February 2008 Archives

Here here's your latest computer fridge news: Whirlpool has partnered with a domestic groupware software company called Cozi. Right now, it's just a branding partnership with Cozi's calendar/to-do list/grocery list etc. software for families, but it's clear where this is going: WP is going to create an embedded version of Cozi's software for their centralpark fridge line and then create other ways to connect to the same service. First it's the fridge, then it'll be an iPhone widget, and if it's a hit, a "household activity dashboard" on Mom's desk at the office, like what Ambient devices has done with some data feeds. Or at least that's the hope.

Electronic household organization tools has been around a long time (I took a half-hearted stab at it a couple of years ago). Not counting pre-Cambrian kitchen computer technology, getting into kitchens was an early goal of the first wave of Internet appliances in the late 90s. 3COM's Audrey, one of the classic failures of this first wave, advertises that it "can be the family's nerve-center in no time, handling schedules, phone books, and notes." Cozi's pitch is similar: "Cozi helps busy families manage schedules, appointments, shopping and communications from wherever you are — the kitchen, car, office or even the grocery store."

Timing is critical in technology adoption so there's no reason why these technologies can't work now when they failed 8 years ago. Many people who in WP/Cozi's likely core audience of affluent 30-something new home buyers are probably thinking much more about their families now than they were 8 years ago, because they probably did have them then. However, the repeated failure of the idea is something to learn from and I hope that Cozi has been studying people's habits and the pattern of earlier similar technologies to see why they didn't work out. Is it purely because the value of the service versus the cost isn't great enough (i.e. dry erase boards are cheaper and more flexible, but don't allow you to check your kids' schedule from the road, but that's OK with most people) or is there something deeper? I'll be interested to see where this goes.

My old friend and former business partner, Indi Young, has just had her first book published by my old friend Lou Rosenfeld and his new company, Rosenfeld Media, for which I'm an advisor. That caveat aside, I think this is a pretty great day for user research, user-centered design and publishing around, even as I've watched it take shape for several years from afar.

First, Indi's book. Although the title says it's a method to construct "mental models" through in-depth task analysis, it's a lot more. In the book Indi documents the many techniques she perfected working with a huge variety of clients. The techniques range from how to structure a cross-functional team, to recruiting people, conducting interviews, analyzing them, and creating effective diagrams that communicate the results. Really, it's Indi's whole rigorous process, which so many Adaptive Path projects hinged on, described clearly and in detail. It's a fantastic resource, a toolbox of highly effective, original tools for doing insightful, in-depth user research. I recommend it without reservation to everyone who does user research. We've all been asking Indi to write this book for years, and I'm so happy she's finally done it. Congratulations, Indi!

[FYI, get 10% off the cover price when you buy it directly from Rosenfeld Media's site when you use the code "FOKUNI10" I recommend doing this, rather than going through Amazon because Lou gets a larger proportion of the revenue, even with the discount.]

Second, Rosenfeld Media. Lou has been a friend of mine for many years. We were both part of the soup at the University of Michigan in the 80s and 90s which led to that school's far-thinking innovations in technology. We were there at the beginning of the Web and although we didn't know each other then, I believe we were influenced by many of the same ideas. When I went to LA in 1994 to design websites, Lou was already thinking and writing about information organization. He then went on to basically invent information architecture as we understand it today. He then went on to found a successful company and, in the process, he managed to write something like 8 books.

When he said he wanted to reinvent publishing based on his experiences writing books and designing information systems, I was very excited. Rosenfeld Media is the product of that redesign process. With it, he's decided to embrace the user-centered principles of site design and applied them to making books. But it's not just books, it's the whole culture of information around the books. In Lou's vision, the physical book is an artifact of a larger process of taking experts' knowledge and matching it to the needs of his company's audience of practitioners. Right now it looks mostly like traditional, if somewhat enlightened, publishing company, but you can start to see some differences in it immediately: books have version numbers on every page and you get the digital copy when you buy the paper one. He's also doing much of his marketing research out in the open and he's beta testing his books. It really is a different way of thinking about how to publish technical books, and I wish him luck and success (frankly, I don't think he needs the luck, but it can't hurt ;-).

Geotagging site history

Reading a geolocation paper gave me an idea that I'm not going to implement, but may be an interesting exploration of using location as a memory aid when looking up browser history.

The idea is to use Firefox 3's new Places bookmarking and history service to store the location of where a web page was viewed, and to allow users to sort their history based on location in addition to time and alphabetically (the two current options).

