[governance] IPv4 - IPv6 incompatiblity (was Re: Towards Singapore)

[Karl Auerbach]

On 06/17/2011 04:27 AM, John Curran wrote:

>    Excellent points.  I also go to great length to stress how the protocols
>    that we use everyday (such as http for web and smtp for email) run without
>    change over IPv4 and IPv6 equally well and that this interoperability was
>    an explicit design goal.

Yes, getting rid of protocols - such as FTP - that tend to carry IP 
addresses at the application level were troublesome.

However there is yet another issue to consider with IPv4 and IPv6:

I have a unique place from which to view some of these things.  Our 
company (InterWorking Labs - iwl.com) builds gear to help network 
implementers exercise their protocol implementations under impaired (but 
usually legitimate) network conditions.  (Recently we've been using the 
gear to explore the bufferbloat issue.)

(By-the-way, we do build gear to test IPv6.)

What we find is that even after decades of implementation experience, 
even basic IPv4 layer code still is often of poor quality.

It's been an unfortunate trait of many vendors that they consider their 
work done and a product ready to ship once a chunk of network code runs 
on their lab network.

That vendor trait probably is not going to improve with IPv6.  Which 
means that along with IPv6 will probably come a lot of buggy code.

For several years I've advocated measures to advance the internet 
towards being the kind of lifeline grade utility that people believe it 
is.  (The first of those measures is simply a change in mental attitude 
so that one views the net as a complex distributed system, with feedback 
loops, oscillations, etc, rather than a bag of separate computers.)

I have concern that with IPv6 that we will see a retreat in terms of 
internet stability and reliability, i.e. that the net will be further 
from my goal of it becoming a lifeline grade utility, at least for some 
number of years.

I wear two hats: I am both a techie and a lawyer.  One thing that has 
always surprised when wearing the lawyer hat me is how software, 
particularly network software, has so far avoided the strong hand of 
what is here in the US called "product liability".

In a world where the internet increasingly approximates a utility, I 
wonder how much longer that immunity can last?

Consumers of goods and services tend to presume that each generation of 
those products and services will be more safe than their predecessors. 
That presumption will, I suspect, not be true for several years with 
IPv6 based versions of IPv4 products.

My (highly anthropomorphized) hat looks at IPv6 and steals a line from 
that great film "All About Eve" - "Fasten your seatbelts.  It's going to 
be a bumpy night."

	--karl--

____________________________________________________________
You received this message as a subscriber on the list:
     governance at lists.cpsr.org
To be removed from the list, visit:
     http://www.igcaucus.org/unsubscribing

For all other list information and functions, see:
     http://lists.cpsr.org/lists/info/governance
To edit your profile and to find the IGC's charter, see:
     http://www.igcaucus.org/

Translate this email: http://translate.google.com/translate_t