Television: single service; Internet: multiple services

In a recent post, we spoke of the ISP's service-oriented business model, contrasting it with the unfettered delivery of information (bits) on the Internet.

As an example, we cited the NBA playoff games being streamed over the Internet as well as broadcast on television. We noted that the service-oriented business model maximized carrier profit at the expense of consumers and the national economy.

In addition to being economically efficient, the Internet allows us to combine the game coverage with services like the LA Times chat room shown here:

This and other services like providing expert commentary, statistics, alternative play-by-play reporting, supplementary video, etc. would add to the viewer experience and create an online community of fans.

Would you enjoy chatting with other fans while watching a sporting event? Would you rather watch a sporting event on broadcast television or your computer screen? Which would you prefer if the Internet speed increased to the point where it could match the size and resolution of your television set?

How will WolframAlpha change math instruction?

When portable calculators were invented, we debated whether students should be allowed to use them in school and on tests. WolframAlpha takes this question beyond grade school.

Stephen Wolfram, co-founder of Wolfram Research, invented and commercialized Mathematica, an incredibly powerful symbolic math program. A command line version of Mathematica is embedded in Wolfram's latest product, the WolframAlpha Web service. Here are a few examples:

• Enter y = 10 x, and you are told that it is the equation for a straight line with x and y intercepts at 0 and a slope of 10, and the equation is plotted.
• Enter x^2, and you are told it is a parabola, which it plots and gives the global minimum, derivative and indefinite integral.
• Enter x^2 + y^2 + z^2= 10, and you get a plot of a sphere along with its center, radius, diameter, volume and surface area and other information including derivatives.
• Enter sin(x)^2 + cos(x)^2, and you get a plot of a constant 1 along with series representations and integrals.
• Enter derivative x^3, and you get the derivative and its plot plus an indefinite integral and global maximum.
• Enter integral x^2 + 3x dx, and you get the indefinite integral along with its plot. If you click on "show steps," you see the steps leading to the solution.
• Enter definite integral x^2 (0,1), and you get the definite integral from 0 to 1 along with a plot.

You get the idea -- try a few for yourself, and you will be amazed at the power of this symbolic math program.

Should students of algebra, geometry, trigonometry, finance, statistics, calculus, etc. be allowed to use WolframAlpha in class, for homework, during exams?

P. S. For links to further discussion, see this post.

The information-service business model is costing us dearly

The incumbent Internet service providers have a service-based business model, differentiating between Internet, telephone, text-message, basic television, premium television, voicemail and other services.

This enables them to vary prices to maximize their profit and discourage competition. For example, they charge exorbitant fees for text messages and discourage Internet video by charging extreme prices when download caps are exceeded.

This differentiation between types of data or service is arbitrary. It is all bits.

For example, the NBA playoff game on Sunday was televised and streamed over the Internet. Both came over the same cable:



As we see, the television coverage, including the ads, is being delivered over the Internet with a four second delay. The Internet image quality is below that of the television signal, but that will improve when US Internet speeds catch up with the rest of the world.

The incumbent telephone and cable companies profit from their service-oriented business model, so they will resist becoming information utilities, delivering undifferentiated bits.

Can they sustain that position in the long run? Would we tolerate a water company that differentiated between drinking and washing water or a gas company that differentiated between heating and cooking gas?

The cost of high-speed international connections -- Africa loses

Telegeography, a market research firm, compiles a database of wholesale Internet prices -- the prices a global organization or an ISP would pay. This chart shows a few price examples:

The price differences are dramatic -- a 2 Mbps link between London and Johannesburg costs roughly the same as a 10 Gbps link between London and New York.

A recent New York Times article points out that the cost of serving Web content over slow, unreliable links in developing nations is high and the income from advertising is low. This has led some companies to block developing nations from their sites or to offer degraded content by, for example, compressing video or serving "light" versions of pages.

What are some of the causes of high international link prices in developing nations? What other factors contribute to the Internet being relatively slow and expensive in developing nations?

