Communications

Cutting telecom taxes increases total government revenue

A major obstacle to the spread of new telecom services in under-developed countries is government dependence on revenues from the existing telecom monopoly.  ITU and World Bank advisers promise that if you remove restrictions and reduce taxes, the growing telecom industry will yield more total tax revenue while also fostering economic growth (thus increasing general tax revenues).  But that's a big risk to take.  How do you pay government employee salaries while waiting for the benefits of growth to kick in?

This article in the Pakistani Business Recorder gives specific tax rates, teledensity growth figures and resulting tax revenues which may help more backward governments to take the leap.

	Activation Charges	Subscriber Growth	GST/Excise Revenues	Activation Revenues
2003-04	Rs. 2000		Rs.   5.1B	Rs.   4.1B
2004-05	Rs. 1000	154%	Rs.   9.9B	Rs.   7.6B
2005-06	Rs.   500	169%	Rs. 18.8B	Rs. 11.4B

Certainly Pakistan has had no lag in revenue gains as they have cut activation fees.  Considering all the other benefits that telecom is bringing to Pakistan, it's too bad they are only discussing a further cut to RS. 250.  The best approach for poor people in Pakistan would be to completely eliminate activation fees.  Either way, the government is certain to gain in total revenue.

Foreign Investment in China

It is now well established that investments in telecom are the most productive of any investments a developing economy can attract — better at increasing per-capita GDP than investments in roads, electricity or even education.  So I'm always interested in better understanding what a developing country can do to attrack investment in telecom.  The answer is complex, but a few things are clear.  Regarding telecom directly:  increase competition, reduce restrictions on foreign direct investment (FDI) and reduce taxes on telecom, including less direct costs like spectrum fees and service activation taxes.  But also an issue is a business climate that recognizes property rights and minimizes corruption.  Just how important is that?

I've seen good data to support the benefits of competition, the good results of reducing restrictions on foreign investment (i.e. on allowing foreign ownership of telecom companies) and the positive gains from reducing telecom taxes.  I have been less confident about the issue of good government as China appears to attract enormous flows of FDI despite its government getting consistently bad marks in the western press.  What's actually going on?

My thanks to Ajay Shah who pointed out:

The World Bank has produced a draft volume Dancing with Giants: China, India and the global economy, edited by L. Alan Winters and Shahid Yusuf. Six areas are covered: global industrial geography, competing with giants, international financial integration, energy and emissions, regional variations in growth within the giants, and governance.

Groundwork towards this was done in the form of 21 background papers. Drafts of the book, and the 21 background papers, have been placed on the web. It is good work.

In particular, the working paper Does ‘Good Government’ Draw Foreign Capital? Explaining China’s Exceptional FDI Inflow, by Joseph P.H. Fan, Randall Morck, Lixin Colin Xu, and Bernard Yeung, shows that FDI in China correlates extremely well with respect for property rights.

Their conclusion in response to the question:  Is too much FDI flowing into China?

Probably not.

We have shown that, within China, inward FDI flows disproportionately into provinces with less corrupt governments and governments that better protect private property rights.  We estimate a cross-country FDI model, without China, that explains inward FDI using measures of the strength of constraints on executive power as well as more general measures of government quality and track record in fostering growth and a set of standard controls.  This model predicts FDI flow into China with prediction errors similar to those for other countries with similar levels of institutional development...

There is indeed an enormous FDI flow into China, but given China’s size and growth track record, this flow is not far from what would be expected for a country at China’s level of institutional development. 

So much as we may dislike government repression, it appears the critical issue for raising people's standard of living is protection for property rights and minimizing government corruption.

In any event, there's a wealth of economic information in these working papers.  I'm grateful to Ajay for pointing them out.

Connect 2006: Disruption - my closing presentation

There are several items from today's sessions to write about, but not enough time right now…  So let me talk about my closing presentation, just completed and apparently well received.  Here is a copy:

Disruption — Where Do We Go From Here?

I was “assigned” the title and choose to use it to talk about telecom economics, suggest when mobile "dumb pipes" may emerge, and then examine what we can learn from the economics of successful Internet applications that might be relevant to the evolving mobile industry.

In high tech, the word “disruption” was popularized by Clayton Christensen’s 1997 book The Innovator’s Dilemma, but for today’s discussion I used the material I wrote about in this earlier blog post. 

