Subclock

The 100 Oldest Companies

2008-06-14T09:32:00.000-07:00

I picked up this link, about the 100 oldest companies in the world, from a post by Kevin Kelly on the Long Now blog. The profiles of the companies are really interesting. What do you want to bet that these companies have consistently worried about cash flow and profitability more than they have worried about share price?

Charter school will focus on homeland security.

2008-05-07T15:49:00.000-07:00

I really want this to be a joke.

Sky marshals on the no-fly list

2008-05-02T07:04:00.000-07:00

This is ridiculous -- but makes me feel marginally better about the trouble I've had with the list myself.

The Next Big Thing

2008-04-29T07:12:00.000-07:00

Since I left Oracle last January, I've been spending time looking around the technology industry for interesting trends. Sooner or later I will have to take a job, so this is my chance to pick a horse that will be fun to ride.

There are now roughly one million Software-as-a-Service companies per square mile in Silicon Valley (they are thicker than that in Palo Alto, but Stanford has always been peculiarly fecund). Even Microsoft is on the bandwagon, so you can be sure that SaaS has jumped the shark. Just lately, I've seen pitches for Data-as-a-Service, Hardware-as-a-Service and Infrastructure-as-a-Service.

Here at the Subclock blog, we're getting out in front of this thing. I'm pleased to announce today the creation of Service-as-a-Service. Want to start your own -as-a-Service company, but not sure what that means, exactly? Just set your business plan down on top of Service-as-a-Service and let that advertising revenue roll in! In fact, Service-as-a-Service is itself built on Service-as-a-Service. It's turtles all the way down, baby!

Service-as-a-Service takes advantage of all the latest Web buzzwords. It's platform-enabled, naturally. It's fully cloud-compliant. We leverage community. Our tech is simultaneously green and clean. We've sprayed so much Xen around here that the floors are sticky.

Service-as-a-Service isn't a company. It's way more Valley than that. Think big O'Reilly-style unconference, with probably some blogs and a wiki. Think Facebook, but cooler. Think tweet storm. Most of all, think huge VC investment opportunity. Got an office on Sand Hill Road? Give us a call here at the Subclock blog -- operators are standing by.

More security theater.

2007-03-22T16:50:00.000-07:00

Tom Kyte posts a story about crossing into the US from Canada.

Tenacious search.

2007-02-01T13:50:00.000-08:00

Well-known database researcher Jim Gray is missing. A group of friends and colleagues is organized to help with the search. We're posting updates here.

Concentrating.

2007-01-24T09:50:00.000-08:00

Steve O'Grady has an interesting suggestion for Sun: Get out of the fulfilment business that your new Startup Essentials program created for you, since you're not a Web commerce company. Steve suggests Amazon or eBay as order and fulfilment partners.

Coincidentally, I finally got around to listening to Werner Vogel's podcast at ITConversations on scalability, reliable services delivery and Amazon's Web Services APIs. He describes an Amazon very different from the bookstore many people think about. Much of what Werner says supports Steve's argument: You'd be crazy to build your own store, these days, with AWS' e-commerce APIs and systems at your disposal.

No-fly list to be scrubbed.

2007-01-19T10:19:00.000-08:00

Even if you shrink the no-fly list by one-half, it is still a waste of time and money.

Cisco iPhone trademark suit.

2007-01-11T07:37:00.000-08:00

Cisco's General Counsel, Mark Chandler, writes about the company's claim against Apple for trademark infringement on the iPhone. This case will be an interesting one to watch.

Steve Wozniak gives my old Apple ][ a hug.

2006-12-22T11:52:00.000-08:00

Bruce Damer runs the DigiBarn, a sort of alternative to the Computer History Museum. I enjoy the CHM, but in a lot of ways I think that DigiBarn is closer to the old-guard hacker ethos -- home-grown and eclectic, it's the product of Bruce's pesonal vision and efforts, rather than a well-funded corporate retelling of the story.

I met Bruce through a conference that we both attend with some regularity.

The first real computer that I ever used was an Apple ][ with 16K of RAM, upgraded periodically over the years. I still get enthusiastic about words like peek, poke and hires. I remember working past midnight with my stepdad, Bill Jellison, writing programs in 6502 assembler. I am still ridiculously proud of the first real computer game I wrote -- a horse racing game with ASCII graphics and betting that used random numbers to control outcomes and some trickery that I invented myself to avoid integer overflow in computing your bankroll after a win.

I was certainly a latent geek up to that point, but that Apple ][ is what turned me into an engineer.

My stepdad bought that computer in the first place, and he and my mom held onto it for years, even after its place on the desk was taken over by a Lisa, and then by a 128K Mac. When they moved out of the house, they shipped it, along with a bunch of other retro computer gear, to me.

I eventually donated the machine to Bruce and the DigiBarn. It was one of the very first Apples made -- serial number 495 -- and almost certainly assembled by the Steves in the garage. It was doing no one any good in my attic, and seemed to both me and Bruce to be a valuable addition to the DigiBarn's collection.

A week or so ago, Steve Wozniak visited Bruce while filming a program for the Discovery Channel. Steve Jobs gets most of the Apple glory in the press, but Woz will always be the original Apple to me. I know the story of the floppy controller redesign over Christmas. The Apple ][ was elegant, and Woz was the engineer in the company. It's really unbelievable what he accomplished with a bunch of MSI components and a connection to a television set.

Bruce has posted pictures of Woz's visit, including several of his cuddling up to my old Apple ][. You need to scroll down to the middle of the page for the pictures I like best, or you can just look at this one.

Woz, that is exactly how I feel about that computer. Thanks, man.

Five things.

2006-12-21T09:41:00.000-08:00

Dave Rosenberg over at Open Resource tagged me in the "Five Things" game that's taking the blogosphere by storm. Under the rules, I have to tell you five things you didn't know about me, and then I have to trickle down the trouble on five other people.

The last time I played in a pyramid scheme, I lost fifty guilders. (You didn't know that about me, but it doesn't count).

My five things:

The third time that I jumped out of an airplane, my chute fouled. I had to cut it away and deploy my reserve chute. I remember falling face-up, sixteen feet per second squared for eternity, watching the beautiful white reserve chute unfurl against the cobalt.
I plan to write a novel. I have the subject and plot. I need to retire from technology for two years, somehow, to do it. The plan is long-term.
I am crazy in love with Shell Knob, Missouri.
I have a minor in Dutch Studies from UC Berkeley. UC Berkeley is one of only a few schools in the nation with a Dutch Studies program. Ik ben ook wel gek op Amsterdam, maar het is heel anders dan Shell Knob.
I have a hard time remembering which is my left hand and which is my right.

I tag Fabrizio Capobianco, Kevin Walsh, Marten Mickos and Peter Fenton. Four equals five for low values of five.

Dweebs, horndogs and geezers.

2006-12-20T10:21:00.000-08:00

Nicholas Carr has a great post on the differences in the top searches by users of Google, Yahoo! and AOL. I've believed for a long time that communities self-organize, but I didn't really think you would see it at the level of major web properties like this.

For the record, I'm a dweeb.

NYTimes on airport security.

2006-12-20T06:38:00.000-08:00

An excellent article on the TSA and the illusion of security.

Living in the haze.

2006-10-14T19:42:00.000-07:00

I'm in Singapore right now for work. This is the final stop on a week-and-a-half of conferences and customer meetings in Asia.

I've been to Singapore only once before. This is my first time to visit late in the year, and my first exposure to an annual problem that the locals call "the haze."

"Haze" is a polite name for an ugly problem. What's happening is that developers and farmers are setting fires in the equatorial jungles and forests of Indonesia, in order to clear land for crops or building. Once those fires are set, they generally continue to burn until the rainy season extinquishes them. A fire set by a small farmer, who is able to farm only a few acres himself, may burn thousands of acres before it's extinguished.

The problem began in the 1990s and has gotten steadily worse since. At this point, Singapore expects to breathe the smoke from Indonesia every October. Locals hope for winds that shift north or south to avoid the country, and an early start to the rainy season. Meanwhile, mothers in masks walk children in sealed strollers down smoky city streets.

Right now, the fires burning in the Kalimantan region of Borneo are producing a thick flume of smoke. That smoke is actually so bad that visibility in Jakarta is just five hundred meters, causing all sorts of problems, including airplane accidents. Prevailing winds carry the smoke across the water and over Singapore. Visibility here is between two and five kilometers, according to offical news sites this morning. Just looking out my hotel room window, I would bet more on two than on five.

The smoke is thick, with a sickly-sweet charred smell as you walk along the streets outside. It burns your eyes and throat. You can taste it. The Singaporean government measures the density of particulates in the air in order to monitor health threats. One hundred parts per million is considered to be unhealthy; today, the number is between one hundred twenty and one hundred thirty-five, depending on which news site you visit.

This is an annual problem in the region, my friends here tell me. The issue is that there is no clear will to stop the clearing of tropical forests -- despite the terrible environmental toll and the public health issues, the governments of the countries that create this problem tolerate it.

Clearly, economics are a big part of the problem. Farmers with few other options who need to feed their families are taking desperate measures to do it. I understand from the locals that corruption is also a problem. Developers looking to clear land flout the law, and pay fines or bribes to avoid any serious consequences.

It is really asonishing to me that a country as developed as Singapore has a chronic environmental problem of this kind. Obviously, it matters a great deal who your neighbors are. The governments of the region really must work together to stop the burning of forests. They must educate the farmers of the region, and offer them economic support, to make the fires unnecessary. They must also enforce laws already on the books, with heavy penalties for offenders.

First, though, I think we need to name the problem honestly. This isn't haze. People here -- adults, the elderly, small children -- are breathing the smoke from burned animals and trees. It is an awful stink of destruction, blowing into one of the world's richest countries from one of the world's poorest. It is an unmistakable message from one of the world's most diverse and productive, and profoundly threatened, ecosystems.

Burglary tools.

2006-09-09T12:29:00.000-07:00

I am a dilletante at lockpicking -- I have a couple of sets of picks, and when I am overburdened with spare time, like to hone my skills.

There's a great article by Liam Bowen on how to recover the combination of a Master padlock if you have lost it, with a link to the HowStuffWorks article on padlock internals. This is pretty neat stuff.

The Open Organization of Lockpickers is a good place to start if you are curious about pickng.

If you're interested in classic picking, you can shop for picks or snapguns online. I like picks better; they require a more delicate touch, but they work about as fast as a gun for me, and it's more rewarding to beat the lock by touch than by pulling a trigger. The guns can also damage the lock cylinder.

If you like snapguns, you should check out bumping. The TOOOL site has lots of information on bumping.

Travel advisory.

2006-09-08T11:06:00.000-07:00

I'm on the cusp of another batch of trips for work. I am not looking forward to it, thanks to the new list of forbidden items in carry-on luggage.

I was pleased today to discover that the Elephant Pharmacy in Berkeley carries Eco-Dent DailyCare brand tooth powder -- not a paste, and not, so far, a tool adopted by global terrorists, so permitted in the passenger compartment. I've scoured the local grocery stores and pharmacies looking for something like this. I am astonished at the number of stores that still display "travel-sized" containers of shampoo and toothpaste in a form you can't carry onto an airplane.

I agree with Bruce Schneier -- I don't believe these restrictions increase air security. I think they look good to casual travellers, but that's all. They're expensive to implement, inconvenient and don't stop attacks.

Testing Turing.

2006-09-07T07:51:00.000-07:00

An eventful summer is over. Labor Day was last Monday, so the kids are back in school and the white shoes are back in the closet. It's time to get to work.

The most interesting thing to cross my radar this week was an article by Tim over on the O'Reilly Radar. It's about Google's new Image Labeller project. When I saw the Labeller, I thought it looked a lot like the Mechanical Turk, but Tim -- thanks to Luis von Ahn at CMU -- had a better insight.

Like the Turk, the Labeller is a way for computers to use humans for computation. Tim points out other places where this happens -- CAPTCHAs, for example.

Then he reminds us of the Turing Test, and makes this fascinating observation:

This is an interesting variation on the Turing test, in which humans generate and grade tests that most humans can pass, but current computer programs cannot pass. Is there another variation in the future, in which computers generate and grade tests that computers can pass, but humans cannot pass?

I've been thinking about it, and I've decided that the fourth variation -- human testers, humans fail, computers pass -- is isomorphic to the regular Turing Test. It would, however, allow us to recognize the androids reliably in Blade Runner.

The idea of computers using humans for computation reminded me of George Dyson's talk at Google.

Interesting stuff!

Links for 20 June 2006.

2006-06-20T11:22:00.000-07:00

A mish-mash of this and that, just in time for the summer solstice:

Coté saves you the trouble, expense and ethical minefield of a Harvard MBA with his post on Econobonics.
A pretty cool story on a practical application of the Mechanical Turk.
Mmmm, Tangwich.
The must-have dating accessory of the new millennium.
Wanna be the biggest geek on the lake?

Following the tyger.

2006-06-20T10:57:00.000-07:00

Serendipity always happens when you least expect it.

After my visits to Singapore and Shenzhen two weeks ago, I spent last week in Beijing, Seoul and Tokyo. I'm back in California now, and had planned to write a follow-on to my earlier post, wrapping up my visits from the second week of my trip. Several other people have saved me the trouble, though.

Tim O'Reilly led the way with a post on South Korea. That post actually digests a long article in Business 2.0. The B2.0 article is excellent, but Tim's coverage is worth reading on its own.

The key observation here -- and I saw this in Tokyo just as clearly as in Seoul -- is that a strong governmental commitment to ubiquitous, reliable broadband has been very good for social services and businesses in Japan and Korea. With the platform that the network provides, both governments and companies can build and deliver interesting and valuable services. In Japan, the general feeling is that networks and mobility are ways to improve the quality of life. That's a sentiment that you don't hear articulated in quite that way in most of the US.

