In 1991, two Seattle musicians designed and patented the first novelty eyeglasses with lenses fitted inside the circles of the two numeral “9”s. It was a brilliant idea and one that lasted because every year since then has had two adjacent 9s or two adjacent 0s in it. That is, until 2010, which still almost works. But this upcoming year, 2011, begins a drought of double circles that will last until 2016.

But it turns out, the two inventors, Richard Sciafani and Peter Cicero, exited the market in 2008 (Seattle Times Dec 2008) because low-priced knock-offs (from a country that must not be named) have made the business unprofitable. Their company Brainstorm Novelties still exists though.


Richard Sciafani (left) and Peter Cicero. Photo from Seattle Times

The New Year eyewear fashion fad received attention from the Wall St. J. Dec 2010, complete with examples of how some designers have solved the 2011 single circle problem.

Oh well, hope you have your party favors, hats, firecrackers, and plastic champagne glasses ready. Happy New Year!

Until the mid-1990s, kidney transplants from live donors who were not genetically related to the recipient were rare. As mentioned in a May 2010 blog post, most hospitals had a policy of rejecting donation offers from spouses or other unrelated people. The thought was that the risk of any live donor surgery was too great unless the benefit of a better HLA match outweighed it. This meant the only allowed sources of transplant kidneys were close relatives and deceased donors.

Policies at hospitals changed as 1) the advantage to recipients of receiving a kidney from a relative fell with improvements in immunosuppressant medications, 2) the risk to donors of surgery declined with the development of laparoscopic techniques, and 3) public attitudes about donor informed consent began to favor the belief that unrelated donors were making their offer for altruistic reasons rather than selfish ones (like being compensated or ego gratification or coercion).

The figure below shows the number of kidney transplants performed each year in the U.S. from 1998 to 2009. As mentioned in an Apr 2010 blog post, the rise in transplants mirrors the rise in the incidence of end-stage renal disease in the U.S. The proportion of transplants from live donors rose from under 20% in 1988 to a peak of almost 43% in 2003. It then dipped for a few years and is rising once again.


All transplants. Data from UNOS

Looking at the figure below, the composition of the live donors has been changing dramatically. In 1988, only 71 kidneys were transplanted from unrelated donors (the New York Times article reports a lower number) representing 4% of living donors. The proportion of unrelated donors has been steadily rising and now accounts for about 43% of all living donations. Notice that after a quick rise, the number of donations by spouses or life partners has leveled off. The most common sources of unrelated donors now are friends, neighbors, coworkers, church members, and even strangers that meet on the Internet.


Live donor transplants. Data from UNOS

Even more interesting is the extremely rapid rise of alternative sources of living donors. The last three lines from the figure above are rescaled in the figure below. The fastest growing source of living donors is paired exchanges. As described in a Mar 2010 blog post, a kidney exchange involves the trade of kidneys between two sets of incompatible pairs. Each pair consists of a recipient and a willing, medically suitable donor who is either blood type or HLA incompatible. Through an exchange, they can find another pair in the same situation but where the donors match the recipients in the other pair. Kidney exchanges have the potential of becoming the leading source of live donor kidneys within a few years. (More about that and the potential of adding compatible pairs to exchanges in a future blog post.)


Exchanges and anonymous transplants. Data from UNOS

A living/deceased exchange is similar to a paired exchange, except that the recipient in the pair receives a kidney from a deceased donor rather than from a live one. Under UNOS rules, people who have donated organs receive 4 points if they ever need to enter the transplant waiting list. This is often enough to move them to the top of the list. In a living/deceased exchange, the donor provides her organ to a patient on the UNOS list and provides her 4 points to her incompatible recipient. The recipient is now at or near the top of the list to receive an organ from a deceased donor.

Since receiving a kidney from a live donor generally produces superior medical outcomes to one from a deceased donor, a living/deceased exchange does not produce the best possible outcome for the recipient participating in the exchange. Yet despite the growing use of paired exchanges, the number of patients participating in living/deceased exchanges is also growing. Hopefully, paired exchanges will grow fast enough to soon make living/deceased exchanges unnecessary. (Incidentally, the process for managing living/deceased exchanges is covered under a patent application. My low opinion of business process software claims can be seen in this Mar 2010 blog post.)

The final fast-growing source of donors are people who donate without a specific recipient in mind. They are called nondirected or altruistic donors. Johns Hopkins Medicine claims to have performed the first nondirected live donor transplant in September, 1999 (though UNOS data shows five other nondirected donors in 1999, three in 1998 and one in 1988, the first year data is available).