I tied the ideas to mobile broadband which, today, is still limited by technology thus vertical integration (and walled gardens) remain the norm.  Once 3G or 4G technology is able to provide “enough” mobile broadband capacity, there will be strong pressure to offer “dumb pipes.”  Unlike the fixed access market where regulation, legislation and lobbying dominate, mobile is a competitive industry, so once the technology can support adequate dumb pipes, they are bound to emerge.

People who say “the operators will never allow themselves to be marginalized” don’t understand that this decision is not under management control.  If mobile VoIP begins to significantly impact mobile revenues, those operators who don’t find a way to split themselves up, will be restructured by corporate raiders, just as the steel industry, the auto industry, etc. have been restructured in recent decades.

There's a lot more in the presentation which I'd be happy to discuss if anyone inquires.

Connect 2006: Mobile data problems and a solution

Andrew Budd was one of the more interesting panelists at Connect 2006 this morning so it was no surprise when his presentation in this afternoon’s session, Explore the Mobile Content Ecosystem, was excellent.  Besides providing a good introduction to MBlox and a good history of premium SMS, Andrew discussed the key problem with content delivery over mobile data today, and proposed a simple solution.

The operators’ biggest contribution to the mobile content value chain is in delivery and billing, not in marketing content in a walled garden.  But 3rd party content providers who wish to sell content to mobile consumers have a problem – unexpected data charges.  If you have a major brand to protect, you can’t afford to sell a € 3 piece of content if there is a chance the mobile consumer will also have to pay € 20 in data charges.

So far, most operators have taken the position that mobile subscribers should sign up for unlimited data plans.  If they don't and they get hit with a large bill, this will be a good way to drive them to sign up for an unlimited data plan.  That’s nuts!  You don’t win subscriber loyalty with unexpected charges.  More importantly, 3rd parties with valuable brands don’t want to be associated with this kind of sticker shock sales tactic. There are 3rd party content offers where it says the price is € 5 if you have an unlimited data plan, but warns you NOT to buy unless you have such a plan.

Andrew’s simple suggestion is to offer wholesale data packages to content providers, so content providers can bundle data charges in with the content and thus be in a position to make offers to their customers that have no hidden snares.

Given operator benefits from differential pricing and given the scalability advantages of wholesaling, this is an eminently reasonable proposal.  I wonder how long it will take for a major operator to adopt this approach?

Connect 2006: Mobile uses of Avatars

The second session at Connect 2006, Innovative Multimedia Applications, had several interesting applications but the most compelling for me was an application developed by Kallideas SpA (in English) in Milan in partnership with Rietek SpA.  The application is called Virtual Girlfriend and is best illustrated in this video.

Boys 'n Girls
Uploaded by Kallideas

Kallideas is the application provider using their avatar technology.  Reitek provides the mobile video technology to get this running on mobile networks.  As an application that appeals to individuals, especially to teenagers, this is a natural for your 3G mobile – frequently a teen has only shared access to a family PC, but sole possession of their mobile phone.

Connect 2006: Operator value is in access & billing

The first panel at Connect 2006, entitled Industry Overview: The Experts’ Perspective, was interesting as the mix of speakers resulted in some genuine controversy. 

Bob Schechter moderated and kicked off a round of panelist introductions with the question “What is the killer app?”  Kari Lahtinen, Business Development Manager at Finnish operator Elisa said IPTV and elaborated by explaining Elisa’s efforts to deliver a triple play bundle.  Philip Kelley from the CTO’s office at Alcatel Mobile Communications Group said Mobile TV.  Matthias Fischer, Technology Manager at Vodafone Group R&D in Germany suggested there was no one killer application and no simple value chain, but rather a complex value web.  If Kari and Philip assumed a walled garden business model, Matthias obviously expected a more complex business environment – a partially walled garden (my words not his).

Andrew Budd, Chairman of MBlox and Vice Chairman of the Mobile Entertainment Forum had the best answer – there are millions of pieces of content and millions of niche applications.  Mobile operators don’t have the marketing expertise or the ability to scale such a marketing effort.  Voice telephony was one application.  SMS is one application.  As individual applications, they easily scale to millions of subscribers.  Content involves millions of individual brands – Tom Cruise is a brand –  Madonna is a brand.  Marketing this content doesn’t scale.  Similarly, the millions of niche (or mainstream) applications that have emerged on the web, require millions of separate marketing efforts – this doesn’t scale either.  Thus Andrew argues that operators should focus on providing access and billing services and otherwise stay out of the way.  Take lower gross margins but achieve higher net margins.  In support of this approach, he showed a graph with data points for EU mobile operators that related the percentage of content fees they pass through to content providers with their success in non-SMS data revenues.

