In computer science we have always learned of ways of transitioning from an old system to a knew system and these ways are designed to allow for seamless change over without disrupting normal business usage. Some of the ways include running two systems in parallel, gradual change over various system components, backward compatibility and so on and so forth.
For software/hardware changes backward compatibility has, been the favourite. This works better if the newer version of software/hardware has made major changes.
IPv6 has been designed without giving room for backward compatibility. This I think is very un Computer Science like.
There are these problems of so much hardware being dumped prematurely for not being compatible with the new IP addressing system. The ways available for cooperating ipv4 and ipv6 require careful study of the system and skilled expertise to make the change. But does this change stay permanently? Until the company can buy new equipment.
I thought the problem of IP addresses being exhausted is because we are using from the same pool world over. There used to be classes of IP addresses which have now being discarded to allow for usability of classless ips to solve the IP address shortage.
Looking at the local area network design and IP addressing, it is possible to use the 192.168.0.0 network in one organization and the same address block in another. This is a range of IP addresses.
Suppose we take the whole ipv4 range and privatize it at ISP level, then we have a big range that one isp can't exhaust(or not atleast for the time being) for no isp provides to customers of multitude the whole current ipv4 users world wide. If then the ISPs can make the changes to make ipv4 compatible with ipv6, the customers whose use ips just for communication for the purpose of their daily business wouldn't even have a worry of what ipv6 is about.
The only businesses I see being affected by the change are those who provide internet services. And not many businesses do so. If they do, it is usually through dedicated internet service providers. Very few companies indeed host their own internet services.
It therefore remains the ISPs problem.
Further more if the ipv4 pool can be privatised at isp level, the isps have chance of selling their old incompatible hardware to their customers, in favour of hardware which makes using ipv4 and ipv6 easier. It is easier to manufacture compatibility hardware than to configure incompatible hardware to be compatible with ipv6.
How about putting a Nat-device between your hardware and isp instead of rebuilding the whole network on ipv6? Let the Nat-device worry about compatibility not you. Are there such hardware yet? I would love to make one.