4G cellular is expected to roll out faster than 3G did

We have discussed three generations of cellular technology, and the fourth is beginning to take shape.



It took nearly six years for 3G cellular to reach 100 million subscribers world wide, but market research firm Pyramid Research estimates that 4G Long Term Evolution (LTE) technology will grow faster. They expect a compound annual growth rate of 404% from 2010 to 2014, with 136 million subscribers by year-end 2014.

Note that Pyramid Research expects LTE to rapidly surpass mobile WiMAX.

At 100 Mbps, LTE will not be limited to cell phone service, but will be used for fixed, portable and mobile Internet access.

If you live in a large city in a developed nation, you might look forward to LTE or Mobile WiMAX by 2011. Others will have to wait. (Some for a very long time).

Is 3G cellular currently available throughout the city you live in? Throughout developing nations? What applications and mobile devices do you foresee for 100 Mbps mobile connectivity? What changes may occur in the cellular companies business and billing models?

Substituing communication for transportation -- can we reduce business travel?

My son-in-law Mark flew from Los Angeles to Seattle last Thursday for a job interview. The round trip air fare plus four cab rides cost $619. He had a long day, getting up at 4 AM to catch an early flight, and was not home until early evening. If he had stayed over in Seattle, the a hotel room and meals would have brought the cost close to $1,000.

Three people interviewed Mark, and he feels that a teleconference would have been as effective as flying up for face-to-face interviews.

Ten years ago, I conducted a study of business travel at Hyundai, USA. Nineteen employees monitored their travel for three weeks, identifying trips which could have been avoided if various communication alternatives had been available. Based upon that data, we estimated savings of 3.6 hours and $298.07 per employee per week.

In the ensuing ten years, the cost of travel has risen somewhat and the cost of communication has dropped dramatically. Why did they fly Mark to Seattle rather than conduct a teleconference?

The standard answer is that information is lost in a teleconference -- you don't get the same feelings and subtle clues as when you meet someone face-to-face.

If you have any doubt that a teleconference can convey emotion, watch this excerpt from A Hole is Space, showing clips from a two-way video link between people in New York and Los Angeles in 1980. A Hole in Space was not a business installation, but performance art anticipating today's emerging high-speed connectivity.

A high bandwidth teleconference can convey emotion, but that is not widely accepted in our culture -- yet. But, my grandson has grown up making Skype video calls to his relatives in Chile. By the time he starts high school, he will be video chatting on his mobile Internet-access device. Will a teleconference seem emotionless or unusual to him when he joins the workforce?

Have you done video chats or conferences? If so, was it as effective as talking face-to-face? Would the experience have improved if you had a very powerful computer and very fast connection to the Internet? What are the economic and environmental implications of substituting communication for transportation?

International VoIP is growing fast -- what are the implications?

As a result of globalization, we are making more international phone calls every year. In 2008, we made 384 billion minutes of cross-border calls -- 12% more than 2007.
This graph shows the annual growth rate and the total number of call minutes for both traditional time-division multiplexing (TDM) and voice over IP (VoIP) calls. Both are growing, but VoIP is growing faster.

For example, traffic on Skype, the leading international VoIP provider, increased by 41% in 2008. Skype carried 33 billion minutes in 2008, accounting for 8% of international voice traffic.

This sounds like good news for the industry, but, as we see below, prices are falling as traffic increases, keeping revenue roughly flat.
(This data was from Telegeography, a leading telephony and Internet market research firm and our class notes on VoIP are here).

Do you make international calls? How does the cost per minute using Skype or other VoIP services compare to TDM calls? Do the traditional phone companies make any money on VoIP calls?

What will happen to telephone companies if VoIP continues to outgrow TDM? How will this trend effect developing nations?

Zappos and Flickr -- excellent and terrible customer service on the Web

Here is a short excerpt from a talk by Zappos.com CEO Tony Hsieh, in which he discusses their customer service.