It’s pretty clear – the more you pass through to content or application providers, the more total money you make.

The most interesting question from the audience was from Dean Bubley of Disruptive Wireless who pushed Andrew on whether there was even a compelling reason to expect mobile operators to retain a position in billing services for mobile content.  Dean's position was that the Internet model would prevail.  Andrew's counter was that mobile premium billing provides the business model that garage-shop startups need to make money from day one (versus the Internet where costs are near zero but revenue can be similarly near zero).  No resolution, but it's a more interesting discussion than the pure focus on bundles services in walled gardens!

Prague

I'm in Prague for Connect 2006 which starts tomorrow morning, but I've had a chance to tour the city on Sunday afternoon and Monday.  Miscellaneous photos here.  Like Germany they take their beer seriously.  Here is beer delivery by hose, something I've never seen in the US!

Off to Prague for Connect 2006

My wife and I are off to Prague.  I'm attending Connect 2006 on Tuesday & Wednesday next week and we're doing some sightseeing in Prague & the surrounding area for another week. Conference insights should show up here on Tuesday & Wednesday.  Other info - you'll have to buy me a beer or check out future additions to my photos on Flickr.

Increased presence of test equipment and services at VON

I got some more time to tour the exhibits at VON during the middle of the day on Thursday, i.e. shortly before the exhibit hall was due to close.  One thing that struck me was the increased presence of the test industry.  I noticed more than a dozen booths before I started paying attention.  Along the way, I talked to a few of the companies and then made it a point to stop by the booths of Empirix and NetHawk, both customers of NMS Communications, i.e. places where I knew people.

Without revealing who said what, and reading between the lines of what I heard, there are several things going on.  It appears major equipment providers are beginning to buy equipment for their internal IMS labs.  They've got different groups building different IMS systems and it's time to figure out if any of it works together.  It also appears there are a few real UMA deployments which are causing operators to buy test equipment.  Finally, generic VoIP deplyments are increasing.  All in all, the test guys seemed upbeat.

Here's Mohamed Khan in the NetHawk booth.  Their pitch emphasized end-to-end call signaling while their signage focused on UMA and IMS.

And here's the Empirix booth - somewhat larger and plenty of orange shirts.  The focus was VoIP & IMS.

Earlier I commented that traditional VoIP has gone mainstream and mobile VoIP is still in the future (although there were some interesting items on display), so naturally Jeff Pulver's attention has turned to what's hot right now, i.e. video over the net. 

But where "mainstream" sounds boring to me, it's evidently good for the test industry.

Chinese clone YouTube

James Seng points out 6rooms, a Chinese clone of YouTube.  Surfing further on the subject, the best information I could find was this post from Brian Curtis, especially his further discussion in the comments area.

Actually, www.yoqoo.com was recently brought to my attention, as was www.56.com. According to Alexa, they rank:

1) 56.com
2) 6rooms.com
3) yoqoo.com
4) toodou.com (dead last)

Drives home the point that being first to market and well funded is not good enough…

Each time I’ve checked, YouTube was accesible in China. However, the censors can run hot and cold on certain sites. I am not in China at the moment to verify current status.

So a Chinese video market is emerging.  6rooms appears to have a bit over 100K videos up as of today - clearly a tiny fraction of YouTube's inventory.  On the other hand, the Chinese market is large and they are actually ahead on some video applications, like peer-to-peer TV.  So it will be interesting to see how this plays out over the next 12-18 months.

What are the business models for Internet Video

This morning I attended a panel "Advertising Models for Video on the Net" at the Video on the Net conference (co-located with VON).  It was interesting, somewhat educational, but not startling.  Worse, they were talking about CPMs, 15 second ads and product placements.  There was some mention Google Adsense, but no real focus on the long tail or on radically new business models for Internet video.  I wasn't left with the sense these people understand what it's going to take.  But then I don't have an answer either, so who am I to speak.

For the next Video on the Net, I am going to suggest Jeff recruit Umair Haque to speak.  Umair hasn't proposed a specific business model, but he does have a good analysis of where the new business model will lie.

