Friday, November 30, 2007

Put a Leopard in your tank

All but my fellow Mac bigots can skip the rest of this post. I upgraded my Mac to OS X 10.5 today and had a few observations.

One is that the branding of the software package says “Leopard” with “10.5” in smaller print. I didn’t get the sense that this was true with Tiger (10.4), so I guess Apple has concluded that using the code name to build buzz beforehand creates a brand that’s worth using after launch.

10.5 has brought some fawning reviews but also a few noting the key omissions (notably Classic). Today’s immediate problem was the loss of NetInfo Manager.

Without NetInfo Manager, I had to solve two problems:

In our printer driver days, we used to have multiple system configurations installed (it was much easier under OS 7.x-9.x). Still, my former coworkers thought I was nuts to set up multiple configurations on my laptop — in this case, two versions of OS X and one with OS 9.

It turns out it’s already paid for itself. When my 10.4 partition died 6 weeks ago en route to the airport, I switched my 5-year-old laptop back to OS 9.2.2 and then 10.2.8. Rather than drop everything to salvage the 10.4 partition, I decided to wait until I could do a clean install of 10.5, which turned out to be painless (other than the $70 purchase price).

Much as I like Apple’s stuff, they can’t make the hard disk more reliable. And this little episode had a good side-effect: I bought an up-to-date copy of Disk Warrior, which worked like a charm. Even my wife (Palomar’s then project manager and procurement officer) remembered how the product bailed us out of jams in the pre-OS X days. It’s a good insurance policy to have on hand.

Thursday, November 29, 2007

Walt says: wait for Kindle 2.0

Thursday the WSJ’s Walt Mossberg reviewed the Amazon Kindle in his main column. (15 years ago, when I was in the computer industry, this was the most influential computer column in the country).

His conclusions: Amazon knows how to sell books and has good partnerships, but doesn’t know how to make hardware.

Amazon has nailed the electronic-book shopping experience. But it has a lot to learn about designing electronic devices.
And, in addition to the dubious idea of charging for free online content (like magazines and blogs), it doesn’t work very well.

Microsoft and some other companies have done very well by selling a lousy 1.0 and a much better 2.0 or 3.0. (Remember, Windows wasn’t usable for anything until Windows 3.1, and the mass adoption started with 4.0 i.e. Windows 95).

Will Amazon get better? Will it stick it out (rather than give up)? I suppose if your business is selling dead tree information goods, you have to do anything in your power to prepare for the day (due to cost or convenience or environmental consciousness) that people stop wanting to kill trees. This would be unlike the (US) railroads, which never prepared for the era when people would decide to spend 6 hours crossing the country instead of 72 hours.

Tuesday, November 27, 2007

Is Silicon Valley still dominant?

In summarizing the Nokia “Web 3.0” article, I forgot to mention one key point. The whole tenor of the article (and the accompanying series) is part of the Merc’s crusade to reassure locals that Silicon Valley is still the center of the technology world. For example, this photo caption from Sunday’s paper (not on the web):

Engineers test imaging technology at Nokia Research Center Palo Alto. The opening of the facility a year ago illustrates how the valley’s pre-eminence as an incubator of innovation has been enhanced by economic globalization, not diminished.
Monday’s Part 2 carries on the theme with the headline “Valley’s edge: Success hard to copy.” I’m sorry, but this is way too Pollyanna-ish for my taste — seriously in denial as to the reality of globalization and the product life cycle.

I have no problem saying that there are times that the jobs moving to Silicon Valley outnumber those leaving the valley, and that more such times may lie ahead. But technology jobs are moving to Singapore, Tel Aviv, Banaglore, Beijing, Kiev, and any number of other places, not to mention other technologies being created and growing where the center of action is elsewhere (like Helsinki and Tokyo for mobile phones). So as with any other high-wage, high value-added industrial cluster, Silicon Valley faces the constant challenge of re-inventing itself to find new innovations that can only be created here. It can be done, but it’s not a sure thing.

I’m sure the CEOs and VCs of Silicon Valley don’t believe that SV is invincible, because for the past decade they have been pouring their money into other parts of the world. For many entrepreneurs, there is no other option — the kids’ soccer league and public schools are here, as are their professional networks — so they will start firms here no matter what. As in the past 50 years, some of these will succeed and many will fail.

But what about employment of the rank-and-file? Like London and Geneva, the Bay Area is becoming impossibly expensive for clerical staff, schoolteachers, cops etc. etc. (For example, for hospitals Northern California has 9 of the 10 most expensive labor costs in the entire U.S., a third higher than the national average.) Eventually these labor costs will push the cost of middle class living even higher. So even with the success of valley companies, how much of that success will accrue to local employees and the local economy?

The so-called Web 3.0

The business section of Sunday’s Mercury had a big cover article on Nokia’s presence in Silicon Valley, specifically the Nokia Research Center Palo Alto. Besides being a great place to host a Mobile Monday meeting, the article notes that the center has about 50 employees — apparently drawn from the alumni of some of Silicon Valley’s best companies. (The NRC PA director, Bob Iannucci, once was VP of research for Compaq and with it, head of DEC’s famed Western Research Laboratory.)

Not surprisingly, the Page Mill Road location is enabling "Nokia's new research relationship with Stanford University." (The normal translation: Nokia is putting up money to support Stanford research or access to that research). In 2006-2007, the two jointly hosted a series of research talks at Stanford.

Most of this was the customary “foreign MNC comes to Silicon Valley” story (to access talent, partners, etc.) But the title was provocative if not silly:

Phone giant Nokia navigates Internet’s third wave from Silicon Valley base
The explanation is buried within the story:
Nokia's efforts, Iannucci said, are aimed at maintaining the company's market lead as handsets equipped with robust Web capabilities - communications, search, video and more - become the third great wave of commercial opportunity.

The first wave, Iannucci said, represented "the democratization of consumption of information. Web 2.0 is about the democratization of information production. Web 3.0 - the next step - takes Web 2.0 and makes it mobile."
The term “Web 3.0” shows up frequently on the Web (2.0? 1.0?). Right now it’s a meaningless buzzword that means “new and improved,” right up there with “4G.”

Apparently I know a little more than the credulous Merc reporter about the mobile Web 2.0, having studied it for the past 5 months while supervising the forthcoming master’s project by Eduardo Sanchez and German Benitez. The short answer is that some Web 2.0 is desktop based, some is mobile, and some is both. So it’s ludicrous to suggest that adding things like ubiquity or mobility to existing Web 2.0 plans is going to transform this into Web 3.0, even though it will make Web 2.0 more widely available and more powerful.

A better definition of Web 3.0 is the semantic web — one where we are finding information based on meaning and not keywords. Interestingly, a Nokia researcher published an article promoting this definition in a peer-reviewed IEEE journal last June.

Tuesday, November 20, 2007

Another mobile carrier's CFIT plans

Earlier this year, I referred to T-Mobile’s denial of the inevitable commoditization of mobile phone network operators. It's happened to everything else in telecom. Now even international long distance is effectively free, thanks to VoIP carriers like Skype and my new favorite Lingo (which includes unlimited long distance to 23 countries for $22/month).

Sometimes I want to say it's a train wreck waiting to happen. But after reading the weekend transcript of the Financial Times interview with Vodafone's CEO, I think the more accurate metaphor comes from aviation crashes: CFIT, controlled flight into terrain.

The CEO is Arun Sarin, a former Pac Bell executive who (only about 100 km from here) designed the technology for its industry-leading cellphone services in California during the 1980s including the 1984 LA Olympics. Sarin then helped Sam Ginn spin off PacTel Cellular to form AirTouch, which later got bought by Vodafone and then merged with Verizon's US properties to form Verizon Wireless.

The FT news article (full text here) highlights' Sarin's atypical opinion of the iPhone, but he's certainly right that 2.5G is a lousy way to watch YouTube. More seriously, The Register notes his excess optimism about his long-term pricing power due to inexorable increases in competition.