Places rewrites the browser's history and bookmark capability as a single database. Being a database, it allows every entry in the database (every "entity"? I don't know my database terminology well) to have additional fields. Places defines two fields up front: Annotations and Tags (Zotro's influence perhaps?), but there's no reason that either location couldn't be squeezed into one of those fields, or defined as a field itself. You could start with the Wifi SSID. Here's a script to extract that on Mac OS X 10.4, for example:


cd /System/Library/PrivateFrameworks/Apple80211.framework/Versions/Current/Resources
airport -I | grep \ SSID - | awk '{print $2}'


Liz correctly points out that mapping meaningful physical location to either a set of GPS coordinates or an SSID ('linksys', for example) is a significant problem, but I think that SSID-based organization would be an interesting place to start. Moreover, as things like browsers become more portable thanks to dedicated browsing devices and mobile phones, using available information to associate information with location may become increasingly more valuable.

The Site Diet

I'm pretty impressed with Firefox Places in terms of the vision. It's the first rethink of an old and crusty corner of web browsing that hasn't changed significantly in more than 10 years. The Places system keeps track of every visit to a given URL implicitly (as kind of a byproduct of it keeping all activity in a database keyed to URLs) and to me this opens the possibility of many new kinds of visualizations of personal online behavior: where have I been? how long was I there? what did I do there? Google currently keeps track of that for you in its history, but that's pretty creepy. Places moves that kind of information gathering back into the sphere of the end user.

I would also like to use it to remind me of what I'm doing on the Web as I'm doing it. I get distracted by the Web easily (as I know many people do) and end up spending hours on things that aren't high on my list of priorities. I've tried many techniques to combat this and the best browser-based one I've found is the Stealth Kiwi Greasemonkey script. It's a good way to remind myself to stay focused and it works pretty well, but it's a pretty crude solution, kind of stomping on my web use regularly. With the Places API, I imagine having a Site Diet countdown timer next to each URL. I could in advance define how many times I felt I was allowed to visit a site per day or per week (I suspect it would be a default for all sites, with some exceptions, much as Stealth Kiwi works now), and when that countdown timer expired, well, that was it: no more checking my friends photos on Flickr or reading about embarrassing hometown political scandals until tomorrow.

I think it would be pretty straightforward to write Firefox extensions to do either the location-based history or the Site Diet, but I'm too busy making ThingM go, so here, Internet, you can have these ideas. ;-) And thank you Firefox team!

Since we began work with The Henry Ford last year, I've been interested in how museums use technology to tell the stories of their artifacts. Having a single timeline narrated by a single curatorial voice and presented on tiny white wall cards cannot explain complex history and the significance of objects. Museums are, occasionally slowly, realizing this and it's fascinating to watch how they use technology to express their new understanding of their role as cultural repositories.

ThingM's focus on the Henry Ford project was communicating context in history museums, but today I watched how well the Detroit Institute of Arts does the same for art. Liz and I spent the afternoon in the recently redesigned DIA, and it was a surprise and a treat.

The DIA's collection is heavy on the classics and the 18th and 19th European art collected by its original auto company mogul donors. Not all of that art has stood the test of time and some is potentially embarrassing to display without explanation (chinoiserie anyone?). Fortunately, the DIA did not shy away from the questionable acquisitions or keep the "embarrassing" art in the warehouse. This is not to say that they don't have a great collection, they do, but what fascinated me was how they used their secondary pieces to tell stories, to explain and to contextualize the other work. Rather than galleries of dusty numbered Greek vases, for example, they had a life-size rear video projection that explained the social purpose of each of the vases in the Greek wine ritual. This was informative, since the pictures on the vases suddenly made much more sense once their function was understood. Moreover, the rear projection was in the style of the vase art itself, which tied together the artifacts to the people and their rituals in an immediate, entertaining and direct way.

(Flickr image by lisawiz)

Another great exhibit used a bunch of late 18th century French decorative art to tell the story of the life of leisure of French aristocrats, devoting each gallery to a time of the day and the artifacts that would have been found in each situation (implicit in the presentation was that this was the before picture; after the French Revolution, well, that's all different). A great installation in this exhibit was the dining gallery, which had a downward-facing video projection of a formal multi-course dinner, filled with Enlightenment-inspired symmetry, royal pomp and dozens of exotic-to-us dishes (wine aspic, for example). You could sit down at the table and watch as hands in fancy outfits places ridiculous dish after ridiculous dish on the table. Great.

(Flickr image by jtwilcox)

Finally, a much lower-tech technological intervention, but one that was interesting (if occasionally awkward because of its placement) was a set of triangular prism-shaped label spinners. Each side of the prism has a different complementary perspective on a given work from a different, named expert (as opposed to a single view by an anonymous curator). Some have historical context, some cultural. It presents a more nuanced exploration of the ideas.

The parts of the museum where they hadn't implemented all of the changes, where it was more like a traditional museum with vitrines, white cards and only occasional explanation seemed impoverished and bland. But on the whole, the redesign (described in detail [2.1MB PDF] in one of the local papers) is highly successful (and judging from the $180 million the museum raised in preparation for it, its success isn't just measured by the quality of the design) and engaging.

Last night I presented a version of the Sketching Smart Things talk I gave last month at CHