This debate was central to the fascinating discussion at the State of Telecom event at Columbia’s Instititue of Tele-Information, held in mid-October. I attended the afternoon sessions.

Wall Street’s vision

Simon Flannery, managing director at Morgan Stanley, described the challenges that are eroding the margins of the cellcos. Of course, the top two cellcos are doing better than the rest. Flannery said that margins at Verizon at about 45 percent, while margins at Sprint are about 16 percent. Apps that are eroding core revenues include free text messaging and free calling.

Meanwhile, cellcos are selling advanced services that require more bandwidth. “Backhauling fiber to towers requires a massive build,” said Flannery. “Smaller carriers lack the cash flow to reinvest, and there is no financing for newtowkrs that are without returns.”

The market is trending towards a duopoly.

Craig Moffett, senior analyst at Bernstein Research, said that the services that people pay for are the easiest to provide: internet access, phone calls, and so on. “People are less willing to pay for information and entertainment, which are services that cost more to provide.”

Voice may require 9.6 Kbps and people will pay $50 per month for it. People will only pay about $30 per month more for the next generation services that multiply data usage by 10 or 100 times. “The sale price per bit is falling faster than the cost per bit.”

At the bottom end of the market, Moffett said, there is the “poverty problem” where households whose net income is negative after paying for food, clothing, and shelter account for perhaps 40 percent of all homes. “Retailers can depend on the upper two quintiles, but telcos have to sell to the full 100 percent of the population,” Moffett claimed.

The consumer advocate concurs

Mark Cooper of the Conumer Federation claimed that he disagreed with everyone on all sides of this debate. He said that in some poor countries, there are 75 cell phones per 100 people. “People who have no electric power at home may have cell phones.”

If a cellco is just adding voice customers, it’s easy to grow. “It is easy to add subscribers but it is expensive to add capacity. Users, uses, and usage all add costs to wireless mobile networks.”

He agreed with the Wall Street analysts that the marginal sale price of bandwidth drops rapidly.

He added, however, that unlicensed wireless spectrum is the great success story of the past two decades. Even AT&T is now selling Wi-Fi. “Unlicensed has no champion in the scrum for spectrum.” Cooper said that at most 10 percent to 20 percent of spectrum should be sold to the cellco monopolies, so that the Washington can avoid the next monopoly and the next “100 year mistake.”

The Economist Magazine has the data

In an article entitled The Limits of Frugality, The Economist magazine warns that those rural cellphone users who have no electricity in their homes will soon be paying higher prices. “Sunil Mittal, the boss of Bharti Airtel, the mobile-phone operator … said the extra cost of servicing rural customers, and their low usage levels, had made things unprofitable. Prices are now expected to go up across the industry, after two decades of decline. India’s low-cost telecoms revolution has, it seems, reached its limit.”

Buildouts will now focus on the urban rich. “Today perhaps 17 percent of India’s population has half of its spending power, according to the Asian Development Bank …. One proxy for the difference in profitability between the urban rich and the rural poor is the price paid for mobile-telecoms spectrum. In the 2010 auctions for 3G telecoms licences, operators bid ten times more for a slice of the airwaves in affluent Delhi, with 18m people, than in east Uttar Pradesh, with 120m people.”

Washington, D.C.’s policymakers should expect fixed wireless and wireline internet to connect the rural poor at an affordable price. The true price of cellular broadband is going up fast, worldwide, and like all price rises, it will harm the poorest the most.

Neutral exchange

Paolo Gambini, CMO of Tinet (formerly part of Tiscali), announced that his company has formed an alliance with PCCW of Hong Kong that will give PCCW access to North America and Europe and give Tinet access to Asia-Pacific.

Gambini also said that the company is growing its own EtherCloud platform, which he descrbed as an alternative to ethernet exchanges.

Transatlantic cable

Hibernia Atlantic announced that it will be laying a new transatlantic cable by the middle of next year and offering service on it in Q3 of 2012. The cable aims to reduce latency between the London and New York exchanges by at least 10 percent, to less than 60 ms.

Automated quotes

Greg Hough, CTO of Global Capacity, announced that Lattis Global, its tarriff quoting system, had likely saved 150,000 man hours of work last year by transforming a task that could take hours into one that took several clicks.

The company has added a new offering, One Marketplace Access, whose first customer is MegaPath (which also owns Covad and Speakeasy). OMA delivers specific price quotes for access to specific circuits, making buildouts easier, especially in areas where an ISP does not already have facilities. “We have reduced MegaPath’s SG&A and can help them reach tier 2 and tier 3 markets,” Hough said.

Fiber in New Jersey

Vincenzo Celemente, the very young CEO of Cross River Fiber of Isselin, N.J., announced that it is extending its fiber to the major internet exchange points across the state. The new build will reach important points in the following New Jersey towns: Secaucus, Clifton, Nutley, North Bergen, Newark, Cateret, Edison, Piscataway, Somerset, Rochelle Park, and Totawa.

A great duct system

Hunter Newby, CEO of innovative fiber builder Allied Fiber, praised the contruction of DuPont Fabros‘ Piscataway, N.J. data center as the two companies announced they had connected. Allied Fiber promises to break open its fiber every 3,000 feet as it builds a network across the U.S. (phase one, which is nearing completion, forms a triangle between New York, Chicago, and Ashburn). Allied Fiber found that DuPont Fabros’ data center was only 1,000 feet away from Allied Fiber’s route, which runs along railroad rights of way.

Newby said that the duct system in the Piscataway data center made it very easy to connect to the data center. “The physical infrastructure, multiple points of entry, and the building ducts made this data center unique. As I go across the country, costs vary widely, and costs are mostly determined by landlords, who also vary widely.”

Vinay Nagpal, director of carrier relations for DuPoint Fabros, said that his financial industry customers are using SAN technologies for instant replication of critical data between New Jersey and Ashburn. “Some are able to run SAN applications for up to 180 miles without regeneration.”

Connecting the Toronto exchange

Colleen Gallagher, vice president of business development at First TelecomServices, said that her company has recently lit the last leg of fiber connecting Toronto to its network. The company is a former subsidiary of an energy company and has fiber assets (see map) that connect Chicago, Washington, D.C., and New York and has points of presence in eleven states. “We now offer industry leading low latency connections to Toronto,” she said.

A fiber blog

Metro NS announced that they have launched an ethernet-centric blog.

A version of Rockbox for a particular MP3 player is deemed stable if it meets all of the following conditions: it supports the latest version number of Rockbox (currently version 3), it can be installed with the Rockbox install utility, and it has a manual. An unstable version lacks any one of these (sometimes just a manual) but it will generally work fine. An unusable version is in development and can crash.

“Can you brick your MP3 player,” one NYLUG member asked. Menes said that there is a risk when using most MP3 players, except that iPod software is “unbreakable” because iTunes can reset the player to its original state.

Menes used a very small SanDisk Sansa MP3 player running Rockbox which allowed him to use it as a clicker to control his presentation. He said that this Human Interface Device (HID) feature had been added by a student during a Google Summer of Code and then perfected by a Rockbox developer.

The Rockbox project is committed to being free GPL software and it therefore does not support DRM and can never support DRM because DRM requires that the software not be open source.

The software supports over 30 different audio formats, far too many to list here.

It is optimized to run on machines with very little power and with weak CPUs, machines running on battery power using CPUs as weak as 12 MHz.

The end of hardware?

This software project is designed for hardware that will soon cease to exist. In the future, people will use Rockbox on their laptop, netbook, or cellphone. The latest development in the Rockbox project is a port to Android that is not complete. How much hardware will be replaced by open smartphone architectures?

The port to Android is slimmer than the software that is installed on an MP3 player or PC. “We did not need to install a TCP/IP stack because the phone already has one,” Menes told me.

Developers unpacked the Rockbox software, which was a self-contained OS-to-software stack, and built universal connectors for it. Thus a stack of software was cut into pieces and the essential functions received software connectors to other parts of the phone’s OS. I think that in the future, many other pieces of software will be cut into pieces the same way.