Paul Graham also posted two relevant, and very interesting, articles that showed up among my feeds while I was on the road.

One of them, How to Be Silicon Valley, explains the combination of factors that conspired to produce the Valley in California beginning in the 1950s and 1960s. Countries and states interested in duplicating that success can learn a lot from the lessons that Paul presents.

Paul's second article is called Why Startups Condense in America. It lists some of the important public policy and also attitudinal reasons that the US has been so successful at creating and building new businesses. This is a very interesting piece, and ends with some thoughts on the ways that the US can continue to foster the policies and attitudes that matter most.

Plenty of good reading here for tygers.

Tyger, Tyger, burning bright.

2006-06-10T06:48:00.000-07:00

I'm in the middle of a two-week trip through Asia for work. I landed in Singapore on Monday of the week just ending. I spent a day and a half there, then flew to Shenzhen in southern China for another three days. I'm on a plane to Beijing as I type this on Saturday afternoon, and will see Seoul and Tokyo next week.

This is the first trip I've ever made to Asia, and if you ignore the time taken to get here, I have enjoyed the trip a great deal. The people and the food and the sights have all been great. The work days have been very full, but productive and fun.

After just a week in the region, I certainly don't qualify as any kind of expert, but I've seen a consistent pattern that I think is interesting and worth writing about. Economic development is a major focus of public and private activity in the cities I've visited so far, and the difference between those cities and the Silicon Valley, where I spend my normal work days, is stark.

Singapore NYP

In Singapore, I visited Nanyang Polytechnic (Wikipedia here), or NYP. NYP is a curious sort of hybrid: It's a training facility for workers in engineering, design and biotechnology, but also a development laboratory for businesses in Singapore and elsewhere that want low-cost, high-quality help with technical projects. In the US, we have academic and commercial collaborations, but nothing so focused and unapologetically commercial as this. NYP's job is to use real commercial contract work as projects for students who want to learn professional skills.

NYP has a full-time professional staff that oversees the contract work, and that does the heavy lifting required by more demanding parts of the projects. They parcel out work, in three-month internships, to NYP students who take a break from their coursework for a quarter of a year to complete these tasks. NYP is funded by the Singaporean government, and by the companies for whom it does contract work. In some cases, NYP even negotiates rights to the intellectual property that its staff and students create, which can pay off significantly in relicensing or reuse later.

NYP produces three things. First, and most importantly, it educates students in the practical skills that they need to get jobs in knowledge industries after they graduate. Second, it generates a continuous stream of quality commercial contract work at good prices for local companies. Third -- and less obviously, but most importantly -- NYP creates an environment that helps small businesses to flourish, by offering piecework and people to help them get started. This is clearly the intent of the government in the operation of NYP, and it has clearly been successful -- five thousand students graduate from the institute every year.

Granted, there are academic and commercial collaborations in higher education in the US, but I can think of no institution that is as expressly focused on this as NYP is in Singapore. Most major US universities have fundamental research as their first mission, with undergraduate education coming in somewhere after that. In no case is there so explicit a mix of practical training and commercial contracting.

South to China

In Shenzhen, the concentration on business is even more pervasive. Shenzhen is in Guangdong province, just across from Hong Kong, and has served for some time as the gateway to Hong Kong from the mainland. In the late 1970s, the city was just a fishing village with some 30,000 residents. After two decades of investment and development, it's a city of 5 million people with a large industrial base, including manufacturing, shipping and high technology.

Shenzhen is impossible to describe in a few words, so I'll give a few statistics that I learned during a visit to the government's Shenzhen High-technology Industrial Park (SHIP).

The port of Shenzhen ships between 20,000 and 40,000 cargo containers per week -- Walmart, the US retailer, has a large operation there to acquire and transport Chinese manufactured goods to its stores worldwide. After establishing low-cost manufacturing industries in the 80s, the city has established electronics manufacturing and, more recently, software development industries. Shenzhen exported US $102 billion in goods in 2005, including US $2 billion in software -- a huge total for a new industry. Where once Shenzhen was just a place where foreign companies manufactured their own designs cheaply, it is increasingly the place where new systems are designed and new intellectual property is created. Per capita income is high and new building and infrastructure are everywhere. I heard, but cannot prove, that 500 new cars take to the streets in Shenzhen every day.

That kind of hypergrowth certainly offers cause for concern. However you feel about it, though, it is astonishing to see. China has decided to join the first world by developing an information technology economy. The question is not whether it will work, but when the West will notice that it is buying its most innovative high-tech products from the south of China.

Just anecdotally, I visited one major telecommunications company in the city on Thursday. Because they're an Oracle customer, I won't name them. Unless you know the China tech market you would not likely recognize their name anyway.

This company designs and builds its own technology products, including hardware and software, which it ships to customers around the world. So far, most of those customers are in Asia, Africa and Eastern Europe -- I bet that low price was one of the reasons for the market success in those places, early. Now, though, it is beginning to rack up customers in North America and Western Europe. The company did US $8.4 billion in business last year. Its products have matured in their established markets and are now ready to challenge competitors in more mature tech markets elsewhere.

We walked through a software development facility to a meeting room at the company. At least a third of the desks had sleeping bags rolled up and stored beneath them -- clearly, the engineers there are used to long hours and demanding schedules. When we left at nearly 7pm, the lab was full and no one showed signs of packing up for the day. It is a small thing, too, but the conference room where we had our meeting had a threadbare carpet, some beat-up chairs, a well-used white board and a cable-strewn conference table with power and network jacks spread across it. This was clearly a place where people got real work done, not a glamorous place to entertain vendors.

I don't remember the last time I saw that kind of concentration at a startup in the Valley, and this was an $8.4bn-a-year company with more than a hundred people just in the room I walked through. Again, you may have some concerns about balance and quality of life in a work environment like that one, but it's there however you feel about it. From my conversations with local people around the city, that company is no anomaly -- the people of Shenzhen, as much as its corporate leadership and government, are extraordinarily focused on improving their economic lot.

I'm no expert on Asia after just one week, but I'll tell you this: Something remarkable is happening over here right now. My impression of the US and Western Europe is that we have become complacent with our economic success, and perhaps it is right for us to relax after working hard to do well. Singapore and China, though -- and I bet the other countries in this region -- are working extraordinarily hard to challenge Western dominance in high-value industries. If we do not notice what is happening here now, we will certainly notice it in five years.

I want to add a brief disclaimer to this post: Despite highlighting the differences between Asia and the West, I don't mean to be xenophobic, here. I've seen some very interesting government and industrial projects in the last week. Smart people are working hard and doing amazing things -- if anything, I am a little jealous of the speed and scope of change I see around the region. I agree with Thomas Friedman in The World is Flat: If you look, you can see that the economic balance is shifting globally. Blake's fearful symmetry is staring right back.

The future of software.

2006-05-27T12:25:00.000-07:00

One of the best things about my job is the opportunity it gives me to talk to interesting people building surprising applications. I have been lucky enough to do a lot of that just lately. As a result, I have become convinced that the way we build, deploy and manage applications today is wrong, and that future systems are going to look and work very differently.

I wrote about this lately in my post on mashups, but I didn't think the thing all the way down to the ground there. This post is the first of two in which I'll develop the idea more completely.

The network is the computer

John Gage at Sun Microsystems dreamt up the catchphrase "The network is the computer" long before it was true. The vision then was that the world would be covered with computing power tied together with ubiquitous communications. The internet from space [1] still shows big dark holes, but it's clear that it's only a matter of time, now. If you're reading these words, then in most of the places you go, you have easy access to broadband and cycles.

Obviously, if the network is the computer, then the software you use is going to run on the network, and not necessarily on the collection of wires and chips underneath your desk. Gage was looking a long way out, but he saw the future clearly.

The timeshare generation

It's true today that applications run on the network, and not on your personal computer. Every time you fire up your Web browser or email client, you're running a distributed application. The client software on your local machine talks to server software running remotely so that you can read the news, shop for good deals on travel and keep in touch with your family.

Important business applications are moving in this direction as well. Before we sold Sleepycat to Oracle, we used Trinet as our outsourced HR and payroll provider, and Upshot (since purchased, serially, by Siebel and Oracle) for sales force automation. These hosted apps allowed us to work from anywhere in the world, to cooperate with one another and to rely on a central service to manage day-to-day operations of information technology that would have been a lot of trouble to run ourselves.

These services, and others like them, are useful and valuable, and I am glad that they were available to us. They are not, however, very interesting. They do not make good use of the network as the computer.

Essentially, these services are exactly like timesharing systems in the 1960s and 1970s. Instead of buying and running a large and expensive computer system for yourself, you contract with a specialist who builds and operates that system for you. You have the illusion that you are the only user of the system, but in order to realize economies of scale, the specialist provider is really sharing the same computers and software with lots of other people.

Hosted apps are the same monolithic standalone software packages that we used to have to manage on our own. We get better reliability and lower cost by centralizing them and spreading the maintenance cost across many users. Fundamentally, though, we are doing the same old thing on the brave new platform.

The IC revolution

A near historic analogue to this situation is the invention and adoption of the transistor in the 1960s and 1970s. When it was first invented, the transistor was widely viewed as an excellent substitute for the vacuum tube in electronics -- it was smaller, much more reliable and vastly cheaper. Vacuum tube systems were rapidly replaced by transistor systems, and radios could suddenly fit in your shirt pocket.

The real power of transistors wasn't unlocked until the advent of digital systems [2], and especially the invention of integrated circuits (ICs) by Bob Noyce and others. ICs are not transistors doing the work of vacuum tubes better -- they are transistors doing something that vacuum tubes never could [3].

Today's hosted applications are nothing more than better vacuum tubes. They are an old idea -- timeshare computing -- copied to a new medium -- ubiquitous networked processor cycles. Hosted apps, like portable radios, are merely better. They are not different.

What will change

The next ten years in technology will see more and faster processing and networking. The change in quantity will drive qualitative change. We will begin to build applications that are different in kind from the ones we use today.

Applications of the future will not be monolithic systems centralized to simplify their management. Instead, they will be composed of small cooperating components, each specialized in a particular task, tied together on demand to perform a particular task. Pieces of the application will run in different administrative domains: IBM may get some data analysis from Microsoft in order to tune its Yahoo! ad keyword selection based on the clickstream it observes among shoppers on Dell's e-commerce site.

You can already see examples of systems like these. Mashups are a halting first step. Sun offers compute cycles for hire. Amazon is selling cheap online storage via S3. Internally, Amazon is building its core technology platform in exactly this way. Hard-core technology companies like Sun and Amazon are several standard deviations out on the high end of the curve, but over time this architecture will become commonplace. One day, ordinary non-technical consumers will not only use network computing apps like this. They will be able to program them themselves, easily tying information and analysis together to answer questions. They will not concern themselves with what work is done where.

The hard part

Software engineers have long ridden on the backs of hardware engineers. Computer programs are fast and sophisticated today mostly because the people with the soldering irons have made chips so fast and memories so big that we can be profligate when we program them. To some extent, we can follow the same strategy here. The technical trend toward ubiquitous computing is almost irresistible.

There are, however, critical problems we have to solve to make this new kind of application work.

When we reach across the boundaries of organizations effortlessly, and stitch together applications from all over the place, how can we trust the answers we get? How can IBM be certain that Microsoft got the right answers when it analyzed the Dell clickstream? Was that clickstream correct?

Just as importantly, how can we be certain these applications will run at all? Systems made of many small pieces have many places to fail. Any single component failure, or the failure of any connection among the components, can freeze the application as a whole. When we build distributed systems, even out of simple and reliable pieces, we introduce complexity. Complexity is a crushing weight that eventually guarantees failure. How can we manage that risk?

Those problems are hard ones -- too hard to explore here. I'll write more about them later.

Notes

[1] Eick's is one of several very cool maps digested by CNET. See in particular the colorizations by Bill Cheswick. Cheswick runs Lumeta, which specializes in building and rendering these maps. They don't show geography -- they show a deeper truth.

[2] Digital systems do not actually exist -- transistors are really just analog devices with very steep transfer curves. I have not mentioned it to anyone, though, because I do not want to undermine the global market for digital technology.

[3] I am not ignoring early work on tube-based computers. ICs are devices that could never have been built on vacuum tube technology.

Links for 16 May 2006.

2006-05-16T20:47:00.000-07:00

Go on. Click through. You'll be a better person for it.

Wonder where that spam came from? Paste the SMTP headers in the message into this way cool maps mashup, and see for yourself.
Jon Udell's got a useful set of bookmarklets.
Google Translate is just SYSTRAN for now, but the site promises to be driven off statistical data from large bodies of translated text one day. Fast, too.
And while we're at the First Church of Mountain View: Trends lets you find out what was hot, when.
There's a fascinating post by the OpEd editor at the Omaha World Herald showing the distribution of religious sects across America.

Friday is Bike to Work day.

2006-05-15T17:58:00.000-07:00

I will be setting a bad example, but I'm in favor nevertheless. Friday, May 20, is Bike to Work day.

Links for 14 May 2006: Special Wee Beastie Edition.

2006-05-14T15:11:00.000-07:00

That dang Internet just keeps making more stuff to read:

Schrödinger's cat is so last millenium. Quantum puppies have entered the light cone.
Origins of the Black Death: In defense of gerbils.
Emergent behavior: Cockroaches act democratically.
Some remarkable pictures among the winners of the National Wildlife Magazine photo contest (Google cache version is here, in case the NWM site failure is persistent).
Schneier's got a wicked squid thing going.

All mashed up.

2006-05-13T08:38:00.000-07:00

I must have been looking sideways just lately -- I have been busy! -- because I was surprised this week by several stories about a new idea in databases. Coté over at Redmonk posted links to several stories on his del.icio.us linkroll. Daniel Druker and Robert Rich wrote an article about it in DB2 Magazine, and Bill Snyder piled on in a story he wrote for TheStreet.com.

