connected...

Who Cares About The iPhone?

I've had a Motorola Droid 2 for several weeks now and I am still gushing over it, marking probably the longest love affair I've ever had with a gadget.  For probably close to a decade I'd been telling my corporate and college students that I wanted a universal handheld device that was a phone, GPS and MP3 player.  After a few products that came close to my ideal (the Q, which drove me away from Motorola for years not just because Verizon disabled the GPS, and BlackBerry), my grand dream is finally realized.  And it ain't an iPhone, which even as a longtime Apple and iPod fan makes me glad.

Yes, iPhones broke new ground and are still wicked cool.  And you can jailbreak them so you aren't trapped in a room in hell with Inez, Sartre and Steve Jobs' controlling vision.  But competition is good, and I couldn't wait for Verizon to get them.

My old phone was toast, and the Android devices were mature enough so I took the plunge.  Never made a better tech decision in my life.

There's a wide array of apps, and most of the apps I want have been/will be ported from iPhone versions.  The biggest app gap forme so far: no CSPAN radio from Jacobs Media yet, and no NASA TV app.  Yeah, I'm a geek.  In that vein, my favorite apps are the Android version of APOD and Galaxy Wire, which provides cool vintage NASA pics daily and This Day in History stuff.

The phone is also incredibly stable.  Not that it doesn't freak out on occasion, but what computer doesn't, especially when you load boatloads of software and hammer it constantly with use?

Battery life is a POS, so I have found myself tethered a bit more than I'd like, but management apps and a little behavior mod makes it bearable.  The overall utility of this thing more than makes up for power weakness.

Ericka wants an iPhone.  Maybe.  We're due for a New Every Two upgrade and she'd been chomping at the bit for one.  Now that it's not just vaporware here in Verizon Wireless Land, she's less passionate about it since she's seen how great my Droid is.  We'll see what happens next month.

ntodd

PS--I was going to blog more about network issues and whatnot, but lost interest.  Gotta get back to Angry Birds...

The Rural Broadbandification Act Of 2010?

As of yesterday, every Finn has the legal right to 1Mbps broadband connections to the Internet.  Bully for them!

I've long wondered why the United States, originator of the Internet, early adopter of e-commerce, in the top tier of Internet usage and e-government development, never aggressively invested in broadband deployment.  It would fit our tradition of universal service going back to the Rural Electrification Act and Communications Act of 1934 and even the Constitutional mandate to establish post offices and postal roads.

As it stands, we're still on dialup on Bog Road (as are 20% of our fellow Vermonters).  Yes, there is a "high speed" satellite option, but the bandwidth is paltry compared to DSL and cable technologies, extremely expensive in absolute and relative terms, and customer service is arguably worse than even your most horrific Comcast stories.  So if I need to upload any large files to my e-learning sites, for example, I have to run out to someplace like Cosmic Cafe in St Albans where I can grab some free WiFi.

Vermont's coverage is actually surprisingly good, but there are still scores of communities that have little or no access to highspeed connectivity.  Our fortunes do seem to be changing as President Obama announced today almost a billion dollars in grants to foster greater broadband deployment.  Vermont nabbed $45M of that:

Vermont Telecommunications Authority estimates that it will directly create hundreds of jobs with its project to build a 790-mile fiber network across Vermont with it's $33.4 million grant. Over 200,000 people and thousands of businesses and community institutions stand to benefit from increasing access to direct high-speed connections. In addition to the jobs this project creates upfront, it will help drive economic development in the community that creates jobs for years to come.

Vermont Telephone Company's project, assisted by $12.3 million in federal funds, will enhance the existing middle mile broadband infrastructure to address bandwidth and transport capacity shortages. The project stands to benefit over 150,000 people and over 15,000 businesses and community institutions. Not only will this project create jobs upfront, but it will help drive economic development in the community that will create jobs for years to come.

350k people?  Hopefully we'll get a piece of that action in Fletcher, Fairfield and St Albans Town.  After years of promises not met by the State and companies like Fairpoint, I am guardedly optimistic.

todd

Come To Vermont And Enjoy Our Foliage And Fiber

Freep:

Using Burlington Telecom's municipal broadband network as a model, 22 rural Vermont towns are poised to pool their resources and launch a fiber-optic project that could go online by the end of 2009.

It's more than a twinkle in a technician's eye. Members of the East Central Vermont Community Fiber Network announced Wednesday that formal agreements are in the works from Windsor to Montpelier that would bring the strength of numbers -- and attractive financing -- to universal broadband Internet coverage.