The moral of the story is simple. You should see Heavy and YouTube as opposites in strategic error. Heavy doesn't create enough value consistently enough to be able to exert enough pressure to capture a significant share. YouTube, on the other hand, is creating a great deal of value - but also can't exert enough pressure to capture a significant share.

Let me put it more simply: focusing on being a channel or being a platform is the gap between value capture in the next media economy. It is going to be a kind of mournful refrain, an error players make over and over again.

Instead, capturing value for consumer-focused players depends critically on being able to plug brands and ads into the social and cultural structures of interaction itself, to make return on attention hyperefficient - like Myspace is doing.

Massachusetts Broadband Roundtable at VON

Don Means invited me to an interesting luncheon roundtable on broadband access yesterday.  Present were representatives of municipal broadband initiatives in Boston, Malden, Cambridge; representatives of Massachusetts companies that have an strong interests in broadband access, including Sonus and NMS, i.e. me; other companies (Cisco!); folks from the Massachusetts government and Congressman Ed Markey who is the ranking democrat on the House Telecommunications and Internet Subcommittee.

Don has been promoting local responsibility to accelerate broadband deployment through the Community Telestructure Initiative.  He's already got state house sponsorship in California and Virgina.

In a group like this, I have two key issues. 

First, I'm for competition in government –  edge-based policy – move the decision making out to the municipal or neighborhood level. As Justice Brandeis commented in NEW STATE ICE CO. v. LIEBMANN, 285 U.S. 262 (1932) :

It is one of the happy incidents of the federal system that a single courageous state may, if its citizens choose, serve as a laboratory; and try novel social and economic experiments without risk to the rest of the country.

Second, I'm for fostering competition at the lowest layer possible – Layer Zero competition.  Rather than a municipality becoming an ISP, focus on delivering access to the rights-of-way or dark fiber to all comers.

In the end, the meeting ran out of time so I didn't press either of my issues.

My key take-aways were:

1.  I met Deborah Shufrin, Director of the Massachusetts Dept or Business & Technology, who clearly understands that the core issue with any broadband is access to rights-of-way.  It's refreshing to meet someone who gets it.

2.  I got to know more about Ed Markey.  I've heard him speak many times, but this was the first time I got to interact with him in a small group.  Not surprisingly, for someone who's served in the US House of Representatives for 30+ years, Ed Markey is the consummate politician.  Most of us don't understand the practical strategies required to get any legislation enacted.  I ended the luncheon meeting with a much better respect for what Ed Markey actually understands about the issues – of course that may just prove he's the consummate politician.  :-)

Not surprisingly, he took the time to tour the exhibit hall and get photographed in the booths of Massachusetts companies.

Congressman Ed Markey visits the NMS Booth at VON.

Reflecting on Tuesday at VON

I'm not as efficient as Alec Saunders, who did an excellent job blogging Jeff's keynote – apparently in real time.  I have accumulated a half dozen ideas for specific posts, but it may take a week to spool out the text.  For now, let me reflect on my first day at VON.

What's new is the focus on video over IP.  Just in time!  The first stage of VoIP has gone mainstream.  It's acknowledged by the major carriers and incorporated into the major equipment providers' products.  Whether the products use IMS or peer-to-peer SIP, these are the guys with the distribution muscle.

Looked at another way, VoIP is just a technology.  It allows for more innovation than was possible in the TDM world, but after ten years, the biggest VoIP success stories are arbitrage plays with extra features.  Indeed, the big revolution in telecom has been the advent of mobile phones but, so far, there is very little VoIP in the world of mobile. And, at a guess, the mobile VoIP breakout is still 3-5 years in the future. 

Meanwhile video over IP is happening now!  The reason video is now, and mobile VoIP still a few years off, is simple.  The video that matters is one-way – recorded video clips, live broadcasts or person-to-person see-what-I-see video.  That means you can buffer playback to overcome QoS issues.  Mobile VoIP still suffers from a lack of upstream data capacity (and likely will until widespread deployment of HSUPA).

From the NMS point of view, video has been a major focus for 4+ years.  With our telephony background, we started with 3G-324M video telephony (a two-way capability).  We rapidly found the real use for the technology was either "see-what-I-see" (only one direction matters) or rapid switching between alternate sources of video content (only one direction is relevant).  As has been true of VoIP, you need to support both IP & TDM.  We architected our video products to be IP-based internally, even when the external connection is 3G-324M (video over circuit-switched data).  Early this year we acquired Openera which, among other things, added SIP-based video sharing capabilities to our product line.