Sarin admits that the mobile phone service will be a flat monthly fee within 5-10 years, although he still hopes to exercise price discrimination with multiple minute bundles. Presumably he's not worried about existing flat-rate mobile phone services.

He also expresses the heartfelt desire that his pipes won't become commodities, and that his walled gardens will somehow compete with the likes of Google and Nokia. Vodafone will triumph because it owns the billing relationship and because no one can build LBS without paying it a toll:

"Most importantly, we have 240m customers. We have the relationship with the customer, they are either buying top-up cards from Vodafone, or we are billing them on a monthly basis. Just the simple fact we have the customer and billing relationship is a hugely powerful thing that nobody can take away from us. We could lose our customers and then, yes, they could be gone.

“The second thing is we know where the customers are, in terms of location. We know precisely where you are. Frankly only we know where you are. The handset manufacturer that sold you the handset does not know where you are. But we know where you are.

“So if you say at the most basic level say we have got a customer relationship, we have got billing and we know where you are, these are hugely important things. So whoever comes into the marketplace is going to have to work through us."
So no one will get on the Vodafone network without paying Sarin's toll — faithfully recreating the AOL/BOL/CompuServe/NiftyServe walled garden model of the 1980s. Meanwhile, industry upstarts will be building a 21st century mobile version of the Internet with as much free (or cheap) third party content as possible. I guess it will be up to the next Vodafone CEO to deal with the consequences.

[Mixed metaphor alert] Sarin notes he still comes back to California (and Hawaii) to surf with his son, so he knows what a wipeout is. I guess he's trying to ride the wave as long as he can, but at 53 the eventual crash is going to come long before he reaches retirement age.

Kindling a feeling of deja vu all over again

Despite his mangling of the English language, Yogi Berra's aphorism has captured an eternal truth that applies to business history. This week's example is yet another e-book reader: in one of the worst-kept secrets ever (too many discussions with gossipy publishing-types?), Amazon's Jeff Bezos finally unveiled the Kindle.

Attempts at establishing e-books date back a decade, with at least a dozen failed efforts. The big names have included Sony, Microsoft and Adobe. Yes, the technology is getting better — the E-ink reflective screen rather than a backlit LCD screen, and Amazon is using Sprint's 3G wireless network rather than sideloading for content access. It also includes limited e-mail and browser capabilities.

These information goods have high returns to scale, hence both the incentives for success and the inability to survive as a niche product. The question still remains: will someone pay $400 ($100 more than Sony's product) for a book viewer when he/she was born with the necessary equipment to view dead trees? Unlike newspapers, electronic delivery has displaced less than 0.1% of book sales.

Here Amazon is trying to use its distribution might (the upsurge of traffic during the Christmas shopping season) as well as its content relationships to succeed where others have failed. But the business model is fatally flawed.

I'm shocked that they didn't try to do more to solve the angry orphan problem. The device's only native format is its proprietary AZW, requires conversion to open .DOC and .html files, and doesn't do PDF at all. Books from AZW are only available from Amazon and only viewable on an Amazon device, so when Amazon throws in the towel there will be no way to view them. At least the iPod value proposition was primed through use of unprotected MP3 files, so that most of the iPod content (initially) could be played on any PC, MP3 player or other device.

There are clever attempts to make money off of newspapers, magazines and even blogs. I think they mortgaged their soul to provide ubiquitous connectivity (which today is still expensive in the US, unlike telecom commodities like e-mail and international long distance). They need free user-generated content to fuel ubiquity and adoption, but their connect fees won't allow it. Take a cell phone, add e-books, improve the MySpace/Facebook/YouTube web access, and stand-alone book viewers are toast.

Theoretically the COGS should be less for e-delivery, but the book pricing does not appear to be terribly aggressive (given the manufacturing and distribution savings), perhaps because the publishers fear cannibalism. Meanwhile, I get a book I can't lend to a friend, can't sell to a used book store and can't donate to my local library. Thanks to Amazon's inept strategy, many more trees will have to die needlessly.

The spin is that this is an iPod for books, but's only spin. John Paczkowski of the WSJ blog mercillessly lampoons it as "the Zune of reading."