Customer service excellence is a core value at Zappos. They publish their 24/7 toll-free phone number on every Web page. Customer service representatives are expected to give friendly, helpful “above and beyond” service, and Zappos does not time calls or set sales-based performance goals. They view a call as a branding opportunity rather than a necessary expense. Reps even direct customers to other Web sites when appropriate.

Zappos also gives free shipping and return shipping, accepts returns for a year, stocks every item in their own 24/7 warehouse, and often surprises customers with an upgrade to overnight shipping.

(Hsieh's full talk, in which he stresses culture building more than customer service, is here).

Contrast that with my experience with the absurd customer service on the Flickr photo sharing site. On March 13, I submitted a question asking why people could not see photos I had placed in a "public" Flickr group. The next day, I received this response:

Just a quick email from Team Flickr to let you know that we've successfully received your recent Help by Email query and we hope to respond shortly.

We'd also like to take an opportunity to remind you that one query is sufficient and multiple queries regarding the same issue make the Magic Donkey cry.

Lastly, you may not be aware that our FAQs and forums are full of help goodness.

The Flickreenos

They tried to blow me off to their FAQs, warned me not to resend my query, and irritated me with their cutesy language, but, that would be OK if they helped me with the problem. The next day, I received an email saying they had escalated the question to a "senior representative."

But, evidently the senior representative had a five week backlog. On April 24 he got back to me with a non sequitur, boilerplate-laden answer.

Customer service clearly costs Zappos more than it does Flickr, and they are targeting a high-end retail customer, but, no matter what the purpose of your Web site is, you should study Zappos' example and do a lot better than Flickr.

Have you seen other examples of very good or very poor customer service on the Web?

Would you like to own and install fiber to your home?

There are many examples of municipal ownership of access networks, but ownership and control could also be pushed out to home and building owners.

Wu and Slater discuss this alternative in a recent paper and a test is underway in downtown Ottawa, Canada, where fiber has been deployed to serve a 400–home neighborhood, but a service provider has not yet signed on.

The Norwegian telecommunication company Lyse Tele reports that 80% of their 130,000 customers have agreed to dig their own trenches and bury their own fiber in exchange for a discount on installation. This has been good for business -- only .2% of customers who have installed their own fiber switch to another service provider.

While this sounds good, there are questions. Who owns the fiber, Lyse Tele or the home owner? Who controls the fiber -- can their customers reach competing services or are they locked in to Lyse Tele?

I have seen estimates that it costs Verizon about $1,000 to connect a home. I would gladly pay that if it meant I owned and controlled the fiber and could get connectivity from competing ISPs over it. I would consider it an investment -- increasing the value of my house -- not an expense.

I own my own sewer, water and gas lines and call a plumber when there is a problem. I would be happy to own my fiber.

Four excellent videos on municipal fiber networking

We have discussed infrastructure ownership options, and Benoit Felten has posted four excellent videos on municipal fiber networking. Three are interviews of people who gave talks at the recent Freedom to Connect Conference: Terry Huval, Tim Nulty and Bill St Arnaud. The fourth is a presentation given by Felten in New Zealand last month.

1. Terry Huval, Director of the Lafayette, Louisiana Utility Service fiber to the home project.

Huval discusses the motivation and business model that gave rise to the project, their legal battles (3 years of fighting until the State Supreme Court ruled unanimously in their favor), the services they offer, and the applications they will be offering in the future.

2. Tim Nulty, Project Director, East Central Vermont Community Fiber Network (ECFiber).

ECFiber plans to build rural fiber connectivity to 21,866 households in and around 22 Vermont towns. (So far 21% of those households have pre-registered). Nulty presents data on costs and revenue, and concludes that the network is a viable business. They will offer both retail and wholesale Internet service over the network, and he explains why a pure wholesale network like that in Stockholm makes sense in Europe, but would be defeated by the incumbents in the US.

3. Bill St Arnaud, Chief Research Officer at CANARIE, Canada's research network with a mandate to develop next generation networks, applications and services.

St Arnaud describes the "G-commerce" model, which combines connectivity with energy savings and pollution reduction. Installation of fiber to the home will be financed by a 1-2 cent per kilowatt hour increase in electric bills and energy-cap savings from reduced power consumption as high speed communication substitutes for transportation.

4. Benoit Felten, Senior Analyst, Yankee Group.

Felten gives examples from France and the Netherlands showing that network sharing is profitable even for incumbent ISPs. Take-up rate is more important than average revenue per customer, and the fastest way to convert 100% of the population to fiber is sharing it among service providers. He outlines and presents examples of several business models ranging from passive sharing of access and rights of way to offering retail service.

Why is connectivty in Stockholm so much faster and cheaper than in US cities?

We've been discussing US broadband policy and the stimulus package, and this table shows the cost of fiber-based, residential Internet service in several cities (Brough Turner provided the European and Asian data):

Where?	Monthly cost	Uplink (Mbps)	Downlink (Mbps)
Stockholm	$11	100	100
Seoul	$24	100	100
Hong Kong	$35	100	100
Tokyo	$61	100	100
Amsterdam	$127	100	100
Lafayette, LA, Municipal	$58	50	50
Lafayette, LA, Cox Cable	$140	5	50
US, where available, Verizon	$145	20	50

Can we explain the large speed and cost differences?

The Cox Cable offering in Lafayette, Louisiana seems to be the worst deal. It is the slowest and only five dollars a month less than the Verizon network. The municipal network in the same city is faster and cheaper. The Cox network reaches more neighborhoods than the municipal network, and they are forced to compete with temporary sale prices.

Stockholm is at the other extreme. They have a municipal network that reaches every block in the city. Unlike Lafayette, they do not offer consumer service over their fiber, but lease network access to anyone who would like to offer service. The Internet service providers, including incumbent telephone and cable companies, compete on an equal footing.

As a result, there are many competing service providers in Stockholm, and, as Turner points out, the city owns the expensive, long-life assets like fiber, rights of way, conduit, and tunnels, and the service providers own the electronic equipment that is relatively cheap and is upgraded frequently as technology improves.

Many factors determine the cost of Internet connectivity, but the ownership model is significant, and it seems the Stockholm model is superior to those in the US.

Note that analysts at the OECD also endorse the Stockholm ownership model, writing that:

Municipal networks can play an important role in enhancing competition in fibre networks. If these develop, governments should encourage them to be open networks, that is providing dark fibre to service providers rather than becoming themselves service providers. Nor should the existence of a municipal network providing dark fibre mean that investment in other fibre networks in that municipality should be prevented.

Should some of our broadband stimulus funds be used for Stockholm-style municipal networks?

Click here for a paper with more on this topic.

Click here for a PowerPoint presentation on this topic.

-----
Update 11/10/2014

Stockholm reports 19 years of financial and user success. The Stokab report should be required reading for all local government officials.

Some neat bookmarklets

Would you like to convert a Web page to a PDF file? One way to do that is to install a browser add-on from Pdfdownload.org. The add-on lets you choose to have the converted page emailed, opened, or saved on your hard drive.

Or, instead of using the add-on, you could use the Pdfdownload.org bookmarklet. The bookmarklet does not require installation, works with most browsers, and will not slow your browser down or consume memory. It can even be stored on the Internet for access from any computer. However, it lacks features of the add-on -- it does not offer the option of emailing or saving the PDF file; it just opens it.

Bookmarklets are similar to add-ons, but they may have fewer features. They are small JavaScripts that execute when you click on their icons in the bookmark toolbar. Here is the code of a simple bookmarklet that re-sizes the browser window to 800 by 1,000 pixels (which would be useful if you were preparing a series of same-size screen shots):

javascript:window.resizeTo(800,1000);window.moveTo(0,0);

The Pdfdownload.org script is a bit longer because it calls the PDF conversion service using their API:

javascript:void(window.open('http://www.pdfdownload.org/web2pdf/Default.aspx?left=0&right=0&top=0&bottom=0&page=0&cURL='+document.location.href));

There are tons of useful bookmarklets, and you can see a list of some of the best at this excellent Digital Inspiration post. If you start using bookmarklets, you will also like this post.