Most of the video presentations at VON are about creating and finding video content that people can access over IP, versus our focus on video delivery, but from a user point of view, we're on the same page.

Applications, not platforms

I spent some time in the exhibit hall at VON this afternoon.  One thing that struck me was the large number of companies offering service platforms.  These ranged from BEA to Lucent to Ubiquity, IP Unity, Personeta.  Operators don't buy platforms.  They may say they want platforms, but they buy applications, and typically only applications that have been shown to be profitable somewhere else.  It's one thing if these companies were at VON trying to recruit developers, but most of them are selling a platform for which they are explicitly or implicitly the sole source of applications.

Personeta ( a customer of NMS) was better than most.  Their signage included applications as well as statements about their platform, but you had to watch their presentation to actually see the rich set of applications they have already delivered to a dozen or more operators around the world.

Why not lead with your applications?  That's what operators want.

And note - Personeta was one of the better such booths!

NMS booth at VON

Here are some pictures taken at 10am this morning just before the exhibit hall opened.

There are four display stands showing: 

• Vision Family Overview - industry's broadest range of media solutions:  video, audio and signaling servers
• Vision VXML Server - voice and/or video media controlled with high-level scripts
• IP Media Server on an ATCA Blade (the MG 7000A)
• CG 6060 & CG 6565 VoIP & TDM components with rich development environment

And here's the crew, ready to help the crowds that probably won't materialize until the current Industry Perspective presentation concludes.

Happy Birthday Jeff

I'm at Fall VON 2006 in Boston and today (September 12th) is Jeff's birthday.  In celebration, and recognizing Jeff's reputation for sartorial elegance, Andy has declared today Hawaiian Shirt Day at VON.

No New Competitors Emerge in US Spectrum Auction

The Economist has a current article (subscription) subtitled "America's wireless-spectrum auction avoids the pitfalls of Europe's."  But that misses the most important point.

It's true US operators are paying less for spectrum in the FCC's current Advanced Wireless Spectrum (AWS) auctions than EU operators paid for their 3G spectrum.  Here is the comparison from the Economist article:

On the other hand, the UK & German auctions were in 2000, at the height of the bubble.  In addition, EU spectrum is highly regulated – you can only run W-CDMA technology in this spectrum, it's the only spectrum authorized for 3G and there is no subsequent market to buy & sell this spectrum.  So existing mobile operators had to win in the auction or go out of business.

But this is not the important story for the AWS auctions. The issue here is no new competitors! 

The new spectrum is going to existing wireless players.  T-Mobile USA, Verizon Wireless and a Sprint/CableCo joint venture came out on top.  Some of the regional licenses are going to second tier players like MetroPCS and Leap Wireless, but nothing much is going to anyone who is not an incumbent mobile operator today.

Today's mobile Internet is a walled garden, in the US and in most of the world.  Our best hope for an open mobile Internet is the combination of improving technology and robust competition.  The technology roadmaps are clear – for UMTS (HSDPA, HSUPA), for CDMA 2000 (EVDO Rev A, Rev B) and for WiMAX.  The competitive situation is more worrisome, given recent consolidation (Sprint-Nextel; Cingular & ATT Wireless;  etc.).

The good news is T-Mobile USA appears to be in the game for the long haul, thus mobile competition remains fairly robust – much better than our fixed line duopoly.

More on layer violations in wireless

In late August I wrote on the technology-driven transition from vertical integration to modular industry value chains and how it applied to telephony in general.  A subsequent post on "layers violations" attempted to relate the concepts to mobile VoIP.

I have additional supporting links.  First, my August post alluded to work by Henning Schulzrinne with Andrea Forte and Sangho Shin at the 65th IETF meeting in March.  At that time, the only reference the abstract of a presentation at WitMeMo’06.  The full paper is now on the web.  It's 802.11 in the Large: Observations at an IETF meeting by Andrea Forte, Sangho Shin and Henning Schulzrinne.  As discussed in the previous post, this work points out the extent of problems with current 802.11 implementations, i.e. problems in today's products.