This new idea is called Master Data Management, and if you buy the momentum stories, it's the Next Big Thing.

I don't think that the emperor is entirely naked, here, but his mother ought not to have let him leave the house dressed that way. Master Data Management is an old idea in database systems, and all our experience so far says that it's unbelievably hard to do well.

First, though, some context.

Of all the ideas that are currently burbling around in the Web 2.0 cauldron, I personally find mashups to be the most compelling. The best mashups that I have seen so far are based on the Google Maps API. People are building sites that show the locations of their first kiss, local public libraries, sex offenders living in the area and more. The idea is to use simple, standard web-based interfaces to combine data from one site with map data from another. Once you internalize this idea, you realize that there are lots of different data sources out there that you'd like to tie together.

The old-fashioned name for this discipline among database researchers is federated databases. The idea is to take a collection of databases, created and maintained by different organizations for different purposes, and combine the information that they store in interesting ways. Much research money, and some investment capital, has been plowed into this idea, with (so far) no big bang. Those efforts have not been a complete bust, but in more than a quarter century of work, no single general-purpose technique has been discovered that works well.

The problem is that the different groups who build and maintain these databases collect and store information with different assumptions. Is my first name "Mike" or "Michael"? Are the prices you publish in euros or yen? Are dates represented in American or European format? The answers make a difference if you're combining records from different sources.

Worse, the reliability of the combined data is generally worse than the reliability of data in any single database. If my phone number is wrong in one database, and my age is wrong in another, then the combination is wrong in two particulars, not just one.

While inaccuracies like that may seem unimportant, they can matter a great deal. One of the example stories for the success of MDM is the casino that recognized a card cheat by a match on his telephone number with a different casino's employee database. Think about it: Do you know anyone that has written your telephone number down wrong? Would you want the companies you do business with to make decisions about you based on information that may be wrong, and that you can't review and correct?

It's absolutely possible to handle these issues, especially for single companies combining data that's all under their control. The established database vendors all offer products that do this, but they require careful analysis and considerable effort on installation. The information they operate on needs curation.

Mashups are much too powerful an idea to constrain to mapping apps. We'll see more, and more interesting, examples. Some will certainly tie together legacy data from a variety of sources, including relational database systems. This is an old technique, though, with a lot of practical experience highlighting problems in the field. Web 2.0 apps that use the technique but ignore the experience are going to deliver wrong answers.

Don't believe everything you read on the internet.

The Invention of Silicon Valley.

2006-05-07T12:53:00.000-07:00

I've just finished The Man Behind the Microchip: Robert Noyce and the Invention of Silicon Valley, by Leslie Berlin. It's a remarkable book. Living and working in Silicon Valley, you can sometimes forget that there was a time before integrated circuits and venture capitalists. Berlin does an excellent job of documenting the creation of the Valley and the emergence of an industry.

Her work on Noyce's boyhood and very early career is fascinating. He grew up in the Midwest. As a boy, he designed and built technical toys, like an aircraft powerful enough to lift him and his twelve-year-old brother when it was pulled behind a neighbor's car. Through a very good high school teacher, he learned of the Bell Labs work on transistors almost immediately after it was published. His fascination with the device led him to MIT, and into industry.

He quickly discovered an entrepreneurial streak in himself. That led him to Santa Clara, still a tiny town in the midst of apricot and orange groves. He worked for William Shockley, but strong personal differences drove him and his colleagues out, and together they formed Fairchild Semiconductor. Several years of success at Fairchild led Noyce and others to found a company called Integrated Electronics, shortened to Intel.

Besides Shockley, the book is loaded with names that industry veterans will recognize: Eugene Kleiner, Andy Grove, Gordon Moore, Arthur Rock and many others. There is the obligatory collection of Steve Jobs stories, in which the unkempt and ill-mannered teenager invites himself into a central role in the Valley and the industry. Noyce even rubbed shoulders regularly with Warren Buffett, as both were on the board of Grinnell College in Iowa. Intel is one of the very few technology investments that Buffett is on record as endorsing, but his endorsement carries an important qualification: "We were betting on the jockey, not the horse."

It's hard for an entrepreneur to read the book without identifying with Noyce -- his mix of passion and pragmatism, and even his professional failings, will be familiar to many who have started companies.

As much as it's a history of Noyce, though, the book is a history of the Valley, and the business and economic forces that shaped it. Berlin documents this wonderfully here:

In the same way that [Tandem, Atari, Genentech and others] build on the previous generation's technical advances, they also took advantage of the network of suppliers, venture capitalists, equipment vendors, specialized law and public relations firms, contract fabs, and customers that had sprung up in the past decade to support high-tech entrepreneurs in Silicon Valley. By 1983, more than 3,000 small consulting firms in Santa Clara County provided new companies with startup expertise and continuing help over the early years of operation. Many of the chip designers, glass blowers, fab houses, and die cutters that catered to Silicon Valley high-tech entrepreneurs were themselves small privately-held firms. This "supply chain," most often mentioned for its support of small companies, is itself an entrepreneurial phenomenon.

You read a great deal these days about the emergence of high tech economies in China, Eastern Europe and elsewhere. Berlin's observation here is critical: You can't create an entrepreneurial powerhouse without a substrate of small, entrepreneur-driven companies competing to provide consulting and services to the tech companies. It simply isn't possible to direct a one hundred million dollar firehose of capital into a region and drive fast economic growth and innovation. It takes time for the ecosystem to produce a diverse collection of suppliers and consumers. In the Silicon Valley, these companies evolved simultaneously with the tech companies they served.

Of course, entrepreneurs today have an important advantage over Noyce: The global Internet, built on the integrated circuit that Noyce and his colleagues invented at Intel, allows companies to reach across large distances to share work and products with others. In that sense, the Silicon Valley that Noyce invented has become a global phenomenon.

Breathless Google-watching.

2006-03-21T09:21:00.000-08:00

Adam Bosworth, who worked on some very interesting mobility applications during his tenure at BEA, left to join Google just shortly before Google's IPO last year. Adam's a smart guy, and lots of us wondered what he would work on for his new employer. I ran into him shortly afterward at Esther Dyson's PHI Conference in New York; he was mum about his plans then, but I thought it an interesting venue in which to find him.

Garrett Rodgers has an article in ZDNet today on Google Health and Adam's job title in a recent PC Forum attendee list.

Just as I said in October: I believe that personal health information management is currently a mess, and a field ripe for innovation. It will be interesting to see what Adam and his colleagues have cooking.

Data tone.

2006-03-15T08:08:00.000-08:00

Coté at Redmonk led me to news of the latest product innovation at Amazon: Cheap storage.

This is just another development in an accelerating trend: Data tone. People have come to expect that electricity and dial tone are just present in the walls of their homes. When they move in, they expect to plug their blenders and phones in, and to have them work. That's led to all sorts of opportunity for innovation, and all sorts of value to consumers. Standard interfaces and reliable service have allowed companies to build useful devices and offer valuable services on top of existing infrastrcture. Home alarms, voice mail and call waiting, DSL broadband access and more are built on these services.

Amazon's announcement means that you're now able to get storage services in much the same way that you get telephone and electrical service today: It's provided by a capable and reliable utility (Amazon), so you don't have to worry about installing your own data storage devices and keeping them running. This first generation is aimed at innovators and entrepreneurs who build storage-based products on top of Amazon's standards-based service, but those innovators and entrepreneurs will develop the devices and services you'll buy for your home.

Want your computer backed up automatically? No problem, just get the client software product, with strong encryption built in, that moves your files automatically to the Amazon-provided storage service. Not happy with the price you're paying your ISP for email storage? Use the email client plug-in to copy all your important messages to the Amazon store, where you can keep them more cheaply.

Next step in this progression will be computing cycles: You'll be able to use a computer remotely and just pay for the time that you consume. It'll be cheaper and much easier than buying, installing and maintaining a system of your own, especially when you can stop worrying about Microsoft system viruses and all the updates and upgrades that your home system requires. Already companies like Sun are offering on-demand computing to businesses; that'll penetrate to the broader consumer market soon.

For now, Amazon's storage service is metered -- you pay based on space you consume and data you transfer. Over time, I expect that this will move to flat-rate service, just as the mobile and wireline telephone market has for most ordinary consumer uses. Remote computing will follow that same progression as infrastructure is built out, and as competition drives prices down in the market.

There's an important business trend that goes along with the technical trend. You used to have to spend a lot of money to start a company -- you needed people, buildings, IT systems and expensive software packages. Increasingly, the physical plant isn't required. You can buy the IT services you need (as in Amazon's storage offering). You still need people, but the start-up costs are dramatically lower because the equipment is no longer needed. You can pay (very little, in general) for what you need, as you need it.

This trend, toward lower capital requirements by new companies, is happening at exactly the same time that venture capital firms are looking to place investments from very large funds. It'll be harder for investors to find places to put substantial amounts of money, which means that it's a good time to shop for funding.

Disclaimer: Amazon's a big Oracle customer and uses a lot of Berkeley DB. I'm a regular Amazon customer and a big fan of the company. Mostly, though, I'm just enough of a geek to think that this is really, really cool.

Go Bears!

2006-03-14T21:09:00.000-08:00

Victoria punks USC. Man, I love Berkeley!

Airport security.

2006-03-14T07:42:00.000-08:00

I've written about this before. I am on a watch list for airport checkin with Southwest Airlines (but not, interestingly, any other airline that I fly regularly). Having my name on that list has meant I fly the middle seat whenever I fly Southwest, because I have to check in at the terminal. Not deadly, no, but a hassle, and especially aggravating because the additional screening I undergo doesn't actually increase air security.

Further proof of the flaws in TSA screening appear in an article in CSO. Thanks to Bruce Schneier for the link.

Hayward fault helicopter tour.

2006-03-10T06:59:00.000-08:00

Sobering mashup of the day: The Chronicle ran a story today about a virtual overflight of the Hayward fault. Anyone living in the Bay Area knows what this is about. It's a Google Earth mashup with some data from the US Geological Survey.

The project page is here.

Two anecdotes.

2006-03-05T11:23:00.000-08:00

Two anecdotes that I particularly enjoyed from The World is Flat:

TiVo collects information on the viewing habits of its users, and analyzes that information to figure out what its customers do. This infrastructure means that TiVo knows what moment in television was most often rewound and re-viewed: Janet Jackson's wardrobe malfunction at the 2004 Super Bowl.

Later in the book, Friedman interviews Bill Gates, and they talk about the "ovarian lottery." Friedman writes:

Thirty years ago ... if you had a choice between being born a genius on the outskirts of Bombay or Shanghai or being born an average person in Poughkeepsie, you would take Poughkeepsie, because your chances of thriving and living a decent life there, even with average talent, were much greater. But as the world has gone flat ... natural talent has started to trump geography.

"Now," says Bill Gates, "I would rather be a genius born in China than an average guy born in Poughkeepsie."

Flat is the new open.

2006-03-05T09:17:00.000-08:00

I've had Thomas Friedman's book The World is Flat on my bookshelf for almost a year, now. I buy more business books than I read. This one has gotten some very good reviews, though, and I liked Friedman's earlier The Lexus and the Olive Tree an awful lot. I finally got around to reading it this week, and am sorry I waited so long.

There are a few general business books that have swept Silicon Valley over the last years. Shapiro and Varian's Information Rules was one, and Christensen's The Innovator's Dilemma was another. If you're an investor or an entrepreneur in technology these days, you've probably read these. You should add The World is Flat to the list.

Friedman's case is simple to state, and his book argues it compellingly. Technology and economics are eliminating barriers around the world. It is easy to collaborate with people today, no matter where they are. Broadband networking makes it possible to send clerical, accounting, software development and other jobs to the places where they are done most cheaply. The quality of the global workforce is on the rise: India, Pakistan, China, Eastern Europe and South America offer educated and motivated people eager for good jobs. The trend is not limited to desk jobs, either. UPS, for example, has transformed itself from a shipping company to a logistics company, with warehouses and expertise all over the globe, able to help companies with storage and distribution of goods, and with supply chain management.

These forces push in one direction: Companies and individuals must broaden their networks, cooperate with providers and consumers regardless of geography, and open themselves to fast and flexible affiliations that let them do more, more quickly, than ever before.

Friedman, Shapiro, Varian and Christensen all acknowledge that these forces are painful. When jobs are more mobile than people, people suffer. Like the others, though, Friedman makes a key observation: Lower costs and easier collaboration create opportunity as well as pain. If your infrastructure costs go down, you can invest more in innovation on top of that infrastructure. Yes, you need to work harder to find ways to innovate, and yes, you have more competition from other companies in other places. But global collaboration can be -- should be! -- about building new value, and not merely shaving costs. Christensen, in his public speeches, calls this "the conservation of attractive margins."

Tremendously interesting stuff, and well worth reading. But why is free the new open?

Friedman identifies ten trends that are flattening the world. One of them is open source. It's an important trend, but Friedman's description makes clear that it matters most because of the way that the open source community behaves.

After reading his book, I am more firmly convinced than ever of a key truth about open source: The source code is the least important thing about it. Successful open source projects thrive because of the charisma of their leadership, and because of the enthusiasm and dedication of the community members at large.

I believe that open source wins because of easy access to expertise. Most of the key developers of open source are directly accessible to users, and most work hard to stay in close contact with the larger community. That ethic, of low barriers and easy direct communication, is a world-flattener, in Friedman's terminology, and a sustainable advantage to the projects.

Certainly the act of sharing source code makes certain kinds of collaboration easier -- sharing the work of finding and fixing bugs, for example, and giving people the tools they need to add features and extensions. Many in the industry, though, confuse open source with open access. Access to the experts makes software more valuable; access to the source code may or may not.