Features

Like many open source projects, Rockbox has many more features than any closed source competitor. It can record FM radio, it can speak menus, it can deliver sound effects (including sound effects for each song), it can fade songs in and out, search files by name or by any of several tags in a surprisingly powerful database, and it can do many other things too.

Check out the 1 star reviews for this company. Avoid them.

And shame on citysearch for allowing fake reviews.

Every system administrator lacks the time to do everything that needs to be done. “I’m not good at time management,” Limoncelli said, “but I’ve created many coping mechanisms to deal with the problem.”

“It’s not your fault,” he added.

Part of the problem is evolution. The human brain may be suited to surviving in the wild, but it is not suited to time management. For example, one key time management skill is memorizing long lists, a skill not suited to the neocortex, which is roughly the most recently evolved part of the human brain. However, the neocortex is good at making tools. So use them: paper, pencils, smartphones.

Another part of the problem is infrastructure.

If you are your own boss, then you made some tough decisions about infrastructure and you have to compensate for your own lack of investment. Limoncelli remembers a time when he had to persuade companies to adopt trouble ticketing systems.

In his speech, he posted a funny graphic: a trouble ticket that said, “The worst has happened. There’s no coffee!”

Limoncelli added that he has worked at makeshift desks, computer rooms that were too small, and with monitoring systems that could not parse the data they were gathering.

“How many people here have worked at a company that tried to save money by purchasing consumer grade networking equipment instead of professional grade equipment,” he asked. Many in the room raised their hands.

Recommendations

Limoncelli recommended keeping a written To Do list. It can be as simple as a Notepad text document. For Emacs users, Limoncelli recommended Org-Mode, a powerful To Do list system that can be partially automated.

Part of your job is getting rid of interruptions — which is as much about dealing with people as it is about dealing with technology. Limoncelli recommended writing down a request — with large and highly visible hand motions — rather than interrupting what you’re doing and allowing someone else to set your priorities.

Limoncelli recommended what he called “the 4 PM check.” That’s the point in the day when you know whether or not you will be able to do everything that you had planned to do. If you cannot get it all done, he said, you have three options: negotiate an extension, delegate the work, or work late and complete the take.

“Stop working late every night,” he advised.

If you are managing desktops, you should be patching and cleaning machines, backing them up, and making sure that you can restore from backups. Limoncelli referred to the Evard Life Cycle of a Machine.

This is common advice, but Limoncelli had powerful stories about the consequences of not following it. He said that he once came into a company, as a consultant, and found that a key machine was generating a tremendous quantity of security problems. When he walked into the administrator’s office, he did not have to ask why the problems were occurring: he could see, on the desk, in pristine, unopened packages, all of the patches that should have been implemented. The company refused to pay for a test environment for this vital piece of hardware, so the administrator refused to patch it for fear that the patch could take down the machine.

If you have systems, they have to be useful. Limoncelli had several stories about the consequences of burdensome security and backup systems. The result of a burdensome system is that users will avoid it and it will therefore not serve its purpose.

Be a leader

Limoncelli said that administrators must be leaders as well as managers. A manager sets priorities and allocates resources; a leader is the person who goes first and makes it possible for others to follow.

Sometimes leadership means helping others do their job. Limoncelli told the story of a security consultant who was brought in to a company that was growing so fast that machines were being added faster than security issues could be fixed. So the security consultant did what the IT department should have done: she automated many IT tasks so that the IT department could do both tasks: fix security issues and also add new machines.

Limoncelli recommended documenting what you do. He said that the documentation could be bullet points. It could be simple. He recommended documenting the tasks that you least like to do. Eventually, you will be able to delegate them. Also, the documents will allow you to occasionally take a vacation.

Eventually, when your department grows, you can recommend that the company hire a cheaper, junior person who you can mentor. When the company asks you what they will do, you can read from the list of documents you have created. You will be delegating all of the tasks you least like to do.

With good documentation, you may also find that you can automate some of these boring tasks.

People skills

Let’s say you’re approaching your boss. Which of the following do you think would be more effective:

“We need a faster and better server”

or

“I figured out how to reduce the time that our salespeople spend uploading data to the CRM system so that they can spend more time selling.”

If people start e-mailing you instead of using the trouble ticketing system, Limoncelli advised that you be firm but polite. “I always say that it’s nice that they e-mailed me but that I’ll get the message if they use the system and that I’d hate to think that their problem could go unnoticed for a few weeks if I was on vacation.”

Conclusion

Limoncelli is a fantastic speaker. In this schematic summary of his talk, I’ve left out most of the anecdotes that give weight to his recommendations. If you have the chance to attend a speech of his, take it! His tour schedule can be found on his website.

In an earlier interview, Grippi said, “it was the first time it made us think of the violence of those that use your data, and of how, behind the scenes of someone who offers you something for free, there’s always someone that uses the data you exchange with your friends. We deleted ourselves from Facebook and we started to think about an alternative. People don’t really understand the risks they’re taking, but even those that understand them don’t know where else to go to.”

Before starting Diaspora, the club had built a MakerBot and had completed other projects together, such as having the door to the club room tweet every time it was used.

The group decided to build a decent social network for nerds. They went to Kickstarter, whose founder is also a graduate of NYU. Kickstarter allows anyone to raise money for any project through very small donations, as little as $5 per person. The group posted what Grippi called “a pretty terrible video.” The goal was to raise $10,000. In fact, they raised $200,000, which at the time was a record for Kickstarter.

Raphael’s brother Michael was a key part of subsequent development. Michael works at Pivotal Labs, which is now an investor in the Diaspora project and which provides the project its office space in San Francisco (Pivotal also provides breakfast to team members).

Diaspora today

The team has been working on the project for seven months now. Its code base is stored on Github and the project’s public face is the Join Diaspora website.

The project now has 106 contributors. It is AGPL3 compliant.

It is federated. This means that anyone can set up a Diaspora seed (the word “diaspora” is ancient Greek for “scattering of seeds”). Each seed houses the accounts of its subscribers. When people communicate with subscribers of other seeds, the message passes between seeds seamlessly. Currently, the project’s founders operate the largest installation but other large seeds include one in Seattle and another in Germany.

The interface for a user’s profile allows the user to maintain their contact list, aspects, connections, and to connect an account to Twitter and Facebook. The user interface is still in development and is subject to change. An account description is like an e-mail but it is not an e-mail at this time. The structure is username@pod.

Grippi said that where possible, the project tries to avoid reinventing the wheel and instead uses existing protocols, such as Webfinger, hcard, and Salmon.

Sofaer explained that with the hcard protocol, a user can decide whether an image is searchable or not.

For security, the project relies on another trustworthy public protocol: the open SSL RSA library.

The Salmon protocol helps distribute comments. Unlike other social networks, a user cannot see a friend’s comment on a wall if they are not a friend of the person on whose wall the comment was made.

The project currently runs on the Rackspace cloud and stores software images on Amazon S3.

It uses Splunk for log analysis. The group also tracks virtual machine statistics and is especially interested in the time required for specific queries. If a query requires too many milliseconds, there’s an error in the code that must be fixed.

Diaspora in the Future

“Moving forward, we will stablilize and formalize Diaspora,” Sofaer said. He explained that the team’s initial goal was to create software that worked. Now that it’s working, the team needs to formalize how it works in order to ensure that it can scale.

“We wanted to get something up and running without debating protocols for two years,” Sofaer said.

“I want my sister to use it because it’s good for her,” said Grippi. “We want to see many users of free software.”

The eventual goal is that as seeds proliferate, they will compete with each other. Some will offer better privacy, for example.

The team wants to make an API. They want to bring Diaspora to cellular networks with a mobile API. They want to work with existing trust communities, for example by supporting OAUTH.

NYLUG members asked whether Diaspora is ready to participate in the Web of Trust, but the team noted that the technical requirements may be beyond the managers of some seeds. Grippi said that although Diaspora was built for nerds, many people who are not very technical want to participate. He said there are currently about 100,000 members and a waiting list of about 300,000 e-mail addresses.

One part of the task of making Diaspora easy to use is nearly complete. The project has tools called Sod and Chef that are designed to automate the deployment of a seed. Chef is made by OpsCode, and Sod was built by Michael Sofaer.