The group's leadership presented the project at a board meeting of the Vermont Telecommunications Authority held Wednesday at the Zampieri Office Building in Burlington.

The presenters made no funding requests but asked the state board for support with credit and regulatory hurdles.

Steve Willbanks, chairman of the Strafford Selectboard and a key player in the emerging network, said commercial broadband providers could not meet the needs of rural Vermonters.

"These fiber-optic connections are absolute necessities; not luxuries," he said. "We need them for our economical and cultural development. We've had seasonal residents tell us they'd move here in a heartbeat; they'd telecommute if they had access to broadband.

"There are people out there who would kill -- almost -- to have reliable broadband service."

More than 1,000 residents in his area have registered for service, Willbanks said. About half of the population targeted by the East Central Vermont Community Fiber Network have no broadband service.

Tim Nulty of Jericho, who left Burlington Telecom in the fall to join the East Central group as chief consultant, said Burlington Telecom's business model would work well in a confederation of smaller, more thinly populated towns.

Nulty said low-interest loans would allow the $70 million, subscriber-funded network to achieve a positive cash flow in four to six years.

Communities, not technology, would shape the size of rural networks, Nulty said.

"Localized service is a huge, competitive advantage," he said. "You're calling on your neighbors for service, and you're serving a common interest. All we really need is a critical mass of about 25,000 -- or about 6,000 paying customers."

Long ago the Friday Coffeeblogging team chatted with Tim Nulty when he was still at Burlington Telecom.  He's a good guy with a comprehensive vision, so this bodes well for rural fiber access in Vermont.

It would be nice if Gov Douglas would put together a more cohesive strategy for getting broadband deployed statewide.  These community coalitions are great, though the patchwork approach doesn't do universal good.  The technological solutions will still likely be patchwork (fiber and WiMax depending on geography and geology), but we really need money and coordination to pull it all together.

In related news my neighbor, the "mayor" of Bald Hill, has been poking around for broadband solutions up by The Fortress of Solitude.  When I was up there around the new year, she and I were talking about possibly getting set up to share satellite access but I observed that we already have a nice tower on our mountain and perhaps we could put a repeater up there to grab local wireless service (I'd actually written to the Governor about that and he wrote back with referrals to people in his administration).  She's been doing some legwork and we just might be able to pull that off.  Fingers crossed.

ntodd

Like Wireless In A Younger Day

So I headed up to St Albans to enjoy some lunch and free WiFi since my favorite cafe in Milton has unexpectedly closed down.  I was looking for The Flying Disk, but it turns out it's actually in Enosburg and the listing at the WiFi directory I used was juuuuuuust a little off (I've submitted errata).  Since I was in town anyway, I stopped into the Park Cafe a block up the road and ate there whilst grabbing signal off a nearby open access point.

I really want a replacement for The Alley because I'm stuck on dialup at home and the only other access near my house appears to be at the McDonald's in Milton, which isn't quite the atmosphere I want to hang out in.  I have a couple of spots I like to go in Burlington, but that's about 45 minutes door-to-door and significant gas consumption.  I was psyched about The Flying Disk because it would be about 20 minutes away, so cuts down the inconvenience and fuel factors.

This is the personal impact of the United States' crappy broadband situation (which we can extend to a global impact if you consider the additional driving involved).  Of course, I live in a rural area in our most rural state so it's generally worse than in the rest of the country, which is why I support efforts like Jim Douglas' broadband initiative.

WiMax-based state projects make the most sense for places like Vermont--heck, Great Auk Wireless could pop a repeater up on the fire tower that's already on top of the mountain by my camp in the Northeast Kingdom--and probably won't run into the same anti-democratic, profit-mongering resistance and lobbying from the telcos like municipal WiFi has.  Given each state has its own unique issues in terms of infrastructure, demographics, etc, it makes a bit of sense for them to be heading the effort to improve broadband access.

However, I'd like to see Federal backing for what is fundamentally a national challenge to keep the US competitive.  I think building upon the old notion of Universal Service would be an important part of that, but nobody ever listens to me so I'll probably be limited to getting excited about 100k over my wireless phone forever...

ntodd

Universal Access

Something I do agree with Google about:

Google wants to put its money where its mouth is...to the tune of at least $4.6 billion.

The search giant said Friday that it's willing to participate in the Federal Communications Commission's upcoming wireless spectrum 700MHz band auction and pay the minimum reserve of $4.6 billion.

But for Google to participate, there's a catch. It's requiring the FCC to adopt all four of its licensing recommendations, no matter who ultimately wins the bidding process.