In addition to increasing the total number of donations, nondirected donors also play an important role in starting donor chains in kidney exchanges. Donor chains reduce the risk to recipients of their matched donors backing out an exchange after the first transplant takes place. Thus, nondirected donors reduce the need to perform the transplant surgeries simultaneously. This simplifies scheduling personnel and operating rooms for kidney exchange transplants.

To learn more about becoming a nondirected donor in a kidney exchange, contact a transplant center and ask if it participates. Lists of some participating centers are available at the National Kidney Registry, Alliance for Paired Donation, Paired Donation Network, and New England Program for Kidney Exchange.

[Update: This data is examined at the transplant center level in a Jul 2010 blog post.]

I recently came across a few articles that featured natural user interface or NUI design that I found interesting. (Note: I’m currently doing contract work for the Microsoft Surface team.) The first article in Tech. Rev. Mar 2010 discussed a way of improving the haptic feedback users get when manipulating buttons on small touch screens, like the ones used in most smartphones. The touchscreen designed by a company called Immersion, uses a piezoelectric material that vibrates 100 micrometers every millisecond as the user’s finger moves over a button and then stops once the finger is no longer on the button. With this small but fast movement, the user gets the illusion that the flat glass surface has a convex button on it.

Improved haptic feedback is an extremely important advance in the design of automobile dashboards, where glancing down at the controls can lead to driver distraction and vehicle accidents. It can also add haptic feedback for controls used in factories, power plants, aircraft, hospitals, etc. where diverting your attention from the work surface to look at a control can be dangerous. Immersion is already licensing its technology to LG for smartphones.


LG Chocolate BL40 smartphone uses haptic technology

This is very different from the traditional motorized vibration technology that most firms were pursuing a few years ago to give user’s feedback when they press an on-screen button correctly. This feedback is coarse, so while it informs the user that the button was pressed, it doesn’t give the user the illusion the surface actually has a button. For more about this older style of feedback see Tech. Rev. May 2007.


State of the art haptic feedback back in 2007


Microsoft Research is working on a very interesting touch screen painting application called Project Gustav. It combines texture mapping, color and transparency mixing, multi-touch input, and a 3D paint brush bristle deformation algorithm to make a realistic painting simulation. The effect is very impressive and could have lots of commercial applications. It will be interesting to see if this technology makes its way out of the lab and into some retail products soon.


An expensive finger painting simulator. Image from Microsoft

Also from Microsoft Research in cooperation with Carnegie Mellon University is a touchscreen interface called skinput. It projects an image onto your forearm or hand and then records the user’s finger taps using a microphone. Somehow it doesn’t seem really practical. Plus it seems a bit creepy. Check out the YouTube video.


Tap tap tap. Photo from Carnegie Mellon

Finally, gamers everywhere are excited about Project Natal, the controller-less interface for the Xbox 360 that is slated to be released for the upcoming holiday season. Although most people have just been amazed that Project Natal works, I wanted to learn more about how it works. That is how does the vision system convert 2D video of a person in motion into a 3D wire-frame model? Microsoft has been fairly quiet, either because they are still patenting the IP behind it or because they want to focus attention on the games not the technology. Alas, the only article I could find was Pop. Sci. Jan 2010.


3D wireframe generated from a 2D camera. Photo from Pop. Sci.

[Update: Microsoft has released Project Natal under the tradename Kinect.]

Apple recently filed two patent infringement cases against HTC, a Taiwanese company that makes the Nexus One smartphone for Google as well as phones that run on Microsoft’s mobile OS (now called Windows Phone 7 Series). Given that Apple doesn’t license its technology, the remedy Apple seeks is to stop HTC from selling multi-touch smartphones. It could be a start of a patent war between Apple and Google. (A brief history of the relationship between these two formerly friendly firms is detailed in Gizmodo on Mar 2.) Many claim that a legal cloud could impede innovation in multi-touch interfaces. I say, not likely.

Farhad Manjoo builds case against software patents in Slate Mar 2010. I’m against software patents because I think protecting ideas is different from protecting inventions, though I have a hard time defining the line between the two. So while I think Mr. Manjoo is being overly dramatic in his article, I believe he’s basically right. Companies clubbing each other with patent lawsuits doesn’t do much to reward inventors for being creative, encourage companies to invest more in research and development, nor provide consumers with a greater variety of innovative, well-designed products.

In a  Mar 4 blog post I mentioned how easy it is to get software patents. In three years at Microsoft, I was involved in the development of code that led to three patent filings. I’m proud of the work I did, and the patent applications were for some clever ideas. But the patents were for ideas and algorithms, not inventions. And I wasn’t any more creative during my three years at Microsoft than during the prior twenty years when I worked for a variety of other companies. And during those twenty years I had never filed a single patent application. What does that tell you about the ease of getting software patents vs other industrial patents?

On the other hand, I don’t think software patents hurt the economy as much as Mr. Manjoo claims. One of the references he cites is a working paper coauthored by James Bessen of Research on Innovation. However, if you read this paper, it actually says that the largest proportion of software patents are not collected by software companies. Rather, as many at 95% of these patents are awarded to large industrial companies that use them for strategic purposes like stalling a competitor or as a defense against a competitor. These companies already have created huge patent thickets and would continue to pursue their aggressive legal strategies through other means if software patents were not easily available. This argument is fleshed out more fully in a book Mr. Bessen coauthored entitled Patent Failure which has a chapter on Abstract Patents and Software.

Thus, one could argue that companies like Intellectual Ventures (see Feb 2010 blog post) could actually help level the playing field by allowing all companies to license patents to the benefit of the investors in the fund. However, that isn’t happening at the moment because the largest investors in Intellectual Ventures are also the licensees. Eventually though, patents may be treated as tradable securities and a market will emerge for these rights. (And if we are lucky, the market will remain rational and will not inflate like a bubble.)

Edward Tufte is one of my favorite writers. He’s a specialist in creating information graphics and a proponent of dense data displays. He is the author of several books, most of which I own. In his latest book, Beautiful Evidence, he introduced sparklines, little graphs that can fit in the same line as text. Sparklines are a wonderful way to combine words and graphics to tell a story with statistics.

Here’s an example of how they can help improve your understanding of numeric data. Assume your company has three products called A, B, and C. Here’s a spreadsheet created in Excel 2010 that shows sales for the products over a seven month period along with total sales.


Sales of three products over seven months

Which product is doing best? Is it Product C with the highest total sales? Or is it Product A, with the highest sales in July? I suppose I could create a graph of the three products, but that would take up a lot of space. Wouldn’t it be nice to create a dashboard that displayed and compared the sales of the three products?


A sparkline of the same data reveals trends

Well, why don’t I create a sparkline? Voilà! Now we can see that sales of Product C are falling rapidly, Product B are flat, and Product A are quite erratic. Pretty nifty, n’est pas? Now I can investigate the reason for these numbers.

Since publication of Tufte’s book, several companies and open source groups have built add-ins for Excel to create sparklines and embed them within a cell of a spreadsheet. In fact, sparklines are so useful that Microsoft decided to implement them in the upcoming version of Excel, as explained in a Jul 2009 MSDN blog post.

I’ve had a chance to play around with this new feature and find it to be quite easy to use. First you select a sparkline type from the Insert ribbon. You select the data range and the destination cell and click OK. This causes the sparkline to be inserted and a Design ribbon to appear on the far right (see below). The ribbon displays quite a few options to make your sparkline pretty. You can group multiple sparklines so that they all have the same min, max, and formatting.


Sparkline ribbon in Excel 2010

I found two annoyances though. First, you only get three choices for chart types, line, column, and win/loss. I wish there were more, especially a scatterplot, since most interesting data is not evenly spaced. Most of the 3rd party sparkline add-ins provide scatterplots, so hopefully Microsoft will add them in the next version of Excel. A good comparison between Excel 2010 and the open source Sparklines for Excel add-in is provided in a recent blog post.

The second annoyance deals with copy-paste. Because sparklines are so compact, it would be nice to be able to embed them in a Word document. However, as you can see in the example below, copying a range of cells containing a sparkline and then pasting it in Word causes the sparklines to disappear. Note that this limitation is due to the way the clipboard works in Office, so will affect all sparklines, not just Microsoft’s implementation.


Copying a group of cells containing sparklines in Excel


Pasting the same cells in Word. Oops what happened to my sparklines?

Having some software experience, I understand why this happened. But most users won’t and will just be frustrated. The sparklines are not static pictures, they are charts generated from the data in the Excel spreadsheet. Word can’t draw the graphs, so when I paste them, it just ignores them. If I want to retain the sparklines, I need to paste the selection as a graphic. But I don’t really want to paste the entire selection as a picture. I want to paste the cells and the text as an editable Word table and have the three sparklines pasted as 3 separate images inside their respective cells of my table.

Hopefully, the Office team will improve paste fidelity of sparklines in the next version of Office, though my guess it would be somewhat difficult to implement and they may decide it’s not worth the effort.


