Re: [v6ops] Enterprises won't deploy IPv6 for IPv6's sake (as they didn't IPv4, IPX, XNS, or DECNet)

[Mark ZZZ Smith <[e-mail]>]

----- Original Message -----
> From: Alexandru Petrescu <[e-mail]>
> To: "Templin, Fred L" <[e-mail]>; "[e-mail]" <[e-mail]>
> Cc: 
> Sent: Saturday, 11 January 2014 9:31 PM
> Subject: Re: [v6ops] Enterprises won't deploy IPv6 for IPv6's sake (as they didn't IPv4, IPX, XNS, or DECNet)
> 
> Fred,
> 
> Thanks for the message.
> 
> Le 10/01/2014 17:37, Templin, Fred L a écrit :
>>  Hi Alex,
>> 

<snip>

> 
> Given that, one wants to make it first IPv4 (IPv4 innovation is still
> needed in some protcols), and plan for as little effort as possible to
> deploy to IPv6.  I.e. make them as much as possible look the same.  If
> not, there is additional software effort, additional money to look for,
> additional time to add to end of project.
> 


So let's imagine IPv6 was just IPv4 with bigger addresses, nothing else changed, which is what I think you're saying some people think it should have been. How would deploying IPv4 with bigger addresses, in addition to IPv4 with small addresses, save or make an enterprise money today or in the near future?

To pick a hypothetical example, imagine a golf club manufacturer. How would deploying IPv6 (of either IPv4 + bigger addresses, or what we have today) decrease the cost of manufacturing golf clubs, or increase the sales of golf clubs? I can't think of how IPv6 would decrease their manufacturing costs, and I don't think a "We've got an IPv6 network" sticker on the box will increase golf club sales significantly (they might sell a extra few hundred to network engineers, but that would be it).

In the views of some, if IPv6 had just been IPv4 with bigger addresses it might slightly improve the business case to deploy it, because it would have been less to learn, and therefore theoretically less cost to operate. Lower costs are great, but less costs don't matter if you don't need what the technology provides.

I used the following in another off-list email about this topic, I think it is worth posting here:

"I once read that when you go to the hardware store to buy a drill, what you're really buying is a hole. The drill is the means, the hole is the end.

No matter how fancy the drill is, there will be some people who don't need a hole, and therefore will never buy a drill. That's not a statement of the success or failure of the drill to make holes, and nor will adding further features and enhancements to the drill increase the demand for holes.

So IPv6 is the means, and an IPv6-only application, or an IPv6 application that works much better than an IPv4 one, is the end. It is only when enterprises (and others) want the IPv6 end will they then be prepared to deploy the IPv6 means."



If people want to stimulate IPv6 adoption in enterprises, then it would seem to me that the focus needs to be on developing or demonstrating the value of either moving applications to using IPv6 instead of IPv4, or creating applications that cannot or cannot easily be operated over IPv4. While I don't know if it is used much, one example of such an application is Apple's Back to My Mac service, described in RFC6281.

Regards,
Mark.