[Air-l] not wishing to start a war . . .

[elw at stderr.org]

>> In my book the packet doesn't care what it is carrying. Carrying a bit 
>> of binary code or a text message is all the same. Doesn't make the 
>> packet a container of expressive content.  That content appears to the 
>> reader on the other end.
>
> While Bob has produced some argument against this position from a 
> conceptual perspective (and I think Bob and Denise are both right 
> depending on one's conceptual starting point), it might be worth 
> pondering the longer-term impact of IPv6 on the received wisdom that the 
> packet doesn't care what it is carrying. One of the empirical reasons


how does an increase in possible address size correlate to what a packet 
carries?

or are we talking about QoS and related *extensions*, now?  (none of which 
necessarily requires ipv6-ish tech to function...)


> remained in transit and not being sent or received. Clearly though IPv6, 
> by increasing the size of the header and permitting significant 
> additional information about the packet might produce experiences in the 
> world which will require some revision of the 'packet doesn't care'.


... a lot of these revisions are more socially driven than technologically 
determined.  backed up by industry 'best practices' and the like, rather 
than simply the technology's capacity to do any particular thing.


> coffee somewhere at an AoIR conference I think) which is mostly divorced 
> from packet contents. IPv6 however has the potential to bring some 
> interaction between packet content and packet-switching system.


IPv6 is highly overrated and often assumed to be the solution for a number 
of problems which are either 1) unrelated or 2) only tangentially 
connected.



--elijah