Brad Stone of the NYT claims the problem is features, but I think it's all the business model. If they asked me what to do (they won't) to fix the business model, I'd make three changes:

  • Support open document formats (HTML, RTF, even PDF) and then like the iPod, make those the native formats for public domain (no-DRM) content.
  • Offer a monthly subscription that makes unlimited e-mail, web browsing, public domain works and free Internet content (like blogs) available at no additional cost.
  • Get publishers to offer aggressive discounts for impulse purchases, such as 1-day specials for novels trying to build word-of-mouth to get onto the NYT bestseller list.
Stone also suggested advertising, which makes sense given the's recent captitulation to the Google-ization of paid media.

I know one reader who today is certainly reciting Yogi Berra — the former COO of NuvoMedia, makers of the Rocket eBook, which exited through sale to TV Guide's parent. The Rocket eBook lasted less than three years, the Apple Newton not quite four, so I'll bet $50 (giving 2:1 odds) that Kindle is off the market by the end of 2010.

Saturday, November 17, 2007

Two months of IT news

I’m done with grading my MBA classes and my undergraduate midterms, and thus can spend (a little) more time on the blog. During those two months, I set aside a large pile of stuff that I set aside to blog about. Since those items aren’t going to get their own article, instead I thought I’d write a combination article with a line or two on the items that I could find. My notes for September aren’t very good, but I can remember the recent run of interesting news:

  • Sept. 10: I heard Richard Stallman give a talk on GPLv3 (hosted by Larry Lessig, show to the right). I even posted the pictures to Flickr, assuming I’d have time to blog about it. Stanford has a recording online in the obscure (open standard) Ogg Vorbis format, which requires installing a special player. According to my notes, Stallman claimed the major GPLv3 benefits are: international legal compatibility, patent retalliation clauses, anti-TiVo clauses, anti-DMCA clauses, and compatibility with other OSS licenses.
  • Sept. 17: After denying the possibility, the New York Times cancelled its TimesSelect paid online content experiment. As Steve Johnson of the Chicago Tribune said: “the white flag has been waved on the notion that content in the digital realm is worth anything close to what it is in the tangible world”; since then, Rupert Mudoch has predicted he will take the WSJ in the same direction. Interestingly, running the numbers for an MBA assignment, it’s hard to see how the NYT could shift from dead tree to online content without a 90% cut in staff.
  • Sept. 28: Palm probably helped its revenues with the introduction of the Centro. But it’s hard to see how a cheap version of the Treo solves the fundamental problem that its R&D no longer produces differentiation.
  • Oct. 4: EBay’s $1.4 billion writedown of its 2005 Skype acquisition for $2.5 billion. As noted by skeptics at the time Skype was acquired, Skype has a lot of customers but not much of a revenue model (freemium or not). Or, as we used to say, “losing money on every customer but making it up on volume.”
  • Nov. 4: The release of the new business memoir by Tom Perkins, the co-founder of Kleiner Perkins who shows why he was one of the few heros of the past decade of HP’s bumbling and mismanagement. The corresponding 60 Minutes interview wasn’t very insightful, as it mainly was about Leslie Stahl attacking him having for bitter policy fights with female business leaders. However, at least we got to see the Maltese Falcon, his 88 meter computerized sailing yacht that’s the size of a 19th century cargo ship.
  • Nov. 9: Three months after I questioned the business model, Sprint dropped plans to implement WiMax in partnership with Clearwire. It’s a big negative for WiMax, a bigger negative for Clearwire, and also bad news for Craig McCaw, the non-executive chairman who controls Clearwire. On Nov. 14, allies of Clearwire chairman Craig McCaw leaked to the WSJ that he has other buyers seeking to bail out Clearwire, but that seems more like a negotiating ploy than real news.
  • Nov. 9. Apple’s iPhone rolled out in the UK and Germany. Unlike the US rollout, I didn’t witness what happened but I gather the iPhone was achieved modest success. To confirm the latter point, on Nov. 13 the CEO of China Mobile (the world’s largest cell phone carrier) said he’s talking to Apple and his stock gained 6.6%.
  • Nov. 11: In my Sunday paper, the Mercury News noticed that the unusual rash of IPOs of the dot-bomb era didn’t last, and (surprise!) that firms are being acquired rather than going public. Acquisition has historically been the exit strategy of most tech startups for the past 50 years — it was rare that everything worked just right to IPO (and in some cases, like, that shouldn’t have gone public)
  • Nov. 12: A great article by Steven J. Vaughan-Nichols announced that the Open Document Foundation has closed shop, alleging that the foundation’s founders Sun and IBM tried to sabotage document interoperability, and instead endorsing W3C’s Compound Document Format. I’ve had my differences with SJVN — he’s on the true-believer end of open source reporters — but he’s really captured well a complex story: in the end, neither the ODF nor CDF faction comes across as completely credible.
  • Nov. 14: Katherine Boehret (Walt Mossberg’s deputy) doesn’t really care for the Zune — saying that this year’s Zune is competing with last year’s iPod. Of course, this conclusion does raise for Apple the classic question for innovation-based business models: if you stop being a moving target, the competition will catch up. Can Apple find something to do every year to differentiates its music-video players? It’s running out of ideas for the Mac, just as Sony did for TVs and VCRs and DVD players. (Yes, this year’s flat panel screens are better than last year’s).
  • Nov. 16: Warner Music CEO Edgar Bronfman admits that “Jobs was right. I was wrong” in pricing iTunes downloads. I think this is evidence that (as with the NYT case), the changes for an online world will be wrenching but survival is possible.
I hope to resume blogging again — not as much as I did in July, but more than the past two months. I’m guessing that my focus will be on the gPhone and (related) Linux Mobile platforms, as well as my own research.

