4k could end up getting skipped as a broadcast standard given how quickly 8k is growing up. It will be a lot easier to dedicate that much bandwidth to a single channel once we eliminate those which serve no useful purpose in the linear domain (those which carry only pre-recorded shows).

What makes a 4:3 ratio so much better than a 16:9 ratio for your monitor?

This is actually much more than 2x standard broadcasts in terms of resolution as far as the US market is concerned. In the US almost everything "HD" is 720 @ 60fps (or sometimes even 30!). This is 8k, which is 16x the resolution of 1080, and twice the frame rate.

Sounds like the issue should be with Apple then, since they developed an app and provided it to folks to pre-install with the option to disable it based on the network configuration. They gave AT&T that option, and now AT&T is changing the way that they implement it on their network. Presumably Apple could have done things differently, but instead decided to do things this way. Why? No idea, but I can come up with various guesses. Maybe Apple and AT&T struck some kind of deal where in exchange for additional promotional consideration, Apple elected to limit the use of their new app to WiFi rather than AT&Ts overburdened network?

I'm not sure I understand your beef here. It seems as though the argument he lays out is that AT&T does not have a dog in the fight when it comes to what apps folks use for video chat over the 3G network. As a network operator they were initially concerned with the burden that those apps would place on the network so they did not allow them to operate. The situation has now changed, and they are willing to start opening it up to certain classes of users that would present a more moderate load than the class of customers with unlimited data plans. I'm not sure how this violates the FCC net neutrality rules, perhaps you could enlighten me.

This post really makes a lot of sense. I think the most under-appreciated factor in your list is

• a cost per byte that's not higher than what people currently pay for the unlimited bytes they actually use

A cursory inspection reveals that the new share plans come with unlimited minutes and text messages. These may not appear competitive relative to the cost of the grandfathered data plans, particularly for those heavy data users out there. But to the average user, this could be a significant net savings to switch out of those holdover contracts due to the discounts you would receive on the other aspects of your plan. It stands to reason that a good number of those folks out there with unlimited plans for a single device may not be using enough data to actually be saving any money on the byte. They could in fact be paying a serious premium for a class of service that is not actually delivering any additional value for them.

How is that obvious?

T-Mobile is having trouble retaining / gaining subscribers. I doubt this is altruistic, they need to draw more customers in so they are attempting deep discounting.

No they are saying that customers can pay more for more features / services. In this case, a bandwidth sharing plan that fits the average consumers usage patterns better, and the ability to make FaceTime calls over 3G which they don't have today. So, actually more for more. Maybe not enough more to compel you to go out and change your plan, maybe you value something you already have more highly (like one of those grandfathered unlimited data plans, for example) but for others this can and probably does look like a good deal.

Don't forget, AT&T is still giving their customers more tomorrow than they have today. Today they have FaceTime over WIFi, and a variety of other apps that do the same thing. Going forward, customers will have access to not only that, but the same great app over 3G service if they have purchased one of the new service plans. This gives folks an incentive to change to the new plan if they are on an older single device data plan that could very well have a more conservative cap than they would have with the share plan, and they are getting additional value added services. Getting more tomorrow than I have today sure sounds like a good deal to me.

"contractors" in a corporate context don't usually mean 1099 contractors but employees of another firm that you have a contract with. So, you can bring someone in who works for a temp agency (their paycheck and benefits are cut by the temp agency) and then your firm pays the agency an hourly rate for the temp. They are "employees" in the IRS context, but called "contractors" because they are an external contractor's employees. Confusing I know but probably not actually an issue in this case.

The same is true for most cable system video on-demand services as well.

Even once they are all IP they should probably not all be treated "neutrally." For example, IP multicast video distribution is a heck of a lot more efficient than unicast when we're talking about a large number of people on the same network consuming the same content (like live linear TV). Some traffic is much more sensitive to latency and jitter, (like VoIP traffic) and should probably not be treated equally either. Another example would be content caching and distribution. It's a lot better for the network if content producers cache content close to the edge, maybe even in the nodes, particularly stuff that has to be delivered in a unicast context.

Here is a link to a quora post explaining how it works. It's delivered over a QAM channel, not IP, for the last mile in most cable systems. So, as I was saying, not actually an issue under net neutrality.

Most (virtually all) traditional cable system VOD is not delivered over IP - it is an entirely different network and it does not compete with your internet bandwidth.