When I talk about "open", I mean pretty much what Friedman means when he talks about "flat": factors like communication, collaboration, community, feedback, componentization, distribution. I believe that open is strategic, but that source is merely tactical. Competitive organizations need to find ways to be more open and collaborative, to solicit feedback more directly, to work faster and to share more easily. Whether or not there is source code in the pot depends upon the soup.

Many in the open source community deride proprietary vendors for their perceived fear or confusion about giving away source code. Open source projects have pioneered new collaborative strategies that drive success, but those strategies are easily learned. The winners in the future will be those who, like UPS in Friedman's book, transform themselves by adopting openness as a strategy, and who work hard and pay attention to their customers and partners. Proprietary software vendors may well have much to learn, but they not are stupid.

I don't usually post disclaimers on my personal blog, but will break my rule this time.

Because we recently sold Sleepycat to Oracle, some might read this post as a dark prediction on the future of Berkeley DB. It's not. Berkeley DB is tremendously successful. It uses the source tactic to advance the open strategy extremely well. It's open source.

Rather, this post describes -- openly! -- what I think will happen next in the technology industry: Source code will fade in importance, and openness will become critical.

The world is flat, and flat is the new open.

Business 101 on Web 2.0

2006-02-28T07:29:00.000-08:00

Russell Beattie woke up on the wrong side of the Internet on Sunday and wrote about a problem with Web 2.0 hype: the absence of profit.

In fairness to the folks at O'Reilly, where Web 2.0 is a major meme, the platform and the path to profitability really are different things, and can be considered separately. You really do need to grok the technology in order to build a collaborative application, and it's sensible to point some intelligence in that direction. But Russell is, I think, exactly right in his central claim: If you don't also think about how to make money, you're just doing charity work slowly. That's not bad, but it's not sustainable.

It's true that there are interesting Internet applications that have turned into real money for their founders. One of my favorites is Joshua Schachter's del.icio.us, the social bookmarking service that Yahoo! bought late last year. Joshua built and funded the service without recourse to either user fees or advertisers. He sold it to Yahoo!, who wanted the page hits. Bully for him.

But del.icio.us was never, under that model, a long-term sustainable business on its own. If Yahoo! had not elected to buy, the service would sooner or later have been forced to shut down, or to convert to a for-pay model. Because its users were trained to expect something for nothing, it's most likely that any such switch would have alienated many, drastically reducing page views and diminishing even the value that Yahoo! saw in the business.

I like free stuff as much as the next guy, but I don't have long-term relationships that are so one-sided. Sure, I'll use a free Internet service, but I won't rely on it, because I assume that sooner or later it'll augur into the ground. I pay for broadband because I rely on it, and need SBC to stay in business from month to month.

I'm posting this on Blogger, and Blogger is free, but it's owned by Google, which is profitable enough on its advertising business to subsidize all sorts of other stuff. Even so, there's a real risk that one day Eric Schmidt will wake up and ask someone in his office why he is losing money on guys like me. I like Blogger, but I don't rely on it.

Russell's key point, I think, is the difference between a business that's built to grow -- one that has revenue that can outstrip expenses -- and one that is built to flip. Building to flip is the 1990s model of entreprenuership: Get in, get out, and make a killing on the greater fool. Building to grow is objectively better, because that growth creates value not just for the entrepreneur, but also for the customer, the employees and the long-term investors in the business.

Go Russell!

Update: Google in China

2006-02-13T13:32:00.000-08:00

Yesterday, I posted this article, which drew a long but detailed comment that's worth reading. Today, digg took me to a story in USA Today about legislation sponsored by House Representative Chris Smith of New Jersey. The legislation is intended to keep server machines that are owned by US companies, and that store personal information, out of China. The idea is that keeping the machines physically remote limits exposure to seizure by the government.

My initial post was simplistic; my opinion hasn't changed fundamentally, but it has gotten more nuanced. This is legislation that at least makes policy sense, in supporting the goal of protecting human rights. I still believe that singling out technology companies makes no sense. I hope that Congress considers a broader policy. But I have to admit that Smith's bill is aimed in the right direction.

Links for 13 Feb 2006

2006-02-13T07:40:00.000-08:00

Just in time for Valentine's Day: Love that Internet!

Two kinds of office artillery: high-tech and low-tech.
Turn your electric toothbrush into a burglary tool.
Google Earth is on the Mac (and the latest download includes virtual fly-through data for Torino).
A little late, yes, but the 2005 Word of the Year is "truthiness." (Caution, large PDF file)
Less frivolous: What is your dangerous idea?
Universe Today tells you what to look at tonight.

Pointless outrage.

2006-02-12T16:14:00.000-08:00

I've been watching this story for a while, but waiting to blog it until the initial frenzy settled down.

China, a country with a well-documented history of human rights abuses, has imposed some strict censorship rules on its citizens' access to the Internet. China is, simultaneously, under pressure to adopt more democratic political processes. And, of significant interest to companies around the world, China's move toward a market-driven economy has created an opportunity for buying goods from, and selling them to, an enormous population.

That's a complicated mix.

Google, deep in the Internet, has apparently cooperated with Chinese authorities in their censorship of Internet content. That's unfortunate, and is at least a partial repudiation of the company's professed motto, "Don't be evil."

There has been widespread and vocal opposition to Google and to its censorship. Some of the loudest outrage has come from the US Congress, which has criticized the company for "bowing to Beijing."

I'm a believer in the power of the market to effect social change. I was at Berkeley when the movement to divest the University of its holdings in South Africa spread among public and private institutions. That pressure, along with political and other social pressure, eventually led the the dismantling of apartheid as an official policy of the country. Businesses certainly have influence, and ought to use it to advance human rights when they can.

It's unreasonable, though, to expect a business to behave altruistically all the time. This is especially true when other institutions aren't leading the way. There was already plenty of pressure, by ambassadors and individuals, to eliminate apartheid before the divestiture movement began. Businesses supported that pressure, but didn't initiate it.

Google has defended -- or at least explained -- its behavior. I hope that it will do better in the future. I hope that it will help to eliminate censorship and restriction of rights around the world. It has enormous power to do that.

But if there is to be a world-wide movement to eliminate Chinese censorship, it must include the policymakers and governing bodies that wield much more influence than any single company. The US has an enormous trade imbalance with China. We ship lots of raw materials there, and buy back lots of low-cost finished goods. That direct economic support is good for the Chinese people, certainly, but also helps to support the Beijing government.

Congress' demand of Google was, essentially, that it regulate its behavior in China. If Congress expects individual companies to put economic pressure on China, it should show the way. If the regulation of trade is to be a solution, then governments, as well as corporations, need to show how.

Outrage from the floor of the House and Senate are good television, but pointless for directing real change. Asking the private sector to lead the way on a public issue like human rights makes no sense. Sure, let's ask Google to behave better next time. But let's show them how to do it, instead of telling them.

The Sneaky Weasel Theory

2006-01-18T03:36:00.000-08:00

Simon Phipps posted an interesting response to my write-up of the GPLv3 conference. I said that the current draft of the new license appears to take no explicit action to eliminate the ASP loophole. Simon disagreed:

[A]s I have looked at it more and more, I believe Eben and Richard have been far more subtle. A crude and explicit ASP clawback would have raised a riot. Instead, the seeds are sowed in section 5c ... and even more in section 1.

It's important to look as carefully as Simon is doing at the language, and to consider ways in which users, lawyers and the courts might interpret it. Nevertheless, I think Simon is wrong.

The language in 5(c) is simply that the program display notices via some convenient mechanism. That's common in licenses, and the provision is present in the current version of the GPL, in section 2(c). There's nothing new here. Section 1 does introduce some new language, but I don't think it's intended to cover Web interfaces -- rather, it's intended to resolve the long-standing debate on what "linking" means in the GPLv2.

So on the face of it, I believe those sections do important work, but not any work on the ASP loophole.

More fundamentally, I don't believe that the FSF intends to use subtlety and guile to get important changes into this draft. I've certainly seen legal documents that do this; they are usually drafted by sneaky weasels who want to use words to conceal their intent when they are making an agreement. The question is, could the FSF possibly be sneaky weasels in drafting v3?

I believe that the answer is no. The review and comment process is intended to create a large corpus that explains clearly the intent of the drafters. That way, if there are questions as to meaning later, developers, lawyers and the courts can examine the literature and find out what particular words in the language mean.

This comment period will go on for about a year. Simon raises an interesting question; does the FSF intend sections 1 and 5(c) to require ASPs to release their code? There's clear language in 7(c) and a statement from the podium that deals with the point. I expect that the discussion and commentary will make clear that 1 and 5c are aimed elsewhere.

Fundamentally, this process will be too public, discussions will be too open, and the review will be too painstaking to support sneaky weasel behavior. Eben, Richard and the rest of the FSF are smart people. Being a sneaky weasel in this instance just wouldn't work.

We can take the language at face value.

GPLv3

2006-01-17T08:12:00.000-08:00

I'm in Boston for the inaugural meeting for the public review of version 3 of the GNU General Public License. Given the venue, there is less international participation than Eben Moglen and Richard Stallman may have liked, but the turn-out is pretty good. Representatives come from businesses using free software, businesses that produce and service free software, and the broad development community. Participants include HP and Sleepycat, the Apache foundation and individual developers working on small projects. A number of attorneys, from private practices and corporate counsel from technology companies, are here, as well.

All things considered, I think that Eben and Richard did a good job of attracting a cross-section of the interested parties to the kick-off.

The first public draft of the new version is available for public review. It's an interesting document to read, and I encourage you to do that. For the time-pressed or the curious, though, here are a few of my own impressions.

The FSF is clearly taking a much stronger stand against software patents than it did in previous versions of the license. This issue has become increasingly important in the industry. Even Microsoft's source-available licenses dealt with the topic explicitly. The provisions in the new GPL are defensive; someone using GPL'ed code who asserts patent rights for that code against others loses the rights to use. Also, anyone who knowingly distributes GPL'ed code under a patent license that they hold is required to provide, by some means, a patent rights grant for downstream users. This appears to me to be intended to require large companies with substantial patent portfolios to take care of cross-licensing for those who receive GPL'ed software from them. Smaller developers aren't likely to have the tools to obtain blanket licenses, but are much less likely to know that there are patent obligations in code. All told, a good stance for the FSF to take.

Digital rights management -- or, as Richard Stallman puts it, Digital Restrictions Management -- is another major issue for the drafters. They're against it (and so am I). The new draft requires anyone who distributes GPL'ed code that uses DRM to distribute the DRM keys simultaneously. This provision is intended to ensure that hobbyists and others can continue to enhance systems like TiVo, which uses DRM in its most recent releases to prohibit modifications to the software that runs on the systems their customers buy from them. My prediction: The language here will be tricky to get right, as there are lots of parties to consider (author, DRM system vendor, company that builds the system, third-party modifiers and end users), and DRM technology is a moving target. Noble sentiment, some work to do to make sure it plays out properly.

Next, the most anticipated or feared provision, depending upon who you are: The elimination of the ASP loophole. Briefly, the old GPL required you to share your changes if you shared your program. The way you shared programs in the old days was to give copies of them to others. Today, many software packages are "shared" by connecting to a central Web server and using a browser to do things. Application Service Providers, or ASPs, don't ship software, but they still use and enhance GPL'ed packages (sometimes!) and perform the changed version for users.

Many in the developer community want to see those enhancements shared. Many large ASPs want to keep them private. Both sides muster reasonable arguments in favor of their positions. This change has been a lightning rod in the run-up to the new version, and many of us were eager to see how the FSF would address it.

The answer is, they ducked it. There's a provision in the draft agreement -- section 7(d), for those of you reading along at home -- that permits developers to add the requirement to share enhancements that ASPs and companies like them develop, but nothing in the official draft requires that. In fairness to the FSF, the politics on this one issue are tremendously complicated, and the legacy of old GPLv2 code running at ASPs forces those ASPs to make some pragmatic decisions on adoption of new code under the new license if the rules change on them. This topic will be one that gets heavily debated during the comment period.

Section 7 of the new draft is intended to make it easier to build software systems that combine GPLv3 software with other free and open source software. That compatibility is a good thing, but the provisions in section have already raised a lot of eyebrows. There is some concern that allowing people to add conditions to the GPL will confuse, rather than clarify, licensing requirements. For a programmer, the question is, should the license be made extensible (which in my view is what section 7 attempts to do), or should the language permit explicit compatibility with other licenses with certain terms? Stay tuned, this one is far from over.

Now we'll all buckle down for a year and do some work. Feel free to pitch in -- the GPLv3 site explains how.

Links for 7 Jan 2006: Deep stuff

2006-01-07T11:38:00.000-08:00

After some weeks of silence, links for the day. I've collected enough of them over the past two weeks that I can afford to split the load. This batch is the stuff that will make you a better person. To wit:

Bruce Schneier discusses the value of anonymity and how it differs from privacy, in a post rebutting Kevin Kelly's claim that anonymity is bad, with commentary from lots of others interspersed. The key issue seems to me to be balancing individual expectations of privacy with society's need to enforce the law. You ought not to be able to hide while you commit crimes.
Bruce also led me to an article on automating the download of Amazon wishlists, then using the results for data mining. The erosion of privacy is surprising, sometimes. It's getting easier all the time to find, aggregate and analyze what you used to consider personal information. If you don't want it known, you ought to keep it off the web.
Obsessed with digital privacy? Get PGPdisk.
Feeding the geeks: Some quantitative data on the relative popularity of programming languages.
Feeding the geeks a little more: One of the many top download lists for 2005 points you at software you might like, for free.
I wrote about my hybrid Accord a while ago. There is some future-car research, on hyrdogen-powered vehicles, going on at Cal.
Want to know how the search engines rank results? An article on search engine optimization answers that question, as a side effect of telling you how to improve your score.
And the old one on the list: Jon Udell had a good article at the end of November on data synchronization and interchange. Sound dull? This is what Google and Microsoft are working on now.