What is your favorite browser add-on? What is your favorite bookmarklet?

Three extremely simple Web services

The first day of class, we ask "what is a network-based application?"

Examples provide a partial answer -- a starting point. We do not want to be bogged down in details, so we want very simple examples. Here are three:

Eggtimer is a simple application where the program is stored on the network and downloaded each time it is used. Go to http://e.ggtimer.com/5 for a five-second timer or go to the home page for other options.

http://drop.io/ demonstrates an application in which the program and data are both on the network. With a couple of clicks, users can upload and share text, recordings, images, etc. Here is an example.

(Drop.io is surprisingly useful for such a simple application).

Aviary screen capture is a third example. You can capture and then edit any page on the Web by simply typing aviary.com/ in front of the http:// in the page URL.

For example, to capture the YouTube home page, you would enter:

aviary.com/http://www.youtube.com

Aviary automatically downloads the captured page image and an image editing program written in Javascript. Since Web clients contain Javascript interpreters, you can edit the page image when it gets to your computer.

Do you know of other network-based applications that are as simple to use and demonstrate as these?

AT&T mobile Internet access -- $480 per gigabyte when over the cap

We have seen that the price of sending a cell phone text message far exceeds the cost. What about mobile data plans?

An Oklahoma City woman is suing AT&T because her first month phone bill was over $5,000.00. She had subscribed to a $60 per month AT&T data plan that was capped at 5 gigabytes. She exceeded the cap the first month, and felt she had been deceived when the bill arrived.

The AT&T plan states that if one exceeds the 5 gigabyte cap, they are charged "$.00048 per kilobyte."

Would she have exceeded the cap if the price had been quoted as "$480 per gigabyte" or, better yet, "$350 to download a CD" or "$2,500 to download a DVD?"

Why does AT&T charge $12 for each of the first five gigabytes and $480 for the sixth? Do you think she should have to pay the bill?

The FCC national broadband plan -- can the Internet help?

The American Recovery and Reinvestment Act of 2009 directs the FCC to develop a national broadband plan within one year. This is reminiscent of the energy policy task force President Bush established during his second week in office. Bush's task force operated in secret, but President Obama has promised us an open and transparent administration, using the Internet as one tool.

It would be a shame if large, incumbent Internet service providers -- telephone and television companies -- dominated the FCC broadband planning process the way large energy companies dominated the formulation of the Bush energy policy.

Before deploying next generation access networks, we should consider new technologies and ownership and business models, but that will not happen if the incumbents write the plan.

Can the Internet be used to open the FCC planning process, to take it beyond the "beltway" and its lobbyists? The administration has established the Recovery.gov Web site to inform us on stimulus spending. How might the FCC use the Internet to open its broadband planning process?

Stimuluswatch.org -- fast, cheap, ad-hoc development on the Intenet platform

We have seen that application development and delivery platforms have evolved over the years, and have discussed the ease of building complex applications on the Internet platform.

Stimuluswatch.org provides an excellent example. It was conceived of in a blog post by Jerry Brito, and volunteer programmers, who did not know each other, responded. They built the site using data from another site and existing Web services.

Brito posted his idea on December 11 2008, and the site launched February 2 2009. You can read more on this ad-hoc, decentralized development project here.

Recovery.gov is Web 1.0; Stimuluswatch.org is Web 2.0

The President has signed The American Recovery and Reinvestment Act of 2009, which includes $7.2 billion for broadband access. You can see a brief summary of the bill here.

As we have seen, the Obama administration hopes to use the Internet for transparent, two-way communication with the public. To this end, they have launched the Recovery.gov Web site, which will be continuously updated, telling us "how, when and where" the recovery funds are spent.

As of today, Recovery.gov is definitely a Web 1.0 site -- it summarizes the Recovery Act, requests comments using an email form, and asks us to check back frequently for data on spending. They don't even have RSS feeds.

Contrast that with Stimuluswatch.org, a Web 2.0 site. Stimuluswatch began by importing a database of "shovel ready" projects that was posted by the US Conference of Mayors. Users can search the database by city, keyword and project type, and view the project descriptions and estimated cost and number of jobs created.