I wanted to see what Edward Tufte thought of the new sparkline feature in Excel and so visited his blog. It turns out he’s a bit apoplectic and so are his fans. But not about Microsoft adding sparklines to Excel. He’s actually happy about that. Rather, it is because Microsoft filed for a patent covering its implementation. This probably isn’t the place to discuss the validity of software patents. However, I will say it is really easy to get a software patent. All it takes is a few hours of thinking about the best way to solve a computer problem and anyone can file a patent. For instance, check out U.S. patent 7437659 that was implemented in Publisher 2003.

[Update1: I clarified my example with screenshots and clarified that the paste problem is not a fault of the sparkline feature.]

[Update2: I just posted a blog entry that talks a bit more about software patents (see Mar 6).]

[Update3: President Obama just named Edward Tufte to the Recovery Independent Advisory Panel. Let’s hope that he can improve transparency and communications in reporting financial matters.]

At the recent TED conference in Long Beach, Nathan Myrhvold, the founder of Intellectual Ventures (and former CTO of Microsoft), demonstrated a laser system that can track and kill mosquitoes. As silly as this sounds, it has potential value. First, with the falling cost of electronics, a device that could protect the doors and windows of a small building could be built for under $50. Second, when coupled with other strategies, it could reduce the death toll from malaria in Africa and Asia. You can watch a video of the system in action. The device was developed after Bill Gates, head of the Bill and Melinda Gates Foundation, asked Mr. Myrhvold for assistance in developing creative ideas to eradicate malaria. The idea was first publicized last year.

Intellectual Ventures is a rather secretive firm. Its stated goal is to turn patents into marketable assets. It has characteristics of a hedge fund in that it doesn’t disclose its investors or the contents of its patent portfolio. It operates a large number of shell companies to disguise its purchase of patents and its licensing activities. The New York Times recently ran an article on the firm.

[Update1: A blog post by Brad Burnham, a venture capitalist, rejects Intellectual Venture’s approach.]


More pivot charts

Also presenting at TED conference was Gary Flake, the director of Microsoft’s Live Labs. He demonstrated a data visualization technology called Pivot that is designed to help people make better use of digital information. Says Mr. Flake, “With Pivot you can swim through the data, taking little twists and turns.” A video is available here. There is also a web site,, where you can learn more and download the app. Unlike the pivot tables and charts that I wrote about two weeks ago, these pivot charts combine bitmap images with quantitative data. The next step will be to make each bitmap an animation.

[Update2: There is a nice review of Pivot in Tech. Rev. Feb 25.]

I own a lot of tools. And I mean a lot. But I am now faced with a project that I can’t complete because I don’t own the right tool.

In creating a new office for Sue, I have demolished a coat closet and now want to remove the sill plate (or sole plate) for the former wall in order to make a doorway to the office. But I don’t want to damage the existing floor on either side of the sill plate and don’t want to remove the entire wall. What I want to do is cut the sill plate flush to the existing stud.

I can’t figure out an easy way to do it. A standard flush cutting saw (of which I own several) won’t fit because on either side of the sole plate the floor has been raised to a finished surface. On one side is a Pergo laminate floor, on the other is ceramic tile, see photo below. I could use a chisel to cut out the sole plate, but that would take a long time, and there is a chance I could hit a nail and ding my chisel point.


How can I remove this sill plate? Photo by George Taniwaki

I’ve seen infomercials for a tool called the Fein Multimaster that can do the job. The Multimaster is a hand-held oscillating tool that looks like a detail sander. However, it has more power and can take a variety of attachments. You can watch the video on YouTube. But it is really expensive. It’s available at Amazon for $389.00. (By comparison, my Ryobi detail sander cost about $35.)

The Multimaster will also be handy for removing the tile from the floor and for cutting out pieces of the of Pergo flooring around the new door. I’ll have to think about it before plunking down cash for yet another tool though.


[Update1: I bought the Fein Multimaster and I use it all the time. It’s great. However, the tool is expensive at almost $400 and the replacement blades are too at $20 each. The patent just ran out in early 2009, so I expect clones to be available soon from competitors like Ryobi, Ridgid, Craftsman, etc. I haven’t seen them yet, but once they do, prices should drop.]

[Update2: I just saw an infomercial for the Rockwell Sonicrafter, available at It’s only $179.00, but the tool is heavier, doesn’t have variable speed control (which makes it hard to handle), and doesn’t have a quick release blade attachment (which makes it hard to switch back and forth between cutting and scraping). Bosch has a their own oscillating tool called the Multi-X, but is only available as a 12V cordless tool. It’s also available at Dremel (which is owned by Bosch) makes a less expensive oscillating tool called the Multi-Max, available at for only $90.]