These are the ones you'll read instead, though, because they're just more fun:

Bump keys, if you don't have the patience to hone your burglary skills over time.
It turns out we are living in the Matrix.
Everbody loves Three Dead Trolls in a Baggie, because the Trolls know that every OS sucks.
Floam at home.
Without Lego, there could be no great art.

Cory Doctorow on DRM.

2005-12-13T10:21:00.000-08:00

I'm at ApacheCon in San Diego, where about 450 ASF hackers are locked in a hotel while the weather outside is spectacular.

Cory Doctorow delivered an excellent keynote yesterday on digital rights management. His views are clear from his writings elsewhere. I agree with him emphatically: Studios and record companies grew up during a time when the way that content got distributed was very different. Ubiquitous digital networking changes the game fundamentally. DRM is an attempt by the distribution companies -- last millenium's winners -- to prohibit new business models around content distribution, so there can be no new winners for the new millenium. DRM isn't good for artists and it isn't good for consumers. It's only good for the middlemen.

During his talk, and in a follow-up discussion we had later in the day, he made a few points that I think are worth repeating here.

First, Cory works with an organization called the Electronic Frontier Foundation to oppose DRM where it crops up. That group has been very active in monitoring and reporting on meetings of the World Intellectual Property Organization, or WIPO. WIPO considers, with some regularity, imposing legislation in different parts of the world so that DRM is required for use of media. The EFF attends WIPO meetings and reports on deliberations (blogging is a tremendous tool for this). Simply by showing up, the EFF has engaged WIPO delegates in a productive conversation, made deliberations more transparent and changed the way that decisions are made. DRM remains a threat and WIPO still considers it, but this is a great example of why you want to work with those whose views you oppose.

In a surprising (to me) contrast, the EFF has had little or no success in engaging Apple on its use of DRM. Apple is a huge distributor of digital media -- everyone's heard of iTunes, and my informal survey says that the iPod Nano is the only thing you need to buy your teenager this Christmas to win his love forever. Apple uses very aggressive DRM technology in iTunes and the iPod. It's impossible, using Apple-supplied tools, to move iTunes songs to competing music players. It's very hard to share content among devices, even if you own all of them. Apple regularly updates the DRM system in its products to eliminate ways that people find to share files.

Let me be clear, here, that I'm not advocating piracy. I'm talking about moving my legally-purchased iTunes recording of Royal Oil by the Mighty Mighty Bosstones from my Powerbook to my Rio MP3 player. I like it, I bought it, and I want to listen to it in the gym, but Apple says no. The EFF has tried to work with Apple to reduce the burden that DRM puts on Apple customers, but to no avail.

Now, Apple builds wonderful technology. I love their computers. An iPod is not merely a nice player, it's a beautiful object. Apple's engineers are outstanding. Many are active in open source projects (in fact, Apple ingested the open source FreeBSD operating system as the basis for its MacOS X platform). I bet that many Apple employees are FSF members.

So how is it that the EFF can engage with WIPO, and can do productive work, but not with Apple? It's really a shock to me. Apple's customers and the people who make the music would be much better off without DRM. Apple's defending last millenium's distributors, but Apple is a this-millenium kind of company. People don't buy iTunes music because it has DRM. They buy it despite the fact that it has DRM, and they hate the way that DRM interferes with their enjoyment of it.

Hey, Steve: Talk to Cory.

The second interesting point from our conversation was about ways to make DRM irrelevant. I believe that the EFF's lobbying work is important, and I'm glad they're doing it, but I don't believe it will succeed in the long term. The only real way to kill DRM is to create viable businesses that don't use it. Build content repositories that contain great music, literature and video, make them available without DRM restrictions, and then find ways to generate revenue for the creators. Steal the audience away from the existing studios and music distributors; show them that they need to abandon DRM or else lose.

I asked about that strategy explicitly. Cory points out that his own books are distributed under Creative Commons licenses, which effectively prohibit the addition of DRM technology. Other artists -- he named several, but the number is large -- are doing the same.

That's great. I'm glad to hear it. I really hope that someone with vision and enough money to do it will create the repository and distribution strategy that draws the audience. Let's get people excited about easily-shared content. Let's make certain that the sharing is legal. Let's find ways to pay the creators of the works. Where we can, let's eliminate the overhead that middlemen impose, driving costs down for consumers and making more great content more easily available.

Links for 11 Dec 2005

2005-12-11T14:42:00.000-08:00

You can never have too many recipes for homemade ballistics gel.
The hardest math class I took in college was Differential Equations and Boundary Value Problems. We spent a surprising amount of our time talking about the ways that Slinky toys move. You can skip the theoretical work, and just go mess with a spring network with damping.
With a name like United Nuclear, you know it's a cool company. With a product like Neodymium magnets, you may be done with your holiday shopping early.
Wonderful photographs from the European Space Agency's Hubble team.
Nature, red in tooth and claw.
Speed up your Windows installation.

A great day in Monterey.

2005-12-11T14:00:00.000-08:00

I spent yesterday morning aboard the Escapade in Monterey for a two-tank dive trip. It was the kind of day that reminds you why we live in California: Brilliant blue skies over smooth seas, with excellent conditions in the water. Visibility at the surface was about thirty feet, and at depth not much less than that.

These two pictures show our mooring at Outer Pinnacles, our first dive site, and the view of open water from the other side of the boat.

I don't have any good shots of our second dive site, at Pescadero Point, but I do have a photo of a happy diver on the trip back home.

Making the GPL.

2005-12-11T06:05:00.000-08:00

I originally wrote this blog entry in early November, but in the press of events, I didn't get it posted. With the recent publication of the official process document by the Free Software Foundation, I remembered that I had this lingering in my drafts folder. Because this includes views from other groups, I thought it would still be useful to put it up.

I was at the GPL 3.0 panel discussion at OSBC this week. The panelists were Mike Milinkovich, executive director of Eclipse, Diane Peters, general counsel for OSDL and Eben Moglen, general counsel for the FSF. It was a collegial group; Peters is on the board of Moglen's Software Freedom Law Center, for example.

In fact, though, this was Eben's panel, and it was his message most of us had come to hear. The panel was convened to explain the process that will be used to produce the next version of the GNU General Public License. I've seen Peter Galli's coverage of the session. It was good, but I want to do something different here. I want to recap the process that Eben described in as much detail as I can, from the notes I took during the session.

What follows is, as nearly as I can make it, a recitation of Eben's (and Diane's and Mike's) points from the panel, with no editorial comments from me. This is a pretty detailed blog post from me, and intended for a specialist audience; if you don't know what the GPL is or if you don't like to read about committee deliberations, you ought to stop reading now.

The current version of the GPL, GPLv2, was produced 14 years ago by a reasonably small group of people led by Richard Stallman. The license has seen very wide adoption since, and GPL'ed software is used extensively around the world. The next version of the GPL will be GPLv3.

To begin, Eben listed four rights that the GPL is designed to protect:

The right to run software without obstruction, and to understand how that software works.
The right to copy software as much as you like.
The right to modify the software, including the right to make private modifications that you do not share with others.
The right to share the software, but to share with all of these rights passed along to others.

These right all accrue to users of the software. The GPL is a document designed to protect users' rights. Eben emphasized that those protections are intended to extend to all users of GPL'ed software, and that explicitly includes commercial users.

In producing a new version of the GPL, which is designed to protect the rights of users, it makes sense to ask users what they want. The revision process will solicit comment by the GPL user base as broadly as possible, including comments from individual software developers, groups using other licenses, commercial concerns, legal experts and others. The process will consult groups around the world.

Eben said several times, forcefully, that this process will be open and transparent. People who want to participate will be able to do so. No decisions will be made in secret. Discussions and decisions will be made public.

Eben emphasized, though, that protecting rights is work for experts, and not for the one billion people who use GPL'ed software around the world. He expects that knowledgeable representatives will participate actively in the process, but that the majority of GPL users won't choose to participate directly.

He also said clearly that this is not a democratic process. There are no votes on GPLv3. Eben expects this process to produce reasoned, intelligent feedback, which will allow the Free Software Foundation to make informed decisions about ways to protect users' rights, but emphasized that the GPLv3 will be issued by the FSF, and not by acclamation.

In the next ninety days, formal review and discussion will begin.

A discussion draft, accompanied by an explanatory document on the language and reasons for changes from GPLv2, will be published.
As many experts as possible will be invited to read and comment on the draft. (A question from the audience asked Eben about the mechanism for participation: On-site meeting? Some other forum? Eben replied that he wasn't able now to describe the mechanism in detail, but that it will allow both in-person and electronic participation, and that "it will be great").
From the collection of experts, committees for discussion will form. These committees should collect around particular points or issues that are important to the members, and where they have opinions and expertise that matter. The committees have responsibilities in the process:

Outreach, to make sure that people with a stake in the process know that it's going on.

Recruitment, to be sure that people who can and should contribute do.

Discussion, to turn comments from reviewers into well-defined issues that can be debated and resolved. If the issue survives this discussion -- if the group believes that it needs the attention of the FSF -- then it continues to the next steps.

Consideration, to explore as thoroughly as possible the different resolutions available for the issues defined.

Resolution and recommendation within the committee, to propose the resolution that makes most sense to the committee for the issues it identified.

Report to the community generally on issues and recommendations.

This process will produce a comprehensive explanatory corpus on the meaning and intent of the GPLv3. The committee documents and recommendations will be public, and will give license users insight into how the license was made.

There will be issues that transcend single groups, or that require attention from the FSF in order to be resolved. In that case, issues will go to the "Supreme Court" of the FSF (Richard Stallman will play a key role here). The documentation and deliberations of the committees will be important in considering the issues, but the FSF will make the final decision on language.

After a few iterations of this process and accompanying draft revisions of the license text, Eben expects to be able to issue a final license for adoption. On issuance, debate on the GPLv3 will end.

Eben's goal is that all GPLv2 code should be able to move to the GPLv3, and he does not want to do anything in the new license that will make that migration impossible.

An audience member asked about plans for revision of the LGPL, and Eben replied that it will be considered in the same cycle as the GPL. For practical reasons, the actual publication and discussion of the LGPL revision may lag the GPL slightly, but the goal is to have both licenses under discussion more or less simultaneously, and to issue new versions of both together.

There were several other audience questions, but I've only captured one in my notes. Someone asked Mike Milinkovich if Eclipse would consider amending its EPL license as a result of changes to the GPL. Mike replied carefully: He would be remiss not to consider the interests of the Eclipse community, and ways that changes to the EPL, in concert with changes to the GPL, could advance the interests of the Eclipse community. However, he cautioned, changing a license is a lot of work, and the EPL recently underwent revision, so Eclipse could well decide to do nothing.

(Violating my no-editorial rule: I think Mike's exactly right. The revision of the GPL presents an opportunity for other open source license authors to look for ways to clean up and reinforce the intent of their own licenses, and to consider whether they want to make those licenses more, or less, GPL-compatible. That's a tremendous effort, though -- just look at the laborious process that Eben laid out for producing GPLv3! -- so we're not likely to see wholesale revision to FOSS licenses.)

In the footsteps of Tom Payne.

2005-12-09T06:56:00.000-08:00

The thing that most interests me about the internet is the way that it creates conversations. The combination of an easy way to publish and good search tools for finding content makes it possible for communities of interest to grow up without geographical constraints. This is a huge deal: Freedom of speech is useless if my speech can't reach your ears.

Digital publishing is great, but I never travel without real books. They're portable, durable and work even in places where there's no broadband or electricity.

I've written elsewhere about search tools for printed matter, like Google Book Search (formerly Google Print -- and the name change is an interesting story all by itself). There are some cheap-and-easy production strategies for people who want to make books, like the ones that Brewster Kahle uses in the Internet Bookmobile. Until recently, though, I didn't know about any simple and accessible do-it-yourself distribution strategy to put self-published material in front of a global audience.

Kevin Kelly's CoolTools site has a great article on how to get your books, CDs and DVDs to buyers by using Amazon.com as your sales channel. The emergence of on-line communities is exciting, but off-line communities are important, too. It's just common sense.

Links for 2 Dec 2005

2005-12-01T19:50:00.000-08:00

Golden rules from people who learned them the old-fashioned way. My favorite is Penn's: Don't make deals with paper. Make deals with people.
Deep cluelessness, phone home.
A bot that digests Slashdot, Digg and Del.icio.us. Now I just need a bot that reads the digest...
This one's for math geeks: The number spiral turns up some surprising patterns.
And as long as we're talking to math geeks: Primes!
The PBS series Science Now has published past episodes online. They're available for download. This is a really outstanding science series.
Lots of chatter about Wikipedia lately. None of it matters much; the site's tremendously useful, and it's good to find ways to use it better.
Remember this post? Perhaps I need to become prime minister of Japan.

Sun rising?

2005-12-01T09:01:00.000-08:00

Sun announced yesterday that it's making its entire suite of server-side software available under an open source license. John Loiacono provides some background. Jonathan Schwartz posted a teaser the day before. In fact, if you read Loiacono's post to the end, you'll see he promises more big news on December 6 -- Sun's beginning to preannounce its announcement dates regularly.

If you've been watching Sun for a while, this is no surprise. OpenSolaris clearly telegraphed the intent. Sun's been talking about releasing Java under an open source license for a long time. Schwartz's blog post links to several other earlier statements that this was going to happen.

So what does it mean?

If you're a customer, it means you can suddenly get Solaris and the applications that run on it under the same terms, and at the same price, as you get Linux and its applications. Sun's using a different open source license than the one used by GNU/Linux, but that difference matters less to IT managers than to guys like me. In practice, if you run an IT shop, you can now download, build and install the server software you need to run your business. You can choose between Fedora from Red Hat (or one of the other freely-available Linux distros) and OpenSolaris. Linux will have to compete more aggressively on features, reliability and scalability with Solaris. Customers can be happy. Competition and choice are good, and the price is certainly right.

Of course, Sun expects to sell subscriptions to IT shops running Solaris, just as Red Hat and Novell do for their Linux customers. Sun clearly hopes that the freely-available software will convince developers to build high-value systems relying on Solaris and Sun hardware, which will drive adoption of the platform more broadly.

If you're a Sun competitor, the picture is muddier.

Among proprietary vendors, Microsoft probably doesn't care much. It already had exactly this headache with Linux, and Microsoft will say that Solaris is competing for market share with Linux, and not with Windows Server. (Sun will certainly disagree). IBM has an established open source strategy and practice based on Linux already, and plenty of muscle in the marketplace to acquire and retain customers. Hewlett-Packard is likely in the worst shape among the proprietary Unix vendors. It has no clearly-articulated open source strategy, depends on hardware sales for a substantial fraction of its revenues, and will have a harder time than ever differentiating itself from IBM, Sun and the commercial Linux distributors. (HP's current struggles are distressing. Let's hope that a great company can recover its earlier success. The garage in which the company began is undergoing restoration now for eventual opening as a museum).

Open source vendors like Red Hat and Novell will be publicly dismissive, but privately concerned, about the development. Without question, Solaris will win some deployments that would earlier have gone to Linux. The key question is how many, and I can't predict the answer to that question. If the losses are mostly to roll-your-own shops that don't buy subscriptions, well, it's a shame, but not a crisis. If Solaris begins to pry away subscription contracts from Linux vendors, though, life will get interesting in Raleigh-Durham and Provo.

If you're a Sun shareholder, you hope this works. Sun's stock closed down twelve cents yesterday after the announcement, at $3.77 per share. The company is a long way into a big transition from a big-iron vendor with a lot of proprietary high-end software, to a nimble player offering commodity systems and freely-available infrastructure on a subscription basis. To make this work, Sun needs to end its reliance on large up-front equipment purchase fees, and to compete on price with vendors of Intel and AMD-based systems. Sun needs also to switch from a model that recognizes large up-front license fee payments for software, and instead move to a subscription-based revenue stream where bookings are recognized ratably over the course of the year.

Even if Sun can reestablish leadership, operate profitably and drive new growth, the cash flow, top line and income in the new business look very different from those of ten years ago. The market isn't yet assigning any significant value to the new operating model. If it begins to drive revenue to the extent that Wall Street notices, the questions will become: What is the appropriate earnings multiple for the new Sun? How does it compare to the multiple for the old Sun? Should a smart investor buy at $3.77? The jury is still out.

For Sun itself, this looks to me like exactly the right move. The decision to put Jonathan Schwartz into the key operational role was a good one. It was time to let a software guy run the company. The widespread adoption of open source development and distribution models meant that the big vendors like Sun, IBM and HP had to figure out a way to participate, before they were rendered irrelevant. Sun's making excellent progress in transforming itself. I hope that it works.

The move isn't without substantial risk. You'll see, if you read Schwartz regularly, that despite my branding him as a software guy, he runs a hardware company. When he talks about value to IT managers, he talks about delivering computer systems that use chips from AMD and Intel, but that consume less power, generate less heat and take less space than competing boxes. Jonathan claims that the main problem in the data center is the electric bill, not the software bill.

Sun's position is interesting. In the 1960s, IBM leased computer hardware to its customers, and installed and managed the software on it for free. Beginning in the 1970s, and continuing until just about today, the model was different: Buy a box, buy some software, and hire some people to keep it running. Sun is returning to a model more like the 1960s. The computers are much nearer zero dollars than they were in the 1960s, but you pay a vendor an ongoing fee to keep your combined hardware and software running, and to make sure you have the latest software all the time. Sun's not alone in this, of course. The whole move to hosted applications and subscription-based IT services do the same thing. Sun is the clearest example of a hardware vendor doing this, though.

Worth saying: I'm not a Sun shareholder. Sun's a long-term and excellent customer of my company. I was an early fan (I was at Berkeley working for Bill Joy when he left to found Sun with Andy Bechtolsheim and Scott McNealy). I have been skeptical in recent years that Sun could remain a long-term leader in a rapidly-changing technology market. Long-term leadership is hard. I'm awfully interested to see what happens next.

Honda Accord Hybrid review.

2005-11-27T13:35:00.000-08:00

Like everyone else in California, I bought myself a hybrid car recently. I have a 2005 Honda Accord Hybrid. This is a larger and more powerful sedan than the Toyota Prius or the Honda Insight. I test-drove those, but I didn't like them very much. They were small, underpowered and felt fragile on the road. The Accord, by contrast, feels solid, handles well, and accelerates respectably when you ask it to.

Of course, I needed some high-tech California personalized plates to go with the car. Fortunately, this one was available. Note the deft combination of my interests in scuba diving and high-performance embeddable database technology.

I used to drive a Subaru Legacy Outback wagon with all-wheel drive. That was our mountain car, and I had bike racks and ski racks mounted interchangeably on the roof. A couple of months earlier, my wife traded in her Toyota Avalon for a Toyota Highlander hybrid with four-wheel drive. That meant that we had two mountain cars. When my Subaru got needy, I decided to trade down to a sensible city car.

I bought the Honda with confused motives. I wanted a car that I enjoyed driving, and I was sure that I would have gone nuts in a Prius or an Insight. At the same time, I wanted to get hybrid mileage. This was about the time that gas spiked above three dollars a gallon, and I wanted to burn less oil.

The Accord's EPA mileage estimate is 37mpg highway and 27 mpg city, but in practice, it does considerably worse. My commute to work is short -- just four miles (I can't bike, alas), so in my first month or so of driving it, I was averaging about 19.5 mpg. For contrast, I used to get 22 mpg in my Subaru! My wife gets around 28mpg in city driving in her Highlander.

Too late, I did a bunch of research on hybrids and gas mileage. Here are the bullet points:

The EPA mileage estimates you see are almost always way higher than the mileage you'll actually get.
Short commutes like mine are the worst on hybrid fuel efficiency.
In order to improve your mileage, you may need to change the way that you drive.

The first bullet bugs me tremendously. If you're from the EPA and you read this post, please fix your mileage estimates. On top of that, there just wasn't much I could do about either of the first two issues. The third, at least, held out some hope for improvement.

The Accord, like other hybrid cars, has a display screen that will tell you what the car is doing while you are driving it. You can use that screen, along with some other dashboard indicators, to understand what kind of mileage you are getting instantaneously, and what your average mileage is.

The Toyota has the same kind of display. (Sooner or later, somene is going to get into an accident caused by staring at the dashboard rather than watching the road.) If you were foolhardy enough to use your mobile phone to take a picture of the Accord's "trip info" display while you were driving, it would look like this.

Pretty fuzzy, pretty hard to read, because it's hard to hold a cell phone steady in a moving car. That blue bar coming in at the top left of the image tells you that the car is getting thirty miles per gallon just at that instant. That's a reasonable number for tooling down a city street. Because of all the stops and starts you do in city driving, it's hard to do much better than thirty or forty miles per gallon in any sustained way on this display. You can do much better on the freeway.

Just a short while after taking that photograph, I pulled into a gas station to refuel, and took this clearer picture of the display. In this shot, you can see that I am stopped (the blue bar says zero mpg), that I have averaged 23.5 mpg for a total of 301.3 miles on the trip, and that the motor ran for eleven hours and four minutes.

Twenty-three and a half miles per gallon is about four miles per gallon better than my average, up to this point. I did that much better by using the fuel efficiency display on the dashboard, and by driving to maximize my instantaneous mileage. For three hundred miles, I had that display on all the time, and tried to make that blue bar as long as I could. I made much slower accelerations from stoplights and stop signs. I tended not to accelerate or brake nearly as much in ordinary driving as I had before. Those are arguably good driving skills to have, anyway. It really does take some attention to change a lifetime of driving habits, though.

Even with those changes, I wasn't able to get near the 27mpg that the EPA promised me I'd get in city driving. My record for highway driving is 35mpg, nearly the 37mpg that the EPA advertises for the Honda. I was only able to do that well by paying careful attention to instantaneous fuel efficiency on the display. Cruise control helps here.

When I researched the car before buying, I learned that Honda's hybrid technology is different from Toyota's. Honda calls its system "Integrated Motor Assist", or IMA. In a Honda, the gasoline engine is always running, and the electric motor kicks in occasionally to deliver extra power. In a Toyota, by contrast, the system is called "Synergy Drive." A Toyota will run either on just the electric engine, or on the gasoline engine, depending on what the driver is doing. My wife backs her Highlander out of the garage on just the electric engine, totally silently. I believe that this difference is a big advantage for the Toyota. Certainly my wife's much better mileage in the much bigger car suggests that there are advantages to Synergy Drive.

Both makes do obvious things, like stopping the gasoline engine when you brake at a stoplight. The braking energy charges the batteries for the elecric engine.

I don't mean to whine, here. I like my car. Having the electric assist gives me lots of power on the highway for passing and acceleration. It's like driving a V8 engine with V6 fuel efficiency. After owning this one for a few months, though, I wouldn't recommend to others that they buy a Honda hybrid. The gasoline mileage is really disappointing, and that is a primary concern for anyone who buys a hybrid.

Happy Thanksgiving!

2005-11-24T08:37:00.000-08:00

The US shuts down today for the Thanksgiving holiday. Jon Carroll explains why.

Aurora borealis in central Nebraska.

2005-11-19T10:22:00.000-08:00

I've never seen the Northern Lights. That's still a goal. It's an interesting phenomenon. High-energy particles that make up the solar wind are captured by the Earth's magnetic field at the poles, and travel down the field along lines of magnetic force. The ions collide with atmospheric gases and produce a glow. Before the science was sorted out, folkloric explanations included fires started by foxes, an earth-heaven bridge for gods, or a game of football played with a walrus skull.

Because the Earth's magnetic field is a polar phenomenon, you generally need to be pretty far north (or south) to see an aurora. Last May, however, a photographer in Blair, Nebraska shot some amazing pictures of the lights from a rooftop. Blair is about 30 miles north of Omaha, halfway to Tekamah. It's further south than Chicago, and much further south than Boston.

Links for 13 Nov 2005

2005-11-13T15:34:00.000-08:00

A study on tin foil hats for mind control protection, and a rebuttal.
George Dyson tunes up his Google talk for Edge: Turing's Cathedral. Dyson's text actually starts a screen or two down. It's long, but fascinating, and as Tim O'Reilly notes, it has a tremendous punchline.
A Future Timeline of Humanity and the Universe. The future's so bright you have to wear shades, until about 10¹⁴ CE.
Only four grams!

Sony's rootkit DRM.

2005-11-11T14:59:00.000-08:00

There's a ton of coverage of Sony's ill-considered DRM software (here, here, here and here, just to get you started). I have nothing to add to the debate, but I'll concur with a popular view: Trying to use technology to solve a legal problem is just as dumb as trying to use the legal system to solve a technical problem.

Need proof? Here's an interesting piece by Declan McCullagh, in which he points out that (a) it was probably illegal for Sony to install the software on users' computers, and (b) it's probably illegal for users to uninstall the software once it's there.

Sometimes the absurdity just reaches right out and whacks you.

[UPDATE on 12 Nov 2005: Microsoft's anti-malware team announces that it will remove Sony's rootkit from infected computers. It would be interesting to sit in on some meetings in Japan this coming week.]

The right kind of greatness.

2005-11-10T20:13:00.000-08:00

An old friend with close ties to the Chabot Space and Science Center in the Oakland hills send me an intriguing email message yesterday: A private showing was scheduled for this afternoon at Chabot's new Digital Planetarium. Entrance was restricted, but she could get me a pair of tickets if I wanted them. She thought I might, because Dr. Stephen Hawking would attend, and she knows what a groupie I am of Dr. Hawking's.

I snapped up the tickets -- I decided to take my daughter, who is approaching the age where girls stop thinking science is cool, and who needs some amazing experiences to keep her head right on the topic.

It was magical. The show in the planetarium spanned 10³⁰ meters, or 65 billion light years: from the surface of the Earth to the edge of the Universe. The journey was projected on the dome roof using a custom projection system built from a collection of ordinary PCs with graphics cards, reading a database of the positions of celestial objects managed on a clustered file system, with the whole thing driven in real-time from a console.

Doesn't sound very sexy, does it? Try it this way: Galaxies went zooming past our heads. We visualized the sphere, at about 75 light years out, marking the boundary of radio emissions from the Earth, and saw the stars in that sphere with planets around them. Is there intelligent life on those planets? Do they know about us from our radio smoke? If they do, I wonder what they wonder about us. We saw the large-scale structure of the Universe, both in the distribution of galaxies and in the cosmic microwave background radiation. Now that's sexy.

We saw Dr. Hawking arrive and leave. We were waiting in the lobby when he passed us on his way into the planetarium in his wheelchair. After the planetarium show, he offered just a few comments -- he had a talk tonight to give at the Paramount Theater in Oakland, so stayed for just a short while. His comments were interesting: If, he pointed out, the atmosphere were just a little cloudier, or if our eyesight were just a little poorer, we would never have divined the deep laws of physics, and might never have developed the science that marks our age so profoundly. Our understanding of physical law derives from our understanding of gravity, and it was Galileo and Kepler and Newton who conceived that understanding, from the motion of stars and planets in the night sky. It was an unusual insight: because we can only conceive of those things we can see, we need some kind of radio smoke ourselves, if we are to learn about the Universe we live in.

My daughter and I had a chance to talk about Dr. Hawking on our drive home. Some people are famous for being rich or pretty to look at, but they don't matter very much. Some, like sports stars, are famous for their skills, and those skills were developed by hard work. Whether they matter in the large sense or not, their dedication and effort demand respect. Dr. Hawking is famous better reasons. His is among the finest minds of our generation, exploring the Big Bang and black holes and the nature of time from evidence most of us can't see. His continued efforts in the face of the progress of ALS are an inspiration. It's impossible for me to imagine what it would be, to be locked inside a body that refused to act, with a mind that worked perfectly. That refusal to despair is the right kind of greatness.

After his remarks, Dr. Hawking left to prepare for his speech at the Paramount, and my daughter and I hit the hors d'oeuvres bar. Chabot opened two of their three telescopes -- the eight-inch refracting telescope (built in the 1880s!) and their 24-inch refractor. We saw Mars low on the horizon through a soupy atmosphere, and the moon higher in the sky and considerably sharper because of it.

We got our amazing experience.

Lessig on ICANN and TLD policy.

2005-11-08T13:07:00.000-08:00

This is interesting. Larry Lessig surprises me. He argues first that splitting responsibility for TLDs among different DNS roots wouldn't cause any technical problems in address resolution, so taking exclusive jurisdiction away from ICANN wouldn't be a problem. He follows that with a second observation: That ICANN has developed a practice of exerting as little policy guidance as possible, concentrating instead on a narrow technical program to satisfy its mission. On that basis, he says, adding TLD servers would be a bad thing. If a new TLD administrator decided to lobby for legislation on content or transmission, it would really mess up the Internet.

Before reading it, I'd have taken the exact opposite position: Splitting address resolution among TLDs requires too much coordination, and ICANN's execution has created much too much concern among stakeholders about its intent. It may not make policy, but it has seemed so far willing to drop its administrative knitting and run off in search of policy battles to fight.

But Lessig makes a good argument. I'm not entirely persuaded -- I think his confidence in the technical administration of DNS roots is too high. If a new top-level administrator wasn't able to properly decide what names it owned and what it didn't, the rest of the Internet, wherever addresses got resolved, would suffer.

Nevertheless, it's a good article, with some interesting reasoning. Lessig has an insightful take on the political maneuvering behind the scenes, and the extent to which US foreign policy generally, and the confused mission of the UN, conspire to make fast, smart decisions impossible. I agree on one point strongly -- with authority split among WIIS, WIPO and others, the UN looks to have terrible execution problems in any real governance role.

Links for 8 Nov 2005

2005-11-08T07:26:00.000-08:00

The military applications of Silly String.
Optical Illusion o' the Day: Tom Stafford and Matt Webb explain the mechanism for this effect in Mind Hacks. Movement processing happens earlier in the visual system than does color processing, and movement processing is more effective at the periphery of your vision than is color procesing. Evolutionarily, it makes sense to notice something at the edge of your vision that's moving, and not to worry about what color it is.
More visual system effects by Stuart Antsis at UCSD.
Buzz Game: Old news by now, but worth a look if you haven't seen it.
From Kevin Kelly's Cool Tools: Shapelock and Friendly Plastic. Dave Boulton says: "ShapeLock and Friendly Plastic are repackaging a polyester-like stuff called polycaprolactone (PCL). This particular formulation is sold by Dow as TONE P-767. It is inert, FDA approved bio-compatible, and biodegradable. Tensile strength is around 3500 PSI. Melts at 60 C,and is WAY fun to play with."
More cool tools: the car chip.

Wonderfest this weekend!

2005-11-04T14:59:00.000-08:00

Late notice, but in case you're in Berkeley (Saturday) or Palo Alto (Sunday), you may want to carve free some time for Wonderfest. Dan Wertheimer of Project Seti@Home sent me a note to let me know it was happening. Looks like a great day, and a fitting way to remember Carl Sagan.

Update: Sound and fury

2005-11-03T16:33:00.000-08:00

I expected to get a few corrections when I posted this, and I have. Let me shine the bright light on a few factual errors in the original post.

First, I was flat wrong: Oracle 10g Express isn't an Oracle Lite derivative. Express is a limited version of Oracle's 10g product line. The Express version pulls out certain features, including support for spatial data, grid deployment, some enterprise management tools, and others. More significantly, deployment is limited to single-CPU systems managing no more than 4GB of data. Total cache size is limited to 1GB.

I guessed early, and I think my guess was pretty clever. Too bad that's not enough!

Second, I said that 10g was a preannouncement, not available. Not quite right. A beta version is available for download, but the production version won't be ready until later this year. I've not downloaded and test-driven the product myself, but I understand that there were a few glitches in the beta installation that were relatively easy to work around, and that the binary then installed and worked as advertised.

Finally, Rex Wang commented on my original post that Oracle's move was the same as an earlier announcement by Sybase. He was right, and I ought to have pointed the same thing out. In fact, if any company should look on this move with concern, Sybase should. If there are two no-charge, low-end databases available from big relational companies, you'll probably take the one Oracle offers.

My original conclusion stands: It's community, not price, that will determine the success of these products in the market. I believe that the open source players are much more successful there already, and I am deeply skeptical that free beer builds lasting relationships.

Sound and fury signifying what, exactly?

2005-10-29T09:03:00.000-07:00

About a month ago, Oracle Corporation announced that it had acquired Innobase, a tiny Scandinavian technology company, for an undisclosed sum. Oracle has lately been on an acquisition tear, so a small buy like Innobase would have passed unnoticed were it not for the fact that MySQL, a strong emerging competitor to Oracle, used Innobase technology in one of its products.

Discussion of the move swamped the mailboxes of open source execs and the commentary on Slashdot. Oracle had moved against open source competition! There was all kinds of speculation on what would happen next, but the smart money was betting we'd hear more from Redwood Shores.

On Friday, Oracle made another announcement, aimed expressly and unapologetically at MySQL: Oracle plans to release a version of its database, to be called Oracle 10g Express, at no charge, for low-end use. Andrew Mendelsohn of Oracle mentions MySQL by name in explaining the company's decision to release the new product.

What does this announcement mean? Has Oracle finally figured out open source software?

There's no question that the Innobase buy was an entertaining and well-played move by Oracle. The news broke simultaneously with a new product announcement by MySQL, distracting press and customers for a couple of weeks. Oracle has gained some influence -- though not, in fact, control -- over a small piece of technology that MySQL uses in its current software release. Instead of simply continuing to use that technology, MySQL must now consider whether it should replace it, offer alternatives to it, or continue to work with its new partner and long-time competitor.

None of that has anything to do with open source at Oracle, though. Note that there is no credible announced plan to continue to develop the Innobase technology as open source software. It's not that Oracle has said it won't do that; it's that open source, as a development and distribution strategy, isn't even among the talking points at Oracle. Mendelsohn has said only that the Innobase technology might be used in future Oracle products. That's an awfully faint endorsement of the team and technology.

What, then, of Oracle 10g Express?

As is true for any preannounced product, it's impossible to say for sure what this is, or what it means. Since no real product is available for customers to evaluate and use, we should assume that this announcement is intended to confuse MySQL's customers, and not to inform Oracle's.

Nevertheless, here's what I think is happening.

Oracle has had a product called Oracle Lite on its price list for a long time. This software, originally developed as a desktop database, has never gotten any serious attention from customers. If Oracle's field sales team hasn't been able to move it, you can bet that it's very hard to sell.

I predict that Oracle 10g Express is a rebranded release of Oracle Lite. Oracle is making a virtue of necessity. Since it makes no money from this low-end product anyway, it may as well give it away for free. In the worst case, nothing is lost. In the best case, Oracle slows adoption of competing open source products like MySQL and EnterpriseDB.

Don't confuse the announced 10g Express product with Oracle's existing Oracle Express Server product line. The existing product is a data warehousing and mining package for Oracle's biggest customers. The new announcement has nothing to do with it.

Oracle's positioning of 10g Express reinforces my conviction. Instead of talking about enterprise adoption, where open source databases are growing most quickly, Oracle focuses on student projects and low-end users. The database is limited in the amount of data that it will store, and in other ways. This is all consistent with the Lite product, but not with Oracle's workhorse enterprise offerings, which are much too valuable to the company to give away for free.

Once again, I'm impressed by the move. It costs Oracle very little, attracts a great deal of attention, and forces competitors to respond, at least in the press. In the long term, though, it's not going to work.

Here's why.

Oracle is confusing free beer with free speech. The success of open source software is only loosely tied to its zero-dollar price tag. In fact, most of us in open source realize that a realistic accounting of software cost has to include the expense of ongoing maintenance. Just because a product is free to download doesn't mean that it is free to use and manage.

What really drives adoption of successful open source is the size and enthusiasm of the community surrounding the software. Products like Berkeley DB, MySQL and PostgreSQL are successful primarily because of the vibrant developer communities that adopt, support, extend and evangelize them.

The market hasn't ignored Oracle Lite because of its price tag. A few hundred dollars is't enough to make a difference, even to most individuals. Two things matter: quality, where past adoption is a meaningful metric, and the size and activity of the global development community that uses the product.

Oracle needs such a community around its products if it is to compete with open source products. The way to build that community is to release the complete source code of a powerful and useful product, so that smart software developers around the world can learn how it works and make improvements to it.

Even then, building a community is a tremendous challenge. Computer Associates' attempts to build community around the old Ingres software have largely failed. Long-term successful vendors of proprietary software have been closed and exclusionary for years. Opening their development and strategic teams, and taking advantage of the power and intelligence of the global Internet community, is very hard.

So what does this sound and fury signify?

Clearly, Oracle has joined IBM, Sun, Microsoft and others in paying real attention to open source software. We'll certainly hear more from Redwood Shores in the months to come. Fundamentally, though, neither of these developments suggests that Oracle understands open source. It's looking for ways to beat open source. That's not going to work.

Hacking the tax code.

2005-10-25T19:14:00.000-07:00

Mitchell Baker posted a good article to her blog recently about the relationship between Mozilla Corporation (of which she is president) and the Mozilla Foundation (of which she is a board member). I want to write about that relationship here, because it's a great example of the old-school hacker ethic.

If you don't know the Mozilla Foundation, think Firefox and Thunderbird. These code lines are descendants of the Netscape browser released under an open source license in the late 1990s. The Firefox browser in particular is enormously popular -- Slashdot reports that more than 100 million copies have been downloaded. Firefox is an excellent web browser with plenty of great features and lots of useful plug-ins. It's very fast at rendering pages. If you're still seeing advertisements on the web pages you visit, well, all us Firefox users are over here laughing at you.

The Foundation is unabashedly about programming in the public interest. It exists to create, maintain and enhance Firefox and Thunderbird, because the world deserves a decent browser. Mozilla is the sort of open source project that gets economists scratching their heads: Talented engineers donate their time to do hard work for no money and build great software that you can have for free. Magic!

Mozilla Corporation was created by the Foundation on August 3rd of this year. The Corporation exists to productize and distribute the open source packages. So why bother having a Corporation, if it's just shipping what the Foundation builds?

Well, US tax law makes it hard for the Mozilla Foundation to be dynamic and agile. In technology, you often need to strike partnerships, build up funding, make investments, enter into business deals and pay for work from consultants and contributors. The Foundation is a not-for-profit organization established under section 501(c)(3) of the US tax code. The details of the code are complicated, but the effect is reasonably simple: The Foundation isn't allowed to earn a profit from its operation. In exchange, it never has to pay taxes.

Paying taxes in the US, especially for a corporation, is a lot of trouble. It's expensive to do -- you generally can't fill in the forms yourself, but need to hire a professional accountant to do it for you.

The Foundation, focused on building software, neither needs nor wants the hassle of quarterly tax filings. On the other hand, in order to make the Mozilla project really successful, the Foundation needs to be agile. It needs money. It needs the freedom to use that money in complicated and interesting ways, without worrying about whether it has transgressed some piece of the tax code limiting the activities of non-profits.

And that's where the interesting old-school hack comes in.

By hack, here, I mean a really clever and novel solution to a problem. The term has lost its meaning lately in the popular press -- what we used to call pirates and crackers and script kiddies are sometimes referred to as hackers, now. When I use the term, I mean the same thing that Steven Levy did. Self-modifying code on early generations of computing hardware? Sweet hack. Woz's floppy disk controller design for the Apple II? Unbelievably cool hack. Building a golf course in Scott McNealy's office on April Fool's Day? Hack!

What Baker and the other Foundation board members did by setting up the Corporation was a really neat hack. Corporations exist to create value for their shareholders. They work hard to make money and invest it in ways that maximize that value. The only shareholder in the Mozilla Corporation is the Mozilla Foundation. The Foundation's mission is to serve the public good by promoting the development and adoption of the Mozilla code base. There's just one way for the Corporation to create value for its sole shareholder: Invest to support the mission. Fund Mozilla's programming in the public interest!

The Corporation has much more freedom to earn and manage capital than the Foundation has. It can sell boxed sets of software. It can negotiate partnerships and development agreements with big platform vendors who want to distribute the Firefox browser on their computers. It can enter into service agreements. Granted, the Corporation has to pay taxes, but it presumably earns enough from its various activities to do that. The Foundation is insulated from the hassles of corporate tax law, and is free to focus on collaborative development of great software.

This adaptation of corporate behavior to serve the interests of a non-profit, by creating a combination that mixes them, but which allows each to behave exactly as it normally would, is really clever. Corporations sometimes create foundations in order to channel their charitable activities and shelter them from tax law. Foundations creating corporations -- well, that's much less common. It's a pretty good hack.