But, the main purpose of the project database is to organize user input. Users who are familiar with a particular project can drill down to the project page and:

• Make neutral, factual changes by editing its wiki page
• State opinions and debate the value of the project by posting comments
• Share comments with friends on Facebook
• Vote on whether or not the project is critical

(Check this video for a more complete description of navigation and user input).

I checked my city, Los Angeles, and the results were sobering. The database includes 321 projects with a total estimated cost of $7.3 billion, creating and estimated 82,341 jobs. Of these, only 11 projects had more yes the project is critical than no votes, and they accounted for only 3% of the cost of all proposed projects. Obviously citizen votes are only one factor to consider in selecting projects, but this indicates that we need to be selective if we are to avoid "porkwatch.org."

It sounds as though the administration plans to use Recovery.gov to let us know what they have done, but Stimuluswatch.org could be used in deciding which projects should be funded. That could be done by extending the database to include requests for Federal grant applications that are actually submitted now that the stimulus package has been approved.

Stimuluswatch.org was conceived of and built in under two months -- the administration should be working with them.

100 YouTube tools and resources

If you make or watch YouTube videos, check out this YouTube toolbox . With over 100 tools and resources, there has to one you like.

If you particularly like one of these (or another that is not listed), let us know.

Berkeley report on obstacles to cloud computing

We discuss the trend toward Internet services as network speed, storage capacity and reliability improve.

While the case for running applications on the Internet -- "cloud computing" -- is improving relative to running them in house, there are also obstacles.

A UC Berkeley research team is studying and trying to overcome cloud computing obstacles. Their initial report is summarized here, and the full report is here.

You can write documents with a program like Microsoft Word on your PC or you can use a service like Google Docs on the Internet. What are the advantages of Word over Google Docs? What are the disadvantages?

Govtrack.us -- detailed information on legislation and legislators

We have talked about the impact of the Internet on elections and politics.

Govtrack.us is a site for those interested in following the actions of congress and individual Senators and Congress people.

For example, here is a page on the economic stimulus bill and all of its ammendments. There are links to the votes of each representative on the bill and each amendment, their floor speeches, etc.

Another page displays a side-by-side comparison of the House and Senate versions (a lot of difference).

Users can also use news feeds to track changes to a given bill or follow the actions of a given representative.

Would you come to this page to research a bill or the record of your representative? Do you feel this sort of information will be used by voters? By journalists and political analysts?

Learn Web development and Google mashups with dynamic examples

Many Web sites invite mashups by providing application program interfaces (API) to their services.

The API for each service is different, and often poorly documented. They generally require only basic programming skills, but figuring them out can be time consuming.

Google has made it a lot easier to learn their APIs by creating the Ajax API Playground, a very cool site for people who want to create mashups with Google Maps, Search, Feeds, Calendar, Visualization, Language, Blogger, Libraries or Earth.

The API Playground has over 170 examples, but they are not static, they are dynamic. Here is one of the examples -- the code is shown in the top pane and the resultant map below.

If f you change the value of the variable address, and click Run, the map will shift to the new address. Try it yourself.

The W3schools Web site offers similar dynamic examples of HTML, JavaScript and other Web development tools. Here is a simple HTML example.

Do you find these dynamic examples useful? Do you know of other dynamic example sites?

Interview of Nicholas Carr

We have seen that the Internet has made application development faster, cheaper and easier.

Well known author Nicholas Carr elaborates upon this point in his book The Big Switch, in which he predicts a move from entreprise data centers to network computing. Carr recently outlined his thesis in an 8-minute interview on National Public Radio. If you are in a rush, you can listen to this one minute excerpt, where he talks about how easy it is to establish a complex Web site.

If you like what you hear, you might subscribe to Carr's blog.

What are some of the key advantages and disadvantages of moving applications from an organization's data center to the Internet?