The question for mobile VoIP is whether fast handoff can be achieved at layer 3 (the IP layer) without a tight integration with layer 2 (the radio link).  Traditional mobile networks (GPRS, CDMA 2000 & UMTS) use vertical integration – what a network engineer would call a layers violation – in order to achieve high performance handoffs.  Obviously, WiFi networks (and other general purpose wireless networks) don't do that today.  We could just wait until wireless technology becomes much better – for example until dual radios cost no more than a single radio thus allowing IP session setup over a new radio link while existing IP traffic flows over the existing radio link.  But it could be years before that approach becomes low cost.

Or we could attempt to modularize the "layers violations." The latter is what the IRTF (Internet Research Task Force - the research arm of the IETF) MobOpts Research Group is discussing.

As background there already are Mobile IP protocols defined for both IPv4 and IPv6, but these don't provide handoffs that are fast enough for seamless voice telephony.  Quite a bit of work has been done on IP protocols for fast handoff of IP sessions.  For example, RFC 4068 Fast Handovers for Mobile IPv6, describes what to do at the IP layer,

without depending on specific link-layer features while allowing link-specific customizations.

Unfortunately, to achieve seamless voice telephony handoffs, you will require either two radios or some link-specific customization.  This is where the MobOpts Reseach Group (RG) proposal comes in.

MobOpts RG is discussing ways to standardize the needed link-specific customization – see IRTF draft Unified L2 Abstractions for L3-Driven Fast Handover.  In effect they propose standardizing the layers violation.  If adopted by radio link equipment vendors, this would modularized the value chain without waiting for extra technology to show up!

Of course this is a proposal, not an RFC much less a standard.  So any implementation is at least a few years off.  That's consistent with my view, expressed in my August 31st post, that widespread mobile VoIP will lag fixed line VoIP by at least five years, but it's coming!

Fall VON in Boston, Connect 2006 in Prague

Next week I'll be at Fall VON in Boston on Tuesday-Thursday where NMS is exhibiting (booth 775), I have a variety of meetings (email me if you'd like to meet) and I'm speaking on Wednesday afternoon's "Bloggers of the World" panel.

Then Saturday I'm off to Prague for a few days of tourism and, more importantly, the Connect 2006 conference on the 18th & 19.  Connect 2006 focuses on multimedia applications and the transition to the mobile Internet.  VoIP plays a part here, but it just an enabler.  The big story, now and for years to come, is mobile and the evolution of mobile communications.  From the speakers and the current attendee list, it looks like a good mix of innovators and large players, e.g. major operators like Vodafone, major equipment providers like Alcatel and leading trade associations like the Mobile Entertainment Forum.  Of course what I find fascinating is small companies with innovative ideas. Session details are here and registration info here.  As an added incentive, I'm closing the conference with a presentation:

Disruption! Where Do We Go From Here?
Evolving technology (bandwidth, storage, processing, batteries, etc.) enables disruption, but actual disruptions spring from new business strategies and new combinations of technology that act as platforms for innovation. What will drive Telecom 2.0? The Internet provides some ideas like harnessing collective intelligence, users as co-developers, lightweight programming tools, mashups (loosely coupled systems), and AJAX. We can also look to consumer trends in communities (Cyworld, MySpace), gaming (World of Warcraft), and music (iPod) for more ideas. But the interesting exercise comes from mapping these ideas onto telecom as today's operators migrate to IMS and 3G with (or without) WiFi, WiMAX, and one or more of the diverse approaches to mobile TV.

Impressive new judge on US patent court

Perhaps a bit orthogonal to my usual interests, but yesterday, the US Senate unanimously confirmed the appointment of Professor Kimberly Moore as a judge on the Court of Appeals for the Federal Circuit (CAFC) – the court that hears appeals of patent cases.

The news clip caught my attention as I'd recently read an interesting paper she co-authored with Mark Lemley, Ending Abuse of Patent Continuations back in 2004.  Also, the fact that the vote was unanimous.  Of course, while a tribute to Professor Moore's qualifications, the vote may also reflect on how little politicians care about patents.

In any event, it's nice to see a Federal appeals judge with advanced engineering degrees, and from MIT no less :-).  Response among attorney bloggers I respect has also been positive, e.g. this from Dennis Crouch:

My two-cents — Professor Moore is an excellent nominee.  She is extremely intelligent, and has a complete understanding of the legal issues faced by the CAFC.  You will find that her decisions will be fair and quick, and, I believe that her vast knowledge of the CAFC and its history will help her to craft opinions that reconcile some of the internal cracks that have surfaced the past few years.