Friday, November 16, 2007

SDK from the Open Google Alliance

I made a mistake in reporting last week on something called the Open Handset Alliance. Despite my reporting (and the press releases and news stories) and what it claims on the Google Code pages, today there is no Open Handset Alliance. There is a Google-owned project called "Android."

A friend asked (quite reasonably) "Just curious if you have downloaded [the] Android SDK ... I don't want to commit my company to whatever the license says so I'll hold off on doing my own analysis."

I did a little snooping and what I found:

  • Looking at the download site and the license terms made clear: the download greement is with Google, Inc., a Delaware corporation, not the so-called "Open Handset Alliance". (Thus putting the lie to the idea of the OHA being "co-sponsored" by four other companies.)
  • There's no source code on the site, under an Open Source Initiative-approved license or any other license.
  • According to the architecture diagram, as much Android code as possible is based on a BSD (or comparable) permissive license. No viral GPLv3 here, so presumably DRM and music players will be available on gPhones.
  • Android is a platform for Java applets, not native C/C++ apps. Symbian, of course, has native apps, and it sounds like even Apple will have a native app SDK before the gPhone ships.
Alas, the SDK requires an Intel-based Mac (or Windoze) to run, and my first Intel Mac is about 2 months away. So unless I can borrow someone else's machine, I won't be doing any development any time soon.

Engineers get no respect

Restaurant owner Scott Adams this week summarized the problems that talented engineers (Silicon Valley or otherwise) face at the hands of pointy-headed bosses.
[Dilbert]My direct experience with such abuse is more than 20 years old. I quit my last engineering job in 1986 because they wouldn't take my product ideas seriously (two ideas they implemented under external pressure within 18 months of my leaving the company). Then for 15 years I ran company founded by two engineers, where it was the non-engineers who got no respect (that's right, Rodney, no respect).

So while I post this cartoon on behalf of fellow Beavers everywhere, the reality is that I'm probably doing my part to make their life worse. Every time we inflict upon a tech company a manager or marketing person who took business because (s)he wasn't good enough at math to be an engineer, we are contributing to this problem. My dual MBA/MSE students are great — like I once was, they may be better engineers than managers (6+ years of engineering training but 18-24 months of business training), but at least they'll understand the technology that makes the business possible.