Let's slow down on broadband stimulus in order to consider ownership alternatives

The economic stimulus package is moving through congress. Republicans are fighting it on the grounds that more tax cuts are needed and we need more time to make sensible investments. I agree that we need to slow down on broadband stimulus to consider ownership alternatives. Here is the "elevator ride" pitch:

• The current strategy of privatization with hope for competition under independent regulation has failed in many developed and developing nations. In the US, regulators have been unable to create competition and our infrastructure has suffered.
  
• The large broadband incumbents have benefited from public subsidy, have failed to live up to commitments, and have used their power to defeat attempts to create competition
  
• The US has little fiber in the access network today, but will have fiber to all urban and many rural homes and buildings in the long run. The question is not whether we are going to deploy new infrastructure; the question is “who will own it?”
  
• We should take the time to evaluate decentralized alternatives to near-total ownership by the incumbents. Local governments, cooperatives, small ISPs, and home and building owners might own parts of our next generation infrastructure.
  
• This evaluation can be fast and cheap. The work of the National Science Foundation in designing and creating NSFNet and connecting universities, colleges and foreign networks provides an excellent example of a small government staff calling on experts from academia and industry to design a network and a strategy for deploying it, followed by procurement via competitive bid.
  
• We need immediate economic stimulus, but that can come from tax cuts and investment in many sectors as well as broadband.
  
• Nobel economist Paul Krugman acknowledges the need for rapid stimulus, but in this article he says we should downplay the “jump start” metaphor and focus on job creation through infrastructure investment over the next four plus years.
  
• We will be living with the fiber and high-speed wireless infrastructure we build today for many decades. We will also be living with its owners.

Click here for a paper with details on the above.

Click here for a PowerPoint presentation on the above.

Broadband policy -- how can we catch up?

The US government funded the development of the Internet, but our infrastructure has fallen well behind many other developed nations. We have little fiber in our access networks, and have failed to establish competitive Internet-service markets.

President Obama has stated “As we renew our schools and highways, we'll also renew our information super highway,” and set a goal of deploying next-generation broadband:

(To) work towards true broadband in every community in America through a combination of reform of the Universal Service Fund, better use of the nation's wireless spectrum, promotion of next-generation facilities, technologies and applications, and new tax and loan incentives. America should lead the world in broadband penetration and Internet access.

Obama intends to involve the government more heavily in planning, subsidizing and procuring Internet infrastructure.

If our only concern were rapid economic stimulus, we could subsidize today's telephone and cable companies, but they have not served us well. We need to consider long run technology and the structure of the industry as well as quick stimulus. The infrastructure we install today will be with us for decades.

For more on this topic see this article or this PowerPoint presentation.

Rating and reputation in scholarly publication

We discuss systems for rating and reputation. Such systems are common in e-commerce and social networks, and we are beginning to see them in scholarly publishing.

I have posted a number of the articles I've written on my Web site, but do not track downloads or citations. Today I received an email from a journal stating:

As a service to our authors, we are pleased to provide you with a monthly report tracking readership for your article "A Framework for Assessing the Global Diffusion of the Internet":

9 full-text downloads between 2008-12-02 and 2009-01-02 29 full-text downloads since date of posting (2008-09-03).

This count came from a service of Berkeley Electronic Press, and you can read about their methodology here.

Citations are also used as a measure of the value of an academic paper, and Google Scholar provides citation counts. If I go to Google Scholar, I see that the above article has been cited by others 77 times.

The Association for Computing Machinery (ACM) is the most important professional society in my field, and they also report scholarly metrics for articles in their journals and conference proceedings. You can query the ACM database and see the number of times an article has been cited by other ACM authors and the number of downloads during the past six weeks and during the past year.

These are "walled gardens." For example, ACM does not keep records on the article mentioned above because it was not in one of their publications -- they only record downloads and citations of ACM publications by ACM members. Still, they are valuable first steps in using electronic reputation systems to assess scholarly contributions.

Would you consider the rating of scholarly writing in deciding whether to take a class from a particular professor? Would you consider the student-contributed ratings at a site like Rate my professors?