Wikipedia - The social network is only part of the story

One of the wonderful phenomena of the past few years has been the emergence of Wikipedia as an incredible reference for just about anything.  If you've heard Wikipedia's founder Jimmy Wales speak (for example watch this 20 minute video from July 2005), you've also heard about Wikipedia's vibrant community, how 50% of the edits are done by just over 500 people and 75% of the edits are done by less than 1500 people.

It appears these statistics, while true, are very misleading!

Aaron Swartz has looked at who actually provides the most content and it's very different from who makes the edits.  First Aaron was just looking at specific articles, initially the article on Alan Alda.  But then he wrote a program to examine the Wikipedia archives and extract statistics.  He's written a very comprehensive article, but in summary:

If you just count edits, it appears the biggest contributors to the Alan Alda article (7 of the top 10) are registered users who (all but 2) have made thousands of edits to the site. Indeed, #4 has made over 7,000 edits while #7 has over 25,000. In other words, if you use Wales's methods, you get Wales's results: most of the content seems to be written by heavy editors.

But when you count letters, the picture dramatically changes: few of the contributors (2 out of the top 10) are even registered and most (6 out of the top 10) have made less than 25 edits to the entire site. In fact, #9 has made exactly one edit -- this one! With the more reasonable metric -- indeed, the one Wales himself said he planned to use in the next revision of his study -- the result completely reverses.
...

When you put it all together, the story become clear: an outsider makes one edit to add a chunk of information, then insiders make several edits tweaking and reformatting it. In addition, insiders rack up thousands of edits doing things like changing the name of a category across the entire site -- the kind of thing only insiders deeply care about. As a result, insiders account for the vast majority of the edits. But it's the outsiders who provide nearly all of the content.

Aaron's a smart guy.  As a teenager he co-authored RSS 1.0, worked on the W3C's RDF 1.0 Working Group and wrote RFC 3870.  Then he went to Stanford for a year before dropping out to become an entrepreneur.  I'm impressed he has the time to work on Wikipedia.  He's currently running for a seat on the Wikimedia Foundation Board.

Of course his article raises all sorts of governance issues for the Wikimedia Foundation.  Today, you have to have made over 400 edits in order to qualify to vote for members of the board of Wikimedia Foundation, i.e. only the editors can vote, not the folks who are actually contributing most of the content.  And not me whose contributions are very minor.  What's critical is that it remains easy and rewarding for stray individuals to add content.  I certainly hope the Wikipedia inner circle takes note of Aaron's data.

Portable fuel cell to recharge mobile phone

When I last wrote on this topic, both DoCoMo and KDDI had shown prototype cell phone batteries based on liquid methanol.  After the Dell battery recall and the new liquids prohibition on commercial aircraft, it's hard to see how anything like that could succeed.  But as this August 30th story and this July press release show, NTT DoCoMo is taking the long view and continuing to work on the technology.

Now their target is a portable recharger, not a battery replacement. 

Clearly a get-started approach, as few will want to carry yet another device, even if it's allowed on airplanes.

Other updates:  Perhaps in response to issues with methanol on airplanes (even before August 9th), the most recent technology appears to be based on a solid fuel and catalyst mixture which produces hydrogen when water is added.  The hydrogen is then combined with oxygen from the atmosphere using conventional polymer electrolyte membrane (PEM) fuel cell technology to produce electricity (and more water).

None of the press information or news reports say what the fuel-catalyst mixture is.  DoCoMo's partner is now Aquafairy, an eleven person company incorporated June 30, 2006, for which very little additional information is available, in English or Japanese.

I'm glad to see portable fuel cell efforts continue and I do expect this technology will power portable devices, someday.  But, at a guess, we'll be in the demo &/or prototype stage for a few more years.

Internet-scale database - Google's Bigtable

There's a very interesting preprint of a paper from Google Research due to be presented at OSDI '06 in November.  It's Bigtable: A Distributed Storage System for Structured Data by Chang et al. (PDF here). 

Bigtable is a database that runs on top of the Google File System.  I first heard of it in the May '06 interview of Google's Jeff Dean by O'Reilly's Radar. Like the databases supporting other major Internet applications, it's not a straight relational database.  This paper gives the details for those of you interested in such things.