Fortunately, many of our undergrads are taking low tech jobs, while our MBA programs continue to be filled with engineers (taking the path required by HP 15 years ago) starting as engineers, getting an MBA, and then resurfacing as marketing experts. Sure, I can see hiring someone to sell digital cameras from Sony or P&G, but I hope you wouldn't find your enterprise computing or semiconductor equipment rep there.

Monday, November 12, 2007

Microsoft's survey of opinion leaders

I was minding my business last week, trying to finish grading when the phone rang. The call came from the 204 area code (Manitoba, I later learned Winnipeg), and the caller identified herself as being with Ipsos Public Affairs. For participating in a survey I was offered a $50 donation to one of a variety of charities, both blue chip (Alzheimer’s, Cancer Society, American Heart Association) or socially conscious (Habitat, Doctors without Borders) but not the charity Peter Drucker told Forbes was “by far the most effective organization in the U.S.”

The donation should have been the tip-off, because the interview ran 48 minutes. Most of the survey asked my opinion of seven high-tech companies: Adobe, Apple, Cisco, Google, IBM, Microsoft, Oracle, although some questions also “Linux” (or “Open Source”). However, the bulk of the interview — nearly half — was on comparing Google vs. Microsoft vs. IBM.

Having covered political surveys as a newspaper reporter, it only took 10 minutes to identify this as a Microsoft-sponsored survey. One hint was the three way comparison: Google and IBM are major rivals to Microsoft but not to each other. Other hints were the questions that matched Microsoft’s PR campaigns or potential Google criticisms.

I started taking notes about 5 minutes in. The major comparison questions between Google, Microsoft and IBM (in that order) are these questions:

  • [paraphrase] Has a technological vision
  • [paraphrase] Works well with hardware and software from other companies
  • [paraphrase] Is good for the US economy
  • Helps the us technology industry succeed
  • Gives customers control over the privacy of their information
  • Helps people and businesses realized their potential.
  • Is a responsible leader
  • Is committed to making charitable and social contributions in the US
  • Is a company I trust
  • Is a technologically innovative company
  • Offers products of the highest quality
  • Prices its products fairly in the US
  • Cares about its customers
  • Makes secure products
  • Is committed to being a leader in online safety
  • Is committed to making secure products
  • Its business practices meet the highest standards
  • Promotes opportunities for technology skills for underserved populations in the US (senior citizen, disabled)
  • Is an agile company
  • Is a dynamic company
  • Respects the individual property rights of individuals (for example, artists) and companies (e.g. publishers) for use for their content on the Internet
  • A company that has access to too much personal information
  • A company whose business practices are overly aggressive

In the second part, I was asked:

  • What company from the list I most associated with innovation [Apple, of course]
  • Any recent news I recall of IBM, Google and Microsoft
  • What firms would want with business partnership, and how good the 7 main companies are as partners.
  • What a company [like Microsoft] should do to reach out to the SV community
  • What technology-related public policy issue is most important [I said interoperability], which brought questions about what about interoperability was important, and asked to rate the 7 main companies on this.
  • I was asked to compare open source (“like Linux”) and proprietary code (“like SAP”) as to which one was better: ease of use, increasing global competitiveness of US companies, work well with products of other companies, security, quality, growing the local IT economy, innovation. [“Local” was ambiguous, but to me that implied the user’s country, e.g. India, China, etc.]
  • A series of questions about the role of IP as an incentive, the optimal strength of IP for various industries, and whether specific companies’ (RedHat, Sony, Siemens, Microsoft, Google, IBM) approach to IPR “fosters innovation and technology development in the US.”
  • Questions about Microsoft’s ads, and corporate citizenship
  • Questions about what news sources I follow — newspapers, websites, and bloggers
  • Questions confirming my political party and job title (which they had as “business school faculty”

Updated Nov. 14 (Thanks Tom) The list of bloggers must be some sort of badge of honor. The list that I was asked about was (in order) Eric Savitz, Thomas Hawk, Robert Scoble, David Sifry, Tom Foremski, Scott Beale, Kevin Rose, Michael Arrington, John Battelle, Om Malik, Fake Steve Jobs, Valley Wag, and Marc Andreessen. I’d only heard of the last four, although I didn’t know Andressen had a blog and only Malik’s blog has been useful to me in the past.