Google has also been pressing the FCC to reserve a portion of that spectrum to be used primarily, or exclusively, for broadband communications. The Internet giant wants "open access" rules, fearing that without them, only a few powerful companies, such as telecommunications and cable providers, will become the permanent guardians of the airwaves.

Google wants the FCC to agree to its four "open" platform recommendations, which include: open applications for users; open devices that will work with whichever network provider customers choose; open services that would allow for third-party resellers to acquire wireless services on a wholesale basis; and open networks, which would allow third parties, such as Internet service providers, to interconnect at any feasible point within the 700MHz licensee's wireless network.

The Issue Of Which We No Longer Speak really is a discussion of TRANSPORT across provider backbones (yes, getting to them is necessary, but it's about categorizing data for appropriate delivery in the carrier cloud).  I think people should be more concerned about ACCESS--that's where the overblown rhetoric about "Internet gatekeepers" is more apt.

I'd like to see us expand the Universal Service concept to become Universal Access, meaning I can get to the Internet and my applications no matter where I am, no matter what device I use, and no matter what provider's service I subscribe to.  This would probably require all players, including the Googles, Comcasts and Vonages of the world, to pay into the USF.  A good thing for all, methinks.

ntodd

[Adding: I like this play by Google from a strategic business POV, too.  While you can be an app provider without owning the network, if you've got the coin it certainly makes sense to invest in access.  They've also been fairly aggressive in WiFi investment, IIRC.

Also note that the reason I want the FCC to carve out some spectrum for open access is that the airwaves belong to all of us.  The reason we need to auction is to make sure we don't step on each others' toes (having no regs governing wireless would be as bad as having no traffic laws on the road), generating revenue for the government is a good thing for everybody, and it makes sense to have providers with their economies of scale essentially manage the spectrum on our behalf.  But they shouldn't be able to mint unlimited money just because we lease them some of our RF property.  Not unlike public access TV, I say if the providers want access to us, the customers, they gotta give us access to stuff that belongs to us.]

I Will Speak No More Forever: My Last Words On NN.

I'm serious.  After this post, I'm done with Net Neutrality.  Nothing else I can say will convince you if analogies, technical discussions, observations about regulatory regimes and economics, etc, failed to do so.  If you don't accept my arguments, fine. I just hope that readers understand where I'm coming from (and hope the FCC, FTC and Congress continue holding the line).

The only reason I bring it up one last time is I read a couple posts over at CNET NewsBlog that spurred some final thoughts.  First, this one from Tuesday:

The Net neutrality skirmish that swallowed up so much of Congress' technopolitical agenda last year may be gearing up for a comeback. A pair of senators who led the divisive push for the new regulations want everyone to know they haven't forgotten the cause.

Sens. Byron Dorgan (D-N.D.) and Olympia Snowe (R-Maine) aired their views in a joint letter (PDF) filed with the Federal Communications Commission just before the Monday deadline for remarks on an open inquiry into "broadband industry practices."

The senators said they were pleased that the FCC was showing interest in the issue but "would have preferred the commission take the more concrete step of proposing rules to guarantee Internet freedom."

Internet freedom, in the senators' view, is the idea that a broadband operator like Comcast or AT&T should be legally prohibited from charging, say, YouTube extra fees to have its services prioritized over other online video sites. In recent years, cable and telephone companies have said it may be necessary to pursue such a business model to recover investments in new infrastructure, and they don't want regulators dictating how they manage their pipes.

Now, I've been accused of focusing too much on the technical aspects of QoS and such, thus ignoring business realities and the like.  I take issue with that--I have learned a teensy bit about coupla non-techy things whilst teaching a variety of audiences in the industry the past 15 years--and simply note that I always begin with a technology discussion because I think legislators and other folks out there are misunderstanding what words like "prioritize" really mean in this context.

I covered Quality of Service in my epic post a while back and simply observe that prioritization is not about giving a person a chance to simply cut in line, but rather about giving different people in an aggregate sense a service that meets their needs.  Not unlike having an express checkout at the supermarket: it's not unfair to give a line just for 10 items or less because those people are presumably not stocking up and just getting a few quick things on the way home or whatever, so shouldn't have to wait in line behind the person buying all the beer and ice cream for the month.

Here's where we start getting inflammatory words like "discrimination" when really what we're talking about classifying data so we know which line to put it in.  Understandably, when people here these words thrown around, they fear that YouTube will be able to buy themselves a certain advantage over their competition and thus destroy the fabric of the Internet.  Well, yes, they can pay for preferential delivery of their high-quality video service, and so could Yahoo or Joe Schlobotnik's Videos R Us.  Then they all would be treated the same in terms of queuing, delivery paths and so on.