A user can buy a domain from Rackspace and deploy in three clicks, Sofaer said.

“We’re all developers and no sysadmins, so we needed auto provisioning,” said Sofaer.

Conclusion

The NYLUG group had many tough questions. What happens if someone malicious tries to participate in the project or tries to deploy a bad seed?

But overall, the NYLUG group was very impressed by the project. Attendance was high for NYLUG and the speakers were mobbed after they spoke.

One of Sofaer’s comments stuck with me. He said, “we wanted to make social networking more contextual and to give users more control over what they’re sending and who’s receiving it.”

Before NYLUG, I had attended a speech on the importance of context in the preservation of privacy (see Privacy in Context: A Speech at Fordham Law School).

An earlier version of this story incorrectly said that the Diaspora project team members built Sod and Chef. Thanks to Sofaer for the correction.

Nissenbaum said that she thought this would be the easiest book she’s ever written, because it was simply synthesizing many papers, but that in fact it was the hardest, and it took two years. In the book (and in her current work) she is building an analytical framework that would identify the aspect of an online transaction or interaction that causes social anxiety.

The miracle of the internet is about the rapid dissemination of information. This has delivered powerful economic benefits, and it has delivered freedom.

The internet has also enabled massive data repositories that have caused concern. Nissenbaum mentioned Choicepoint and you can see the concerns of the Electronic Privacy Information Center here. Nissenbaum also mentioned the Total Information Awareness (TIA) program of the federal government. Of TIA, EPIC wrote, “TSA has failed to meet its legal obligations for openness and transparency under the Freedom of Information Act and has violated the spirit of the Privacy Act for the protection of privacy rights.”

Getting the terms right

Nissenbaum said that when we talk about privacy, we need to reject the dichotomy between public and private. If all information is available on the internet, what matters is not whether it’s there, but the context in which it was posted to the internet.

Her prescriptions can be broad. In 2004, she wrote a paper on surveillance called Privacy as Contextual Integrity in which she argued that surveillance footage, for example, must be used within its security context but must not be exported outside it.

She wrote, “when violations of norms are widespread and systematic as in public surveillance, when strong incentives of self-interest are behind these violations, when the parties involved are of radically unequal power and wealth, then the violations take on political significance and call for political response.”

In her speech, Nissenbaum said that information transmission is critical to society. In the paper, she noted that many surveillance activities today could violate any law that attempted to codify our social norms and expectations of privacy. In her talk, she noted that any overly broad law would have a serious negative effect on society and on the benefits that the internet has brought to the economies and societies of the world.

In a conversation, she noted, we control the transmission of information. We talk to a specific person. In the context, we understand how our friend will use the information. In another context, such as in a therapy session, one person talks and the other listens and the conversation remains confidential.

“Privacy issues arise when everyone instead of a specific person receives the information,” Nissenbaum said.

Nissenbaum said that laws already understand that privacy issues invole one actor transmitting information and another receiving and then using that information. See, for example, section 6802 of the Gramm-Leach-Bliley act which begins:

Except as otherwise provided in this subchapter, a financial institution may not, directly or through any affiliate, disclose to a nonaffiliated third party any nonpublic personal information, unless such financial institution provides or has provided to the consumer

The act recognizes the obligations of the recipient of private information (the financial institution) with regard to information received from an actor (the consumer).

The internet brings change

“The world is changing, upsetting entrenched norms,” Nissenbaum said. She said that the changes affect vested interests but they also effect core values. In order to evaluate privacy questions, we must consider the context.

For example, she noted that the results of HIV tests must remain private or those who are infected or might be infected will avoid health care.

She said that PACER has put all federal cases online but added that there is debate — good debate — occurring at the state level.

She said that data mining can be good. It can improve health care and education. But many people are not aware of the scale and scope of current data mining by businesses. She recommended a series of articles in the Wall Street Journal called What They Know About You. Of particular interest to readers: After sharp internal debate, Microsoft designed Internet Explorer so that users must turn on privacy settings every time IE starts.