I’m not sure how I ended up in their sample, nor how they got my phone number. They called me on my GrandCentral number (which is on my most recent business card and my website). This is what Ipsos promises its clients:

It's about understanding and managing issues, and advancing reputations.

We've been doing just that for clients in the private, public, and not-for-profit sectors. Every day. …

We provide boutique-style customer service and work closely with our clients while often also undertaking research on a global scale. Clients receive forward-thinking solutions to data collection, innovative use of research technologies, and the strategic insight to evaluate and respond to ever-changing client demands. …

The essence of what we do is help our clients listen to what their audiences are saying, understand what they are thinking and anticipate what they have in mind. We know how best to determine and measure their views and opinions. But we go beyond delivering data. We analyze it, put it in context, and then let our clients know how they can best translate this understanding into efficient and effective policies, programs, communications strategies, and marketing initiatives.

I’m not sure what Microsoft is going to learn from its survey, other than perhaps quantitative measures of Google’s vulnerabilities like privacy and resentment of its march to Total World Domination.

Monday, November 5, 2007

Incompetent gPhone coverage

Rushing to work to give a midterm this morning, I didn't open my paper to see the WSJ previous coverage of the rumored gPhone intro. If I had, I would have seen one of the most inept examples of technology journalism this year in a major publication.

So for my former technology strategy students, what is wrong with this table from page B2 of Monday's WSJ?

C'mon, is it really that hard?

Hint: if "smart" is orthogonal to "open", then is it possible to have all combinations of each dimension? (I.e., "smart"+"open", "smart"+"closed", "dumb"+"open", "dumb"+"closed")

Extra credit option: Dozens of publications have unquestioningly reprinted the Google claim that the gPhone platform "will power thousands of different phone models." So,

  1. How many phone models does the world's largest phone maker (Nokia, a non-OHA member) currently sell?
  2. How many different cell phone models has the world's most popular "open" cell phone platform (Symbian OS) shipped since its first product in March 1999?
  3. If beginning in July 2008, the gPhone platform is used for every new phone model by the two largest OHA members (Samsung and Motorola), how long would it be (at the current rate) before they together had shipped 1000 new models?

gPhone becomes Open Vaporware Alliance

Today Google announced the gPhone. What is it? Yet another version of mobile Linux with yet another trade association —— the Open Handset Alliance — which is led by Google, T-Mobile, HTC, Qualcomm and Motorola as the major sponsors (in that order according to the press release).

There are so many levels on which I could dissect the announcement, but I still have to keep my day job, so let me mention only a few:

  • Motorola is also a member of the LiMo Foundation, Gnome Mobile and other mobile Linux initiatives, but only some of its allies from these initiatives (Intel, DoCoMo, Samsung) are present here.
  • Google CEO Eric Schmidt bragged that this is not a mere phone (take that, Apple) but a platform: "Today's announcement is more ambitious than any single 'Google Phone' that the press has been speculating about over the past few weeks. ... Our vision is that the powerful platform we're unveiling will power thousands of different phone models."
  • Now we know why Google paid untold billions in July 2005 for Android Inc., the company at the center of Google's new initiative. Co-founders Andy Rubin (who also founded Danger, OS suppliers to Sidekick) and Rich Miner are firm believers in a post-PC world, as evidenced by Miner's quote in the NYT
    "There is a good chance that the number of potential consumers of Google…may never have a PC. ... So one of our clear goals was to find a way and have a platform (to) deliver Google services, Google content and Google search into those markets, and the mobile phone is going to clearly become much more of a development platform."
  • With Qualcomm are its two most bitter US rivals, Broadcom and TI, both of whom would like to destroy its business model.
  • The Japanese carriers represented include both Qualcomm-loyalist KDDI and nemesis DoCoMo. Among CDMA carriers in the US, Sprint is aboard but Verizon is notably absent. Of the three major European carriers, T-Mobile is present but Vodafone and Orange are missing. Other GSM carriers including China Mobile and (in Europe) Telefonica and Telecom Italia, but not AT&T, the only major US GSM carrier.
There are some reasons for skepticism:
  • In addition to the carriers, absent were the major European and Japanese handset makers. In particular, Nokia and Ericsson are firmly committed to the Symbian platform, and even if Google's Web 2.0-enabled vision becomes a reality, can have the best of web- and native-client applications through their existing smartphone investments.
  • The fawning pro-Google euphoria was reminiscent of, well, Apple's iPhone roll-out. Even 10 months ago, Steve Jobs did a demo of his vaporware, and all the OHA had to offer today is a phone call. Plans to ship "second half of 2008" puts them 12-18 months behind the iPhone.
  • Instead of the fawning headlines, the term "open" should have been used in quotes (but then the reporters don't study open standards for a living). This is another pay to play trade association, and right now it's not clear how open it is (although almost any Linux group is more open than LiMo).
There are other holes in the story. Android co-founder Rubin said "We've created the platform to be open source," but most of the platform components and layers (as far as we can tell) existed long before Google/Android came along. Rubin also claimed that the phone is the ultimate Internet phone:
Our partners will say that the phone does the Internet, in a way other phones don't do the Internet. We don't have to brand it. When you walk into the store and see a phone with lots of Net applications, vs. others with hardly any, people will get it very quickly.
Of course, the iPhone also does the Internet (pretty well) and most of Apple's major rivals plan their own mobile web strategy. A year is a long time, and it's now certain Nokia will be ready. Motorola and Samsung are both Symbian and LiMo participants, so maybe they hope someone will give them a first-class mobile web experience.

Embedded Linux is doing what it does best, which is fragment like crazy: 2007 has brought more fragmentation and not less. Maybe Gnome Mobile was going nowhere (the website looks a little stale), but the joint effort of Intel's Moblin and Nokia's Maemo seemed promising both from a business and technical standpoint.

Sunday, November 4, 2007

Alleged end of the PC era

On Sunday, AP reported from Japan that declining Japanese interest in PCs marked the beginning of a global trend. Balderdash.

Sure, I'm biased. I made my career and reputation in the PC industry — first in industry in then in academia. And that by my family members, I'm considered one of the fastest touch typists they know (from spending several years as a teenager touch-typing conputer programs). Even allowing for that, I think the evidence suggests this claim is overblown.

Here's what the dispatch claims:

Japan's PC market is already shrinking, leading analysts to wonder whether Japan will become the first major market to see a decline in personal computer use some 25 years after it revolutionized household electronics - and whether this could be the picture of things to come in other countries.

"The household PC market is losing momentum to other electronics like flat-panel TVs and mobile phones," said Masahiro Katayama, research group head at market survey firm IDC.
Part of this is just arithmetic sophistry. Certainly I could understand if (worldwide) PC replacement rates slow down — that doesn't mean that their installed base or importance as an access terminal has gone done. Someday, everyone will have a flat panel HD television and the run rate for flat panel TVs will go down, too — but that won't mean that television is no longer important as a medium, or that people no longer value their flat panel TVs. (Remember color CRTs were a mature category for 25 years, but they certainly were important).

More fundamentally, Japan is an atypical market on several fronts.

I am willing to concede that part of the world loves QWERTY (or AZERTY) keyboards and part does not. Until we get portable voice recognition, there is no speed comparision for Western languages for data entry on a laptop vs. a mobile phone. For Japanese and Chinese, the keyboard is not nearly as efficient and thus the mobile phone is a bette substitute. That's not true for large amounts of text production in Western countries.

Finally, we've been hearing for years that the Japanese will waste money on electronic gadgets rather than saving for a house, because they have no hope of ever owning one. Does this presage a global trend of consumers uninterested in home ownership? I hope not, as home ownership proved to be the bedrock of economic and political stability during the 20th century.

The key question to me is: how far off is true speech-to-text, as in the Asimov Foundation novels or 2001? Once I can dictate to a portable device instead of clacking at a keyboard, I'm outta here. My hunch is that this is at least a decade off — my retirement is set for somewhere around 2025, so it may or may not arrive in time to do me any good.