Yet non-paying content providers would still have their stuff delivered in the same manner as it is today in the "best effort" Internet.  Consider another analogy: airlines generally preboard certain customers before take off.  Why?  Because some people need wheelchair assistance, or have a few unruly kids who need more time to settle in, etc.  Letting them board first gives them "priority," but the rest of us still board eventually and we all get to our destination.  However, changing who boards when just a little really makes the entire boarding process go more smoothly, thus improving the chances of an on-time departure and arrival, which is really all about improving service to everybody.

So in this sense, "prioritization" is inherently a good, not a dangerous, thing.  Providers must be able to offer this technically to deliver multiple applications (flying customers) over a single infrastructure (the plane, not a truck, right Ted?).  Therefore the FCC and FTC do not want to regulate this aspect of things: that's getting into business practices and micromanaging operations, which is not where we want ANY government entity to go so long as general fairness reigns.

The next issue is whether a provider can charge for such premium offerings.  Of course the airline analogy above breaks down here, because nobody charges for pre-boarding per se.  Although if you're a frequent flier and are part of affinity programs, you are in a sense being charged for status in buying more tickets or paying extra for first class.  Anyway, I think my previous analogies of first class versus overnight shipping is more apt here.

If you accept that technically a network must be able to categorize and prioritize, then the question is simply whether you believe in the general capitalist principle that people pay more for more service?  Or to put it another way, if you're really into Net Neutrality it is truest sense, I have a right to choose Low Rent Video service or pay extra to get Super High Def Video (of course, I'd rather get that for free, but I'm a realist).

Speaking of charing, at one point I was also told that I shouldn't conflate network evolution with specific business models.  Indeed, I do not, yet that's precisely what the proposed NN legislation is all about.  Telling providers they can't charge for specific services.  Even in the Bad Old Tariff days, I would not have wanted Congress to come in and determine just how a service is technically deployed and billed for.  Whose demesne was that?  The FCC's and local utilities commissions'.  Congress cannot adapt to changing realities quickly--just look at how long we've gone between communications overhaul legislation. 

We don't want Ted Stevens tinkering with any of this stuff.  That's why I was so pleased with the government's traditional hands-off approach to the Internet in terms of taxation and the like.  Let it flourish and innovate without heavy regulation beyond general strategic principles that acknowledge the power and benefits of the Internet.

 

This is why the FCC came up with its broadband connectivity principles in 2005, to wit:

1. consumers are entitled to access the lawful Internet content of their choice
2. consumers are entitled to run applications and services of their choice, subject to the needs of law enforcement
3. consumers are entitled to connect their choice of legal devices that do not harm the network
4. consumers are entitled to competition among network providers, application and service providers, and content providers

This is, in effect, what the FCC already has enforced when providers have broken the rules (e.g., Madison River).  They did also announce an inquiry to get public comment which will also inform their posture, but really this is all they need and thus the Commission (along with the FTC) have decided they don't need any more Congressional authority to preserve Internet openness.

While I think it would be redundant, I can understand that maybe people want the specific force of law to back up these principles, and I'd be totally fine with this being put into a Net Neutrality amendment.  They are reasonable, limited and precisely reflect what the Internet has always been and always should be.

And now the worst reason ever why somebody should be against NN:

Hollywood hasn't decided what it thinks about the whole "Net neutrality" debate, but it knows one thing: Any rules that would stunt roll-out of the next new whiz-bang filtering technologies or encourage unfettered sharing of copyrighted works over peer-to-peer networks would be very, very bad.

That's the gist of the 9-page comments (PDF) that the Motion Picture Association of America filed with the Federal Communications Commission this week. Monday was the deadline for comments for an FCC inquiry into "broadband industry practices," and most of the some 27,000 filings focused on the thorny issue of Net neutrality--that is, whether the FCC should impose a new policy prohibiting network operators from making special deals with content providers for priority service.

Whatever the FCC decides on that front, the MPAA said it's "crucial" not to take any steps that would inhibit the development and rollout of new technologies like watermarking, deep packet inspection, acoustic fingerprinting and other forms of filtering--all of which are still in the early stages and are far from perfect.

"Any policy efforts relating to Net Neutrality must promote the protection of intellectual property," MPAA lawyer's wrote. "It is crucial that FCC policies not interfere with the efforts of broadband companies and content providers of all kinds to solve problems of free riding."

A prime example of such "free riding," the MPAA said, is the "unauthorized" swapping of massive amounts of copyrighted materials via peer-to-peer networks. The group argued such activities pose a triple threat: to the copyrighted content creators, to the consumers who actually pay for the works and to the "light" Internet users who may be indirectly forced to pay for the high bandwidth use of those file swappers.