If human beings become concerned about the privacy implications of online activity, Nissenbaum noted, they might stop shopping online, which would be bad.

A Fordham student noted that although she herself has opted out of Facebook, she still worries that her friends will post photos of her online.

The systems that Nissenbaum is developing that will be used to analyze the context of online transactions seem most useful for online venues where structured activity takes place. For example, as states analyze the privacy implications of online educational record systems, her schema will provide a valuable framework.

But in areas of the internet that more closely resemble an actual conversation, it seems unlikely that any schema can list all of the layers of meaning and context that surround an actual conversation between friends.

Nissenbaum’s talk shows that while governments can solve many of their privacy issues, social networks face greater challenges. This was timely: after visiting Fordham Law School, I went to a meeting of the New York Linux User’s Group (NYLUG) where the open source Diaspora social networking project would present (see Diaspora at NYLUG).

However, the VoIP services that exist today are a shadow of what the technology makes possible. VoIP has been choked so that it no longer disrupts telephone service. VoIP has been fenced in by the FCC so that it offers no more than telephone, a move that was intended to protect cellular and wireline phone companies.

I remember that in 2004, we held ISPCON Spring in Washington, D.C. The vendors brought their office phones with them to their booths. They were thrilled to be able to answer their office phones while on the road. While this was a boon to the CEOs of small companies, it was a miracle for road warrior salespeople. All of these businesspeople were using a feature of VoIP that wireline service cannot provide — and that directly competes with cellular too.

In 2005, the FCC ruled that VoIP service must support 911 emergency service (the E911 order). This was the beginning of the end. For nomad businessmen such as those who had been so happy at ISPCON in 2004, there was no need for 911 on their VoIP service. Residential users who had taken their home phones with them on vacations could no longer do so. A key advantage of VoIP technology over traditional wireline phone and cellular services was eliminated by FCC fiat.

At the time, VoIP expert Jeff Pulver said, “It is my hope that the FCC’s E911 rules for Interconnected VoIP providers do not extend to services that no user would expect to offer localized emergency response capabilities, such as a circumstance where a computer has both SkypeIn and SkypeOut downloaded … [the order] might relegate America to a VoIP ghetto, where anyone wishing to use both an inbound and outbound voice application on her computer had better not step foot in America. Should Skype consider turning off either its inbound or outbound service in America?”

In 2006, the FCC attacked VoIP’s inherent price advantage by forcing VoIP providers to pay universal service fees. This rule hit the smallest companies the hardest.

VoIP providers were also required to enable wiretaps (CALEA), a ruling that was reasonable, but which was made unreasonable by the federal government’s unwillingness to certify systems as being compliant. VoIP companies were expected to pay $100,000 or more for a system that might or might not do what it was supposed to do. This hit the smallest companies the hardest.

In 2007, VoIP providers were required to register for a CPNI, another bureaucratic procedure that hit small innovators but provided little encumbrance to the monopolies.

VoIP, caged

The result of these rules is that today, the largest VoIP companies are massive companies, and the largest single VoIP provider in the U.S. is Comcast. As of September 30, 2010, Comcast had 8.4 million VoIP customers, almost half of the VoIP market all by itself (before counting the other cable companies).

The “digital phone” service offered by cable companies today is the result of FCC policies: it is a service that runs over the internet but whose features are restricted to mimic as closely as possible the phone service of the previous century. It competes with wireline voice services on price but, more importantly, it no longer allows its users to roam and therefore no longer competes with cellular service. The phone companies are now selling off their wireline voice services, have stopped investing in internet service, and are spending all of their money on cellular. They do not mind if VoIP competes with wireline phone services; they must be thrilled that it no longer competes with cellular service because “digital phone” cannot move.

Some providers are doing wonderful things with VoIP — the innovators are the smallest companies on whom the burden of regulation falls the hardest. M5 Networks, a business-only VoIP provider in New York City, provided hundreds of virtual phone numbers to real estate firm Coldwell Banker. Every advertisement that the realtor placed that week had a different phone number, so the realtor was able to track, in minute detail, the response rate of every ad. Of course, the FCC has left the rules for business VoIP services more liberal than those for residential. As a result, many of the most innovative local players offer business VoIP services only.

