Jdevil3 said:
Oh, I see... thanks for clearing that up ^_^
So we have some more time left to fix this... but reading this thread makes me think that whatever time we have left before that happens is not enough to fix it... maybe I'm getting things wrong? O_o
|
the main problem will be that new isps will have an extremely difficult time getting blocks of ip space to allocate addresses to their customers. Larger isps will have the ip space and ability to extend this somewhat (maybe like 1-2 years at the absolute extreme edge of possibility) but as soleron mentioned it is already far, FAR too late to prevent at least a moderate-scale problem from developing though we really don't have any idea how this whole situation is going to turn out.
Personally (and a caveat here, probably soleron is the most knowledgeable person here as I don't know a great deal but bear with me..) I think the most realistic scenario to prevent a large problem from occuring would be for the major telecom companies to install ipv6 equipment for their commercial networks and provide subsidized ipv6 equipment to their customers' access points so they could talk to the servers. (maybe supported by some kind of government program like the one the US did for the digital tv converter boxes?) Then the companies could simply (and I realize it isn't actually simple, but yeah..) perform the necessary ip routing at their main data centers to allow their customers to perform ipv4 and ipv6 requests that have to be forwarded to another network.
Again I am not an expert on this so this may not actually work. The major problem I CAN foresee would be figuring out how to allow network devices that have embedded non-flashable ROM or that lack the necessary storage and/or communication ability to work with 128-bit ipv6 network addresses (vs the 32-bit ipv4 address) to still be able request an ipv4 address and then have that request translated to a valid ipv6 address to be forwarded to the isp. To me, this is the biggest problem with ipv6 --> namely that it was not designed to be easily backwards compatible with ipv4 hardware and while I can see why they might want to start over with a clean slate they should have been more realistic when they were designing it. To restate what soleron said, we really should ahve started thinking about this 5 years ago (or more).
Although the previous problem would be quite simply a pain in the ass, it still pales in comparison to the problem of convincing telecom companies and/or governments to invest the necessary millions (billions?) of dollars that would be needed to successfully transition to ipv6 without basically shutting down the entire internet for like 6-12 months and THEN spending it.... Because not having enough ip addresses on the internet will not really affect a telecom company and upgrading to ipv6 would not help them personally, all isps across the globe have essentially decided that it is the other guy's responsibility to spend the dough to upgrade. This obviously results in the situation we have now, where nobody is ready and ip address continue to trickle away... real great isn't it?
oh and btw, the problem with y2k wasn't that computers would think that 00 meant 1900 or 2000 because they could just have easily thought it was 1898 or 1998. the real problem was that a lot of programs (and I mean a LOT) involve some sort of arithmetic operation involving the system date. Why is this a problem you ask? well, a program will happly add, subtract, divide etc on 89, 99, 1293847 or whatever but when you start dividing by zero you can watch your program crash and burn... real fun when that program is running your respirator or dialysis machine. Though unless those programs were running Windows it probably wouldn't have mattered anyway because they would have used unix time - which means your program will crash at 03:14:08 on January 19, 2038....