That's the thing though, you can't effectively measure data quantity caps to bandwidth requirements. You could cap all your users to 5 GB a month, but if all your users burst their 5 GB in the same 2 hour period, you're hosed.
This is the paradox with this vod-ip.
If it's made too attractive, it will hose the internet. Let's hope that old people still connects their tv's to the cable for a long time instead of ethernet.
So the "download" caps were/are just one big scam. For the longest time, we didn't even charge them to half of the customers because our monitoring software took a big dump on the old infrastructure after a router software upgrade from Motorola and a bug in the SNMP stack. It would make much more sense for ISPs to throttle bandwidth than data quantities in peak hours. Some do actually, on top of the caps. Download on-going for more than 1 minute at X bps ? QoS that mofo down at the last-mile router.
Aren't "caps" usually just throttling? I've read lot of news about ISP's "prioritizing" rates after certain amount is reached. At least in EU, ISPs have hard time trying to introduce hard caps even with pirating, since internet is being seen as "civil right" and consumer rights are good. Eg. if ISP doesn't provide easy way for consumer to monitor the amount of traffic, there can be no immediate actions when caps are reached.
But ISPs are getting their traffic based monitoring and charging systems working and they will be using them as the difference in consumers usage widens. Already it's like 1% of customers are using 99% of bandwidth. When it's 0.1% using 99.9% and their usage is 1000 times more than now, ISPs have to act. Otherwise average user's cost would be way too high for consumer to accept.
Then comes the issue what is fair cap/throttle for consumers?
Can you "save" your bandwidth to have it more in peak hours?
Then the system would be based on amount of traffic and isn't the situation this right now?
How could you cap the bandwidth by only looking present time usage?
Wouldn't that be same cap for everybody regardless of your usage and so pretty unfair? Wild wild west?
Isn't "download on-going for more than 1 minute at X bps" throttling more based on traffic than bandwith (time*bps)?
About QoS: usually ISPs offer 3 kind of pipes: dedicated (defined endpoints), QoS (X bps, x% of time) and bulk. From what I've heard there's at least 10x price difference from one to another. If ISP want's to give consumer a nice ip-vod experience from outside of IPS's own network, the need for QoS quality bandwidth is increasing and therefore costs for ISP for same amount of bandwidth are also increasing.
Of course ISPs that are also cable operators try not to give benefit for some other content providers. But they know they are fighting against windmills. But instead giving cheap pipe to Apple, they develop their own vod system.
In near future ISPs don't have their own vod service still working, so they might not make Apple hard time delivering iTunes movies.
So for Apple to get iTunes movies successful, they need to make very good deals with ISPs.
In the medium time scale, ISPs try to compete with iTunes movies and this will be hard time for Apple if they don't became ISP themselves.
For long time scale, pipes are getting so fast, that bandwidth isn't an issue any more.
I must be lucky then, I again watched my now 11 year old Macross Plus DVDs just last weekend. Flawless as ever. Maybe I should try out my first ever DVD purchase, Costner's Robin Hood : Prince of Thieves. It too is 11 years old...
Then again, my Ride the Lightning CD still plays fine and it was purchased in the early 90s...
IIRC, most problems when optical storage was new in mainstream were with corrosive inks on label side.
I guess that single layer gold dvd-r burned with slowest speed is the most reliable today. I'm not sure which is more problematic: cd's have data very near of label side or dvd's having 2 layers that can separate. But anyway, it's one the most convenient to freshen the archive, since the physical sizes of different optical storage generations are the same.
15 years from now, you can take that 1000 disc archive, have couple of normal basic computers, choose the most reliable optical media at that time and freshen your archive in a week.
Usually this is much more complicated with other media. In 15 years it might be hard to find even IDE-controllers, like now is not easy to copy microchannel hdd to newer one. Or dd floppy in mac format.