Digium, the business end of the Asterisk project, offers a wide open technology that can be used to do many things. I spoke to one developer several years ago, Professor Brian Capaouch, who set up an anti-raccoon device in his chimney using Asterisk. When a video camera sees a raccoon, it turns on a kitchen mixer that is broken — it’s missing a ball bearing and squeals like murder. The video camera also takes a still photo and e-mails it to the professor. The professor can “call in” to the camera and view the live image on his cell phone.

Capaouch told me that VoIP enabled a completely different kind of collaborative development. Because the calls were free, developers would open a party line and talk to each other for hours while working on code together.

Simple disruptions of the telco business model were also important. You could obtain a phone number in another city or another country and avoid international and long distance phone charges. In the same article cited above, I interviewed Greg Boehnlein who told me, “our operations manager is from London and he was spending $400 per month calling his parents. Now, with VoIP, he pays zero.”

Today, while experts can still do truly amazing things with VoIP, the services that residential consumers see are limited. They compete on price, not features, and deliver something very similar to traditional phone service. Most use phones instead of computer microphones for audio.

Business services are somewhat more sophisticated, but for the truly nifty things that VoIP can do, even a knowledgable business IT department usually needs to hire a consultant like M5 Networks to get elaborate things done. Many IT departments also pay consultants whose job is simply to get them the telephone services they need and make sure they’re not paying for anything they don’t need.

VoIP technology is widely deployed, even within the phone company, but it has been prevented from fundamentally changing the phone business, largely through the intervention of the FCC.

And so journalism declined.

The Federal Trade Commission, worried about the decline, considered trying to stop it by subsidizing newspapers, a move that many on both the left and right said confused “newspapers with journalism.”

In fact, there are journalists, but they no longer work for newspapers, and news is no longer broken by newspapers. News breaks on the internet.

At a recent symposium that I covered here, Harvard Law School professor Yochai Benkler said that the broadcasters’ and newspapers’ eagerness to ask the Obama adminstration for permission to publish the Wikileaks story means that they will never break important news in the future. “The next Ellsberg will not risk his career and liberty to go to The New York Times. The whistleblower will publish on the internet and the media will report on it.”

Leaks are inevitable in the future, as The Economist recently noted. These leaks will no longer be screened by the editorial function of newspapers but will instead be dumped. The raw data will be available to everyone on the internet.

Some are arguing for a fascist response, and others hope that future leaks harm the powers that be. As the Economist article cited above noted, “Some of us wish to encourage in individuals the sense of justice which would embolden them to challenge the institutions that control our fate by bringing their secrets to light. Some of us wish to encourage in individuals ever greater fealty and submission to corporations and the state in order to protect the privileges and prerogatives of the powerful, lest their erosion threaten what David Brooks calls ‘the fragile community’ — our current, comfortable dispensation.”

In a separate article, The Economist argued that the U.S. government should accept current and future wikileaks, arguing for prosecution but not for persecution:

Calling Mr Assange a terrorist, for example, is deeply counterproductive. His cyber-troops do not fly planes into buildings, throw acid at schoolgirls or murder apostates. Indeed, the few genuine similarities between WikiLeaks and the Taliban—its elusiveness and its wide base of support—argue against ill-judged attacks that merely broaden that support. After a week of clumsy American-inspired attempts to shut WikiLeaks down, it is now hosted on more than 700 servers around the world.

The big danger is that America is provoked into bending or breaking its own rules, straining alliances, eroding credibility and — because it will not be able to muzzle WikiLeaks — ultimately seeming impotent. In recent years America has promoted the internet as a menace to foreign censorship. That sounds tinny now. So did its joy of hosting next year’s World Press Freedom Day this week. Chinese and Russian glee at American discomfort are a sure sign of such missteps.

The best lessons to bear in mind are those learned in such costly fashion during the past decade of the “war on terror”. Deal with the source of the problem, not just its symptoms. Keep the moral high ground. And pick fights you can win.

One thing is certain: the newspapers will no longer insulate governments and corporations from the consequences of the revelation of secrets.