on.

LSN/CGNAT will not provide these facilities and in many cases will be implemented in a second layer
on top of the existing consumer NAT.

This combination will break all of those services.

>> End users want stuff that works today to work tomorrow.
>> 
>> All of those things represent a strong interest in avoiding LSN at  the carrier level.
> 
My statement here stands.

Owen