It also includes the following gee whiz items:

• The Google web crawl creates two tables totally 850 terabytes of data. 
• Google Analytics apparently keeps all clicks, forever.  It's currently using two tables totally 220 terabytes.
  Hmm..., if they really keep everything forever, this could end up larger than their web crawl.
• The maps for Google Earth occupy a mere 70 terabytes.

Large Speech Corpora

I didn’t attend SpeechTek this year and I haven’t had much involvement with speech technology in the past 12 months, but this post from Google Research in August 2006 brings me back to the topic I was discussing in August 2005.  Google is offering researchers some interesting data.

We processed 1,011,582,453,213 words of running text and are publishing the counts for all 1,146,580,664 five-word sequences that appear at least 40 times. There are 13,653,070 unique words, after discarding words that appear less than 200 times.

Over one trillion word seuences!  Of course this is a text corpus — useful to an automatic speech recognizer as a ranked set of likely word sequences, but only to the extent people’s speech patterns match our written language…

What’s really needed is a large body of annotated speech recordings — millions of hours.  In my presentation at SpeechTek last year, I suggested one way such a speech corpus could be accumulated.  On a free video storage site, provide machine translations of the audio associated with user-created videos.  Offer users the ability to search for specific events in their videos based on words in the sound track and also offer them a simple interface to correct the machine transcription.  Useful search is the incentive, a human verified transcription is the byproduct.  I recognize these annotations would not be "professional", but it’s more productive for researchers to focus on automating quality measures for millions of amateur annotations, than to produce another 1000 hours of professional transcription.

When I prepared my 2005 presentation, I hadn’t heard of YouTube or Google Video — two sites that appear to exactly match what’s needed.  I doubt YouTube has the resources or can spare the attention to tackle this, but Google has the resources and the researchers — after all they did attract Kai-Fu Lee last year.

Money in dumb pipes?

In the preceding two posts, I've applied Christensen's view, of how technology based products and services are disrupted, to fixed and mobile telephony.  While 3rd party VoIP is not yet dominant in fixed telephony, it's made enormous progress and most people can see the writing on the wall.  But what happens to the lower layers as the value chain modularizes?

In the computer industry, we've seen all sorts of specialization.  But we've also seen companies, like Dell, focus on operations and make money in a layer that R&D-centric businesses avoid like the plague. 

Back in February, Craig Moffett and Amelia Wong of Bernstein Research wrote an interesting paper, The Dumb Pipe Paradox which was distributed to Bernstein Research clients (but unfortunately doesn't appear to be available on the web).  In March, Craig testified on this subject before the Senate Subcommittee on Communications.  At the time, the press reports I saw didn't adequately reflect what was in the original report, for example this from the San Francisco Chronicle:

... financial analyst Craig Moffett of Bernstein Research articulated the dilemma phone companies face as they plan these upgrades.

More recently this from Bill St. Arnaud prompted me to write to Craig Moffett and get a copy of the original report.  It's great stuff !!!

Quoting from Craig's original report:

Our conclusions regarding the economics of “dumb pipe” status are, we suspect, quite counter-intuitive (and starkly counter-consensus).  In short, the economics of such a scenario, as measured by purely financial metrics, are actually better than those of the business today.

• Revenues would fall significantly
• Operating expenses would fall even more significantly
• Operating cash flow would therefore fall more modestly, while operating cash flow margins would  actually expand
• Capital expenditures would fall sharply
• Free Cash Flow would actually increase
• Capital employed in the business would also fall sharply, yielding much higher ROI and ROE

To be sure, our analysis ignores potentially painful milestones on the way to such a “raw connectivity”  future, such as the stranding and subsequent write downs of a huge portion of Comcast’s current asset  base.  Growth prospects in this “end game” would likely also be lower than today’s (although not necessarily).  Nevertheless, our analysis leads to the surprising conclusion that a disintermediated business would, on most financial metrics, at least, actually be a better business than the current one. 

Behind these counter-intuitive findings is a simple, and perhaps more intuitive, truth.  In a high fixed cost business, it’s not such a bad thing to be the low cost provider.

Wow!  Dell's operations focus allows them to make money as the low cost provider of what others consider a commodity.  Similarly in telecom, there is money to be made being the low cost provider of dumb pipes.

Note:  The report discusses both Telcos and CableCos and concludes CableCos would win in the pipes business. But that is very much an artifact of the US regulatory environment.  Think what might be achieved if we had competitive access to first mile rights-of-way.