Links for 23 Oct 2005

2005-10-23T08:34:00.000-07:00

Science is fearless: Park and Newman aim to fix the BCS.
Read a page a day of Leonardo's notebooks over RSS.
Eben Moglen is on a mission.
Fragmented attention is the flip side of always-on, easy communication. This NYTimes article is by Clive Thompson, whose blog deserves a spot on your subscription list.

Web 2.0.17r3?

2005-10-21T08:22:00.000-07:00

Most buzz-heavy tech neologisms bug me. I can't stand the word "webinar." In the late 90s, it was "e-everything" -- e-commerce, e-publishing, e-nough! I'm even tired of iStuff from Apple.

Most recently, Web 2.0 makes me cringe. It's not that I don't like Google Maps. I love Google Maps. I just don't know anything new from a term like "Web 2.0". It scares me a little bit: Did the old one stop working? Will there be a patch release or service pack for the 2.0 version in three months? What happened?

Joel Spolsky has posted an article to his blog that explains why Web 2.0 is the wrong thing to say. There was no New Economy in 1998, and there is no new web today.

You have the right to make copies.

2005-10-19T05:31:00.000-07:00

Copyright law and software licensing have been top-of-mind this week. After a considerable gestation period, O'Reilly has published Open Sources 2.0. I wrote Chapter 5, which explores the interaction of business models, copyright law and software licensing practices in an open source business like Sleepycat, so I am glad to see the book in distribution.

The 2.0 book follows an entirely different book, Open Sources: Voices from the Open Source Revolution. The first book was published in 1999, and explored the underlying politics and technology of the open source movement. The new book focuses on how open source has transformed business and the economic systems on which they rely.

The original book is published on-line, and you can read it right now if you like. I enjoy the debate between Linus Torvalds and Andy Tanenbaum tremendously, and re-read it about once a year.

The new book is 445 pages of collected essays, and my freebie arrived in the mail just Monday, so I have only skipped through it lightly so far. I particularly enjoyed Wendy Seltzer's chapter on copyright politics. She articulates the argument for copyright limitations very well; Larry Lessig and others have been making the same argument for a long time, and it's good to see the case made so clearly in the new book.

This morning, I saw Jason Matusow's announcement that Microsoft has published three new source code licenses. Microsoft assiduously avoids the term "open source" for these, calling them "shared source" instead, but a pragmatist is more concerned about effect than naming. In effect, these licenses look a great deal like open source licenses the community understands.

The Microsoft Permissive License resembles the Berkeley Software Distribution (BSD) license. It permits proprietary and open source use, modification and redistribution. The Microsoft Community License has an effect similar to the GNU General Public License (GPL). Like the Sleepycat Public License, which was designed to do the same thing, the Microsoft Community License is reciprocal, and requires sharing of modifications. Like the Sleepycat license, it's considerably shorter and simpler in construction than the GPL. The Microsoft Reference License is most like Microsoft's earlier source-available licenses; it's effectively a look-but-don't-touch license, allowing selected partners and customers to view Microsoft source code, but strictly limiting rights to modify or distribute copies.

One significant addition to Microsoft's licenses, compared to some older open source licenses, is the explicit grant of patent licenses. This reflects more recent thinking on intellectual property ownership, and builds on work done at Sun, IBM and elsewhere on patent licensing for open source software. Patents are a big issue in the debate over intellectual property rights and open source. It's good to see that Microsoft has considered them.

There will certainly be blowback in the community. Disdain for Microsoft is reflexive among some open source proponents. I can't speak to Microsoft's motivations in creating these licenses, but it's a safe bet that the company believes that it will derive tangible benefit from them, or it would not have spent the effort to produce them. I am not personally bothered by Microsoft's success, and I am always glad to see smart people do interesting things.

What is, to me, most interesting is that Microsoft has chosen to create these licenses at all. It is a tribute to both the strength of open source in the industry -- its success just can't be ignored -- and to Microsoft's legendary agility and willingness to adapt. Everyone has heard the story of Bill Gates' whole-hearted embrace of the Internet in the middle 1990s. That was certainly good for Microsoft and its customers. It's much to early to say that this is as significant a shift in philosophy, but this is the company that just a few years ago viewed open source as a threat to its existence. Just by publishing these licenses under its own brand, Microsoft has moved a tremendous distance quickly.

The real test, of course, is what software Microsoft makes available under each of these licenses. Open Sources 2.0 makes the case strongly that open source is about collaboration, so that widely distributed people can create valuable artifacts by working together. To the extent that Microsoft embraces this philosophy, and to the extent that it gives its customers an opportunity to work together in that way, this is an exciting development.

Good job, Jason.

Links for 12 Oct 2005

2005-10-11T06:42:00.000-07:00

I love this Internet thing. I really think it's going to catch on.

The only way to see the City

2005-10-10T06:29:00.000-07:00

Sleepycat is having its sales kickoff meeting this week. We have people in town from the UK, Spain, Germany, and from Seattle and Boston, in addition to those of us who live in the Bay Area. It's Fleet Week, so the Blue Angels were flying (as was Oracle, by the way -- I didn't know about this until I was putting together the content for this blog entry, but I'm amused that the US Navy and a database company were patrolling the skies). The Bay was packed with boat traffic and the ferry services were all running late, but the weather was gorgeous.

We took a really great tour, offered by the San Francisco Electric Tour Company. These guys take groups out on Segway Human Tansporters -- the Dean Kamen two-wheeled scooters. They were a lot of fun to ride.

Top speed on these scooters is eighteen miles an hour, according to our guides. The scooters have lots of sophisticated software controls, including speed governors that are tied to specific keys you use to start your scooter. As novices, we were issued four-mile-per-hour and eight-mile-per-hour keys. By the end of the two-hour tour, we were all feeling bulletproof, gunning our scooters at top speed and cornering sharply. Probably a good thing we didn't have the full-speed keys.

I had planned to take lots of pictures, but it turns out you can't shoot and drive simultaneously, so I got only a few. They're here.

If you want to know what to get me for Christmas, look no further.

PHI conferece addendum

2005-10-06T07:48:00.000-07:00

On Monday, I wrote up the PHI workshop and mused on identity management.

Dick Hardt from Sxip delivered a fantastic keynote at Web 2.0 that is exactly on target. I missed the conference; found the link on O'Reilly Radar.

Gotta love the blogosphere.

I Zimbra!

2005-10-06T06:39:00.000-07:00

Lots of chatter about the launch of Zimbra, formerly Liquid Systems, just lately. There's plenty to like about the company. The principals are good people and plenty smart, they're working on a very important problem -- client-neutral, shared calendar and email support -- and they named their company after a Talking Heads song from the best album of 1979.

Most of what I do every day is driven by my on-line calendar and email. I'm constantly frustrated at the difficulty of working with other people and sharing information. The people I work with use lots of different platforms -- Windows, MacOS, Linux, BSD variants -- with lots of different client software. It ought to be easy for us to share calendars and schedule meetings, or to get to our email over the Web. It's not. In fact, just this week we had a mail server failure at the office that shut us down for half a day. Scary stuff when you rely on email as much as we do.

So I want Zimbra in the worst way.

I've taken the test drive. The Ajax client is responsive. Google Maps really taught the world a new way to build thin-client apps, and Zimbra's learned the lesson well. The UI is well-designed and intuitive.

As much as I like the company and product, though, the software needs some work before we'll be able to replace our current systems with it. The main problem is that it is missing the critical shared calendar features that we need. Right now, you can't share calendars with other users and you can't synchronize your calendar with your PDA. That's my worst problem with our current systems, and until there's a good client-neutral solution, the pain of switching from the devil we know is just too high.

I'm sure that the Zimbrainians will have this stuff working soon. Their forum posts say they're already working on the features. I'm looking forward to my next test drive.

Links for 4 Oct 2005

2005-10-04T05:24:00.000-07:00

Three interesting sites from the PHI conference that didn't make it into yesterday's writeup:

RealAge, an online service that quizzes you about diet, exercise and other health habits, and tells you whether you're older or younger than you ought to be.
DNA Direct, an interesting personalized health service that, for a fee, produces a DNA sequence for an individual and scans it for specific marker genes related to particular diseases like breast cancer and others.
The Body Journal, a tool for managing personal health information for you and your family. Keeps track of illnesses and treatments for different people, and allows you to publish data to a secure site for access by your doctor or other caregiver.

My summer lake reading this year was Jeff Hawkins' book On Intelligence. This will probably be the best book I read this year when December winds down. I learned today that Hawkins' research institution, the Redwood Neurosciences Institut e, has moved from the Peninsula to UC Berkeley (Go Bears!). I hope to make it to a few of the seminars. Hawkins also has a venture-backed private company working on the theories he espouses in his book; that's Numenta.

I've not gone crazy about Sudoku the way that some of my friends have, but I think the puzzles are interesting and a lot of fun. They're a sneaky way to teach logic to kids; they look like simple number games, but it's really not about the numbers at all. Robert Woodhead has published Sudoku Susser. It's a useful tool for learning the technique of solving puzzles, and a good cheat if you're stumped by a tough one.

Esther Dyson's PHI conference

2005-10-03T06:00:00.000-07:00

I spent all day Friday, September 30, at the Personal Health Information workshop, hosted by Esther Dyson of Release 1.0, at the New School in New York City. It was interesting and a lot of fun.

I am a database guy, not a health systems guy. I signed up for the workshop mostly for mercenary reasons. My company makes data management tools; if there is some new work going on in the health field that will require a lot of new personal health information to be stored, then I figured I should find out about it. I learned a lot, so satisfied my main goal. More importantly, though, I came away with a better sense of where health delivery systems, at least in North America and most of Europe, are likely to go over the next decade.

A "Personal Health Record," or PHR, is a record that captures the interesting information about you and your medical history. It records, for example, allergies, medications that you take, current and past diseases or injuries, and information about any treatment you're receiveing, or have received. Obviously, a PHR is useful to a doctor who sees you for a normal checkup, or during a visit you make because you're sick or hurt. A PHR would be a tremendous advantage to emergency services personnel treating you for a serious injury if you're unconscious.

The conference attendees included a number of physicians, some practicing and some who had moved into public policy or entrepreneurial jobs. The strong concensus was the PHR is critical to the future of health services delivery. Several stated baldly that people are dying today because they get no care, or the wrong care, because doctors and EMTs don't know enough about their medical status or the medications they use.

I'm not a health services guy; I can't prove that claim. I do know that the understanding of, and treatments for, many diseases are getting more sophisticated. Drug interactions are increasingly complicated. The availability of individual genomic information, and a steadily increasing understanding of the genetic basis for many diseases, mean that future prevention and treatment will rely on a doctor's access to the actual nucleotide sequence for a patient.

Obviously, there are a lot of important social policy issues that need to be addressed in order to make PHRs work. An individual's right to privacy is one. At the same time, doctors and others need access to the record in emergencies. Insurance companies ought not to be able to deny coverage to someone based on the contents of his PHR.

Though no one at the workshop said so out loud, the PHR is just another form of digital identity. Much work has been done on building digital identity management systems already (see, for example, directory servers and the Liberty Alliance). The technical infrastructure for managing digital identity is already sophisticated. Granted, a PHR has more, and more complex, data (nucleotide sequences, medical images, lists of medications and dosage frequencies and more) than does the list of preferences you have for your Gmail account, but those differences are technical, not fundamental. We should expect to see early PHR systems build on the digital identity infrastructure already used to store and manage other kinds of identity.

There was a good deal of discussion at the workshop on who would store PHRs and where they would be kept. My bet is that there will never be a single central repository of patient identity; rather, your PHR will be broken into pieces, with each piece stored and owned by the person or organization who cares most about it. An insurance company and a patient are interested in different pieces of a PHR; ownership will split along organizational lines. A chronically ill patient under the ongoing care of a doctor needs much more frequent access to his rapidly-changing PHR than does a healthy patient who sees her doctor once a year; the chronically ill patient is likely to have much more control over his PHR.

Since PHRs need to be readily available, it's reasonable to assume that they will be on-line, protected by access restrictions and encryption. I expect that independent services will spring up to store and manage these records, or at least the parts of them that are of most concern to patients. I would not be surprised to see Google, or some competitor, offer this service. PHR is digital identity, and the big online service providers are good at managing digital identity.

Already there are analogues for the PHR. In the US, your insurance company knows a lot about your medical history, because it processes claims that doctors and hospitals make about treatments they provided to you. The insurance record concentrates on services, procedure codes, allowed fees and accounting for payments, but there is a lot of information there about you. Your doctor knows when you came in and what happened during your visits. Your pharmacist knows what drugs were prescribed for you, and what your dosage and course of treatment were. Not all of this information is in electronic form today, and tying it together is often impossible, but that will change.

The PHI workshop was exciting: New trends in computer and medical technology are driving us toward big distributed systems with important privacy and public policy issues that have to be solved. There are plenty of smart people working on this stuff; I bet that the demos at next year's workshop will be good.

Why subclock?

2005-10-02T17:12:00.000-07:00

I've been meaning to set up a blog for a long time, but kept putting it off because I thought that it would be too much trouble to get started. Stephen O'Grady published an article on starting a blog in his blog and proved me wrong. I followed his instructions, which led me to blogspot.

All of the blog URLs that I wanted were taken. It turns out nearly everyone in the world decided to start a blog in 2001. They grabbed all the good blog names on blogspot, staked out their claims, posted a single article and left, never to return.

It took me half an hour to set up my blog, virtually all of which I spent searching for a free name. I finally settled on "subclock" because I have a clock from a Russian submarine mounted on my home office wall. I've used it for about ten years. It requires some discipline -- it needs winding once a week, and will drift from the true time if I don't adjust it slightly when the humidity changes.

Besides that, there's too much overclocking in the world today; it makes sense, sometimes, to slow down, take a look around, and think deliberately about the next thing to do. I don't always do that, of course, but maybe the name will remind me.

Here's to paying attention.