There is
clearly an argument to be made that the telecom bill wending its way through the House means trouble for the freewheeling Internet. The biggest concern is that the Internet could be bifurcated into a high-speed "Cadillac" network for those who can pay, and a lower-speed "McDonald's" network for everyone else on the public Internet.
I do feel that it is better to come at an issue like this from a position of knowledge & awareness. What I want to do is illustrate what it appears this bill is trying to accomplish, and the effects it may or may not have on our beloved Web.
I am not sure if there is a reason to be too worried about this bill. At least, the performance bifurcation part of it. I'm offering my take on this as my first effort at a diary in DKos. After reading this, you can feel free to tell me to go pound sand :)~
We might profit from considering the very nature of the Internet before deciding that the Net Neutrality issue is something to get too stirred up about.
I've also worked in the internetworking equipment industry for some time (10+ years) and have written several books on associated topics back in the day, so I feel qualified to comment on this issue.
I will avoid acronyms and technical complexities as much as I can.
There are two primary methods for transmitting information across the Internet. One is oriented towards guaranteeing that all transmission units (called packets) for a particular request or application will get successfully sent from one end to the other. This is called TCP. The other transportation method is strictly a "best effort," take -you-own chances approach that offers no guarantees that any individual packet will reach its destination. (This is called UDP.)
Now, although every Web page on the planet uses the TCP method, it goes another layer deeper. What the telecom bill is really talking about is applying a phenomenon called Qualify of Service (QoS) to specialized types of traffic such as AT&T/Verizon-provided Internet video and voice. What it means for you and me is that if some of us wind getting movies piped to us over a fast Internet connection, those live transmissions will have a better chance of having no flaws or skips.
There are very few if any Web page transmissions that get anything beyond what is called "best-effort" service on the Internet. Google, Amazon, whatever. "Best-effort" is an established QoS term that denotes the lowest service quality in the Internet heirarchy. With this new telecom bill, this would not change one iota.
I'll point something else out. When you click on any link in a Web page or select any bookmark, a new session gets established between you and the server sending the page, using a highly specific TCP exchange that I won't describe here. It usually only takes a moment or two for the desired information to download. While that is taking place, you make certain demands on the network. However, this only happens for a brief time. After your page finishes downloading, that session you established with the server is torn down, again using another highly specific sequence. Most of the time, you're hardly using any bandwidth at all, and what IS used, is occupied by small quantities of traffic between equipment, saying things like "are you still there?" and "Yep. I am."
So you really do not have a "Web connection" as such. What you have is a digital connection that allows you to establish many hundreds or thousands of little, short-lived data transfer sessions between your computer and a distant host, serving those pages up. This will not change, and your level of service will likely remain the same, even if you're Kos himself running the server farm. It is this scheme that makes the Web possible. If everyone ate up scads of bandwidth all the time, the Internet would choke on its own traffic. But it doesn't work that way. Even with so-called tiered service, this would not change anything for you and me.
Voice and high-def video are two applications that are completely different from anything that has ever been attempted on the Internet. Neither type of traffic is really suited for the Internet, because video is highly bandwidth-intensive and latency-sensitive. Video on the Internet particularly is like putting lipstick on a pig, because it really is not designed for the protocol stack that runs the Internet. I've used and written documentation for video-on-demand equipment, and all I can say is it is an incredibly complex kludge even when it's an elegant implementation. The end result is that a video feed to just ONE neighborhood occupies an immutable 30-40 Megabits ALL THE WAY THROUGH THE NETWORK into the neighborhood, where it is split up into individual copies for each house using another protocol suite that is far too complex to describe here (multicasting). There's none of this twiddling back and forth like with all of us viewing Web pages.
OTOH, Voice is very sensitive to latency but doesn't need a lot of speed. Your phone line virtually never drops ANY voice info, does it? That's because while it uses only a tiny piece of bandwidth, it is guaranteed bandwidth. Thus, virtually no latency. Voice on the Internet attempts to reproduce this using a brutally complex protocol suite called VoIP which is far beyond my scope to explain here. (Hell, I barely understand it myself.) Voice on the Internet is basically a toy. A popular one, though.
The thing both voice and video have in common is the need for ironclad service guarantees. This is illustrated by the cable companies, who get enormous grief if their transmission starts producing blocky pixelation effects on their viewers' screens (as well that viewers should complain, given what they charge). However, this doesn't happen very often (however, it sure did in Berkeley when I lived there, where Comcast's service is terrible) because the cable companies have a big advantage. It's a truism that they have their own dedicated network completely bypassing the telecom network. So they were able to start offering Internet access as an afterthought and gradually roll out and improve their Internet service, to the point where it is far superior to what the Bells have to offer. (At least, mine is. I shouldn't generalize from my experience, though.)
The telcos know they need to replicate the same video quality that you get in your cable connection or they're toast, with no way to grow their business. (Supposedly.) Also, think about how the Net is set up right now. It's totally random, right? Anyone who wants a connection even if it's just dial-up, can get one. Then, think about 10-20 million new Internet subscribers all pounding down those spanking new 40 Mbps connections for Internet video service. The telcos have no choice but to implement a massive expansion of the Internet's capacity, from the core to the edge, to support even the smallest expansion of their customer base. The root of all this is that telcos want to make sure that Google and Yahoo don't eat up all the bandwidth (from all those new routers and switches they're gonna buy from Cisco and others). Which, for the technical reasons I've outlined, is frankly NEVER gonna happen.
Ed Whitacre's bellyaching about Web companies in his core network is a straw man. He has no idea what he's talking about.
But what WILL happen is that a lot of new equipment is gonna get bought and some jobs will result. (We need this, you know. Good paying jobs are hard to come by in this country.) The capacity of the Internet will expand. And while there may be some inequities as a result of the so-called "tiered service," it won't be as bad as you think. Insensibly, Internet service will probably improve for everyone and more people in the boonies and in poor urban neighborhoods may get better basic Internet service too. It will take awhile. But in this instant-gratification society of our, people don't want to wait for anything. If we do wind up with 10-20 million well-to-do Americans getting 40 Mbps Internet service with fancy-schmancy VoD, so what? Read a frickin' book! I know I won't be getting VoD anytime soon, and frankly I don't need it. There's an economic phenomenon called "substitution." I already have a product that's good enough - cable. (although they could stand to have some competition, given their pricing power.) Do I need 40 Mbps? Yeah, its nice-to-have, but not really.
The fact is that the telcos don't know very much about IP networks. Oh, they think they know every freakin' thing, and they'll tell you so if they get a chance. But they don't. The telecom industry is full of Republican bellheads who grew up configuring stupid PBXs and think every network should be a sickeningly complex web of point-to-point cross-connects or it just isn't worth a sh*t. I hated working for a telecom equipment maker and am glad to be back in the IP world, where sanity generally reigns and it's reasonably safe to be a Democrat. The telecom companies do not understand IP protocols. This makes me feel a bit safer than I otherwise would. And Congress? Good God, man. They can't find their asses with both hands and a map. Hell, neither can the telcos.
So let 'em pass their stupid bill. The genie's out of the bottle and he ain't going back in.
So buck up. It's not as bad as it sounds. Kos is not going to have his packets censored by big bad AT&T. I don't even think most of those guys know what a freakin' packet is. Maybe, if we're lucky, after the telcos take 20 years to roll out their stupid VoD service to three cities in the US, the state-of-the-art will have gone so far ahead of them the only thing they'll be able to do in the end is accept the fact that everyone's packets will go through their dramatically expanded pipes because that's the only traffic there is.