"It would be Pyrrhic indeed to adopt a set of principles asserting that consumers have a right to a cornucopia of excellent content, but fail to provide an environment in which such content can actually exist," the movie industry concluded.

This filing, which was noted Wednesday night by Ars Technica, isn't first to evoke copyright issues in this proceeding. The general counsel of NBC-Universal has already suggested the FCC should require broadband providers to be more proactive about filtering copyrighted content that traverses their networks. About a dozen public-interest and consumer advocacy groups hit back at those comments this week, arguing such a proposal is not only technologically impractical but could also threaten fair use and free expression rights.

The MPAA and NBC are taking a position is actually directly contradictory to NN from where I sit.  Adding regulatory regimes to police content is anathema to the Internet's principles of neutrality and openness,  goes beyond what network providers want to do in terms of classifying data, and really violates the broad principles outlined by the FCC.

Really, the MPAA is trying to invent an issue, just as Google is--I suspect the former will be less successful in rallying the Nutroot Hordes than the latter...

ntodd

My Epic Post On Biofuels

Okay, this isn't really about biofuels.  It's about a major burr in my saddle: Net Neutrality.

Before I get into the technical, regulatory and economic meat, I must needs begin with a variety of analogies I've pulled out a number of times when people start acting all crazy about NN:

• When you order a supreme pizza, you get charged more than if you ask for a plain cheese.  Pizza parlors often have other options such as ala carte toppings, individual slices and even salad!  All available if you choose to avail yourself, all available for a price.

• Some people buy new Cadillacs with all the spiffy options (including crappy gas mileage), yet I see other cars on the road like 1985 Datsuns, 2007 Priuses, motorcyles, buses, and other vehicles.  In many urban areas there are toll roads available if you want to go faster or more directly to certain surrounding areas that are not limited to Caddies, but for any car no matter how expensive or cheap or gas guzzling or biofueling it is.

• Say you go the Post Office to send a letter.  You have a number of choices for delivery: if the letter is not time-sensitive and/or you're a cheapskate, you could just mail it First Class for 41 cents; should the letter be rather urgent, you might buy an Express Mail flat rate envelope for $16.25; add a larger item to the letter and it will cost you more because of the additional weight.  First Class has always been "best effort" but generally arrives at some point, whereas the other higher-level options have guarantees associated with them.

As I tell my students all the time, every analogy breaks down at some point.  People can't start comparing the specific realities of pizza making, for example, with telecom because of course they ain't exactly the same.  Yet if you accept the basic premise that one may charge more for increased levels of service, then you're halfway there on Net Neutrality.  If you don't accept that, well, that's beyond the scope of this discussion anyway--I'm not going to argue about basic assumptions of capitalism.

Now, I'm going to try to do this without the benefit of a week-long class that I teach on basic telecommunications, so some of the things will be grossly oversimplified (my colleagues and other bit-weenies better cut me some slack!).  I hope the general concepts will be clear despite that.

So what the heck is NN, anyway?  I find most people I argue with have a vague notion about it, not unlike the old line about porn, I know it when I see it.  Fortunately Wikipedia has a good breakdown of the three basic flavors:

Absolute Non-Discrimination: Columbia Law School professor Tim Wu: "Network neutrality is best defined as a network design principle. The idea is that a maximally useful public information network aspires to treat all content, sites, and platforms equally."^[2]

Cardozo Law School professor Susan Crawford: insists that a neutral Internet must forward packets on a first-come, first served basis, without regard for Quality of Service considerations.

Limited Discrimination without QoS Tiering: American lawmakers have introduced bills that would allow Quality of Service discrimination as long as no special fee is charged for higher-quality service.^[10]

Limited Discrimination and Tiering: this approach allows higher fees for QoS as long as there is no exclusivity in service contracts: If I pay to connect to the net with a given quality of service, and you pay to connect to the net with the same or higher quality of service, then you and I can communicate across the net, with that quality of service.^[1]. [We] each pay to connect to the Net, but no one can pay for exclusive access to me.^[11]"See Sir Tim Berners-Lee.

I'll address these in a bit.  First, I need to talk about what QoS is since it's featured so prominently in each definition.  An explanation from my old company's wiki:

Quality of service (QoS) is a network's ability to support varying levels of network performance that can then be mapped to the needs of the applications supported by that network. The performance parameters we seek to control include such things as delay across the network, variations in delay, and total bandwidth available for a connection or information flow, to name a few.

Basically what we're trying to do is use a single networking infrastructure to deliver different applications that have rather different requirements in terms of service from that network.  For example, voice hates delay.  Imagine such a conversation:

Person 1: I love you.

PAUSE

Person 2: I love you, too.

That delay imparts meaning in a usual voice interaction--were you not listening to me, did you have to think before saying you love me, etc?  Thus we want a network to deliver our voices without adding any delay (nor delay variation) because that could in effect change what we are trying to communicate.

On the flip side, consider an e-mail that says, "Come get bagels on the 3rd floor."  Whether that takes 10 seconds or 2 minutes to deliver, the message will always retain its intended meaning.  That's not to say if there's a lot of delay the bagels will still be there after your officemates devour them, but that ain't the issue.  We're simply talking about the needs of e-mail and voice in terms of retaining their meaning.

So each of those applications have certain expectations and requirements of the service provided by a network.  Voice wants low delay, e-mail doesn't care about delay.  They also have different needs regarding other aspects of the network, such as errors--voice is rather tolerant of errors (the human ear can figure out meaning even if parts of the "message" are lost), whereas data needs to be 100% correct (it would be bad if you sent a note promising to transfer 100 bucks and it became 1000).

In the past, we've had separate networks for different applications.  There was a voice network, for example, and a variety of different data networks basically tuned to the requirements of each specific type of application, and ne'er did they meet.  Yet now we're using the public Internet and Internet-like private networks to run multiple applications together.

What providers are doing in the enterprise space is deploying IP networks with some extra capabilities that are not inherent to IP.  The Internet Protocol is "best effort" and provides absolutely no guarantees of delivery across the network--obviously stuff gets to the other side most of the time, otherwise the Internet wouldn't work at all, but other mechanisms (the TCP in TCP/IP being a major one) on the ends deal with that.  That's fine when pretty much all applications are cool with being jumbled together and have essentially the same needs for delivery, which is what the Internet environment was all about decades ago when IP was first implemented.

Nobody thought about running voice or video over the Internet back in the Dark Ages (you know, the 80s).  Text reigned supreme and nobody was really running very complicated stuff.  Today we are continuously coming up with new applications (i.e., uses) and they diverge quite a lot from the old, quiet, tradition-bound world of the pre-web world.

So we need new approaches within the network itself and that's just what the providers are implementing.  In the enterprise environment they sell the same old vanilla flavor "best effort" service (marketed with better-sound names life "basic service" or "standard service") and for a little more coin, corporate customers can buy a "premium service" that offers differentiation between data from their various applications so they can improve overall performance and ultimately, the end user's experience.

An example of what happens if you treat everything the same in your network: Champlain College implemented a Cisco Voice-over-IP solution a couple years back for the usual reasons, including cost savings.  A major, unanticipated issue was a severe drop in voice quality at certain times of day--what some people called "the monster on the other end of the phone" problem.  They'd try to make a call, but for some reason the person on the other end would sound horrible, almost scary.

Turns out that students were getting done with their classes and firing up programs like BitTorrent and their interactive gaming stuff and effectively bogging down the network so much that the little teensy packets filled with voice samples were getting overwhelmed.  Lose too many of those things (most such networks just throw out packets if there are too many), or have them caught behind a lot of big data packets (think of waiting at a RR crossing for a long train) and you can't reconstruct the actual voice on the other phone.

Solution?  Start giving voice packets some logical differentiation on the same physical network.  They examined the traffic patterns, tweaked things in their network devices and POOF!  The Monster was exorcised and the students never noticed any change in their gaming or perfectly-legal (ahem) downloads.

So now I can address one misconception about how the Internet currently works and will work in this Fantabulous New Age of Quality.  One thing I hear from folks is they're afraid AT&T will create two Internets: a highspeed King's Highway one for rich people, and a lowspeed, crappy, pothole-pitted Low Road for us peasants.  In many minds The Evil Empire is literally building separate physical networks and will "discriminate" between who can pay for access to the Internet and those who will be shuffled off onto the back roads where you can't get to your favorite political blog and so forth.

It's actually a single infrastructure that AT&T has deployed to carry everybody's stuff, and it's just a matter of what CONTENT PROVIDERS, not end customers like you and me, pay for in terms of levels of service from that one network.  So YouTube can stay with the traditional "best effort", which means no additional cost to them and we can continue to enjoy embedding fun videos in our blog posts.  Yet if they wanted to start selling a premium high-definition movie service over the same Internet, they could put some extra skin into the game and have their video perhaps delivered in such a way that it looks lovely when you watch it on your computer.

I think this is where people get off the bus because they don't see how this can happen without "degradation" of our regular service.  Perhaps a simple drawing will help:

Let's say I've got a router in my IP network that is receiving three types of traffic, each at different rates and with different needs.  One of the things that happens inside the router is it looks at the "class" each packet belongs to and then puts them in appropriate queues.  Depending on the rules the router's configured with, it will then pump these packets out in a "fair" fashion based on how much stuff is in each queue, what the QoS parameters are that each application has contracted for, and choose the appropriate paths for those packets. 

So maybe the HD video gets put onto a path that offers high bandwidth and the voice and e-mail are interleaved on a lower speed path that still meets their basic requirements.  Possibly some of those green packets will be shunted off on the lower path, too.  Yet each colored packet will get access to the rest of the network, albeit in an arbitrated fashion to make sure the green video stuff doesn't take up all the room in the "tubes" (thanks, Senator Stevens!). 

Voice and video, being more sensitive to delay, might get a bit more space than the e-mail, but the e-mail will still be delivered fairly.  Yes, perhaps it will get slightly more delayed than if this were solely a "first come, first serve" algorithm, but we're talking milliseconds here, which are far below the threshold of any human to notice.  All the while this is improving the experience for people watching video or using Skype or Vonage to talk to friends in Oz.

Now it would certainly be reasonable to expect if YouTube wanted to offer a new, high-quality video service that they would charge more for the increased value.  That better, premium service is going to require extra quality and priority delivery across the network, which has cost the network providers money to implement with new, fairly pricey equipment.  So why shouldn't some of what they charge us go to the providers upon whom they rely?  Honestly, I still find it a puzzlement that people raise such objections, but perhaps that's because there are some other misconceptions about the Internet of which I need to disabuse them.

First, the biggest worry people seem have is that somehow AT&T is going to block political blogs, or create such conditions that it will "take 10 minutes to load Eschaton," thereby limited political speech.  I hope at least from the QoS discussion you can see that there's no technical reason why that would be the case.  As for the potential to begin filtering content the way countries like China and Vietnam do, well...the additional features providers want to charge for aren't necessary if they wanted to play along with a fascist regime, nor would the Net Neutrality language I've seen address the issue. 

It's trivial to block sites, specific content, etc.  Yet we haven't seen any of that without Net Neutrality because, as it turns out, Net Neutrality in its purest form already exists: no provider wants to be in the business of filtering content and, in fact, aren't, because that would destroy the very thing they want to make money from, namely the Internet.  Anytime a provider has done bad things, such as blocking VoIP phone calls from a competitor, the FCC slapped them down with current regulations.  That won't change.

Second, some people have expressed concern that AT&T will eventually stop supporting the old free services in favor of only the premium pay services.  I suppose at some point one can expect all older services to fade away as more people move to the newer, splashier offerings.  But the level of penetration in the market is low in the beginning of any product lifecycle (Fischer-Pry, anyone?), and as more people adopt the new technology, it will become cheaper (as we've seen with residential broadband deployment).

Beyond that, however, there's still a technical reason "best effort" won't go away: there's no benefit to the provider to get rid of it.  IP inherently supports it, so there's really no way to turn it off.  Remember, we're not talking about separate infrastructure (in contrast to AT&T's old TDMA network versus the newer, better CDMA network that Cingular had).

Third, somehow people think that AT&T is trying to make money off something that should be free because "the government built the Internet with our tax dollars."  Nope, the government hasn't been in the networking business in years, and even when they were (e.g., with the NSFNet quasi "backbone"), providers were still building their own infrastructure with their own capital without government assistance.

Yes, the DoD did originally fund ARPAnet for research purposes many aeons ago, and out of that research came the TCP/IP suite of protocols that we use today.  But TCP/IP is still free and is running over networks that the DoD had nothing to do with.  The Internet is really not one monolithic network, but really a "network of networks" where AT&T's network connects to Verizon's and Deutsche Telecom's and even Ma And Pa Kettle's Basement Internet Service.  They all use the same protocols so you can send me e-mail to bitch about this post even though my ISP is different from yours (not to mention I'm on dialup while you've got DSL, you lucky bastards).

One of the best examples of knee-jerk reactions against the providers is when people were freaking out in the wake of the patent infringement decision in a suit Verizon brought against Vonage.  Folks who don't understand patents were scared that somehow Verizon was claiming a patent "on the Internet".  Nothing is further from the truth, but because people naturally (and with great cause) distrust corporations and don't understand some fundamentals, they assume somebody's trying to pull a fast one.  They ain't.  At least, you know, where the Internet's concerned.

Anyway, many bloggers have been decrying the recent FTC decision regarding NN:

In the absence of significant market failure or demonstrated consumer harm, policy makers should be particularly hesitant to enact new regulation in this area.

That's essentially in the same camp as the FCC:

Kevin Martin, the FCC chairman, said during his portion of the keynote address that he believes the FCC's existing principles are sufficient to address problems that may arise should network operators block traffic.

"I think the FCC has authority to act," he said. "And it has done so in the past."

That's where I sit because having Congress try to legislate anything regarding fast-changing technology is a recipe for disaster.  Really, do you want Ted Stevens authoring laws about "the tubes"?  Nope, neither do I.  And as the FCC and FTC already can deal with any truly unfair practices by providers, I'd rather follow the "first, do no harm" guidance when it comes to new regulation.

So back to those NN definitions.  First:

Absolute Non-Discrimination

This is completely unreasonable and impractical in a modern Internet that is supposed to deliver new, different innovative applications.  Hopefully the stuff you know about QoS will make that conclusion fairly obvious.

Limited Discrimination without QoS Tiering

Why do Democrats hate capitalism?

Seriously, here's one example of some the zombie legislation wandering the halls of Congress:

With respect to any broadband service offered to the public, each broadband service provider shall—

(1) not block, interfere with, discriminate against, impair, or degrade the ability of any person to use a broadband service to access, use, send, post, receive, or offer any lawful content, application, or service made available via the Internet;
(2) not prevent or obstruct a user from attaching or using any device to the network of such broadband service provider, only if such device does not physically damage or substantially degrade the use of such network by other subscribers;
(3) provide and make available to each user information about such user’s access to the Internet, and the speed, nature, and limitations of such user’s broadband service;
(4) enable any content, application, or service made available via the Internet to be offered, provided, or posted on a basis that—

(A) is reasonable and nondiscriminatory, including with respect to quality of service, access, speed, and bandwidth;
(B) is at least equivalent to the access, speed, quality of service, and bandwidth that such broadband service provider offers to affiliated content, applications, or services made available via the public Internet into the network of such broadband service provider; and
(C) does not impose a charge on the basis of the type of content, applications, or services made available via the Internet into the network of such broadband service provider;

(5) only prioritize content, applications, or services accessed by a user that is made available via the Internet within the network of such broadband service provider based on the type of content, applications, or services and the level of service purchased by the user, without charge for such prioritization...

Part 1 I'm cool with, though the "lawful" content can open up a can of worms.  Part 2 is also fine, though it's probably unnecessary in light of Carterfone.  Part 3 makes me scratch my head, though it's certainly not fundamentally offensive. 

But all those sections seem to be much ado about nothing and are not what NN advocates are up in arms about anyway.  Parts 4 (specifically section C) and 5 are what I take issue with because of the charging issue.  Which brings me to the last option that Sir Tim advocates:

Limited Discrimination and Tiering

This is coming from somebody who understands the Internet from a technical perspective, as well as the implications of new applications on the network.  It's a balanced approach and I think when people get some of the fundamentals, it's the one that everybody can find common ground with.

The funny thing is that ultimately being able to charge for differentiated services ain't gonna do providers much good anyway.  The real money is in developing applications, not delivery per se.  But they do need to recoup investment in the network now to bridge them to the application-centric era in which they will be competing.

Long post boiled down: decisions by governing agencies to reject adding Net Neutrality regulations do not mean "we're fucked," as I've seen some in the blogosphere breathlessly claim.  It's reasonable to avoid imposing anything new and unnecessary as we continue to develop a modern networking infrastructure (not unlike the moratorium against Internet taxation that was in force for so long).

ntodd

PS--As with all my long posts, I will be fixing the inevitable stupid things and also be updating in response to queries and new thoughts.

[Update: NYMary was asking me a few clarifying questions, so let me add one more boiled-down note. 

The networks are NOT looking at the content in the IP packets.  They don't care what the content providers are sending.  The content providers, however, are marking their packets with the priority levels they want and then the network delivers that data according to contracted rules.  So the decision about prioritization comes from THE CUSTOMER, not the network providers.]

Fire Isn't Obsolete

Our CTO has a nice post on broadband in Vermont and makes some good points about DSL in general.

ntodd

Converging Is Hard Work

Chatted with the VP of Engineering and founder of Digital Fairway about a software suite that can help enterprises (and carriers) deal with all the disparate networks they are converging...

ntodd

OSS? We Don't Need No Steekin' OSS!

My boss and I did a podcast today with the CTO of Sigma-Systems in Toronto about OSS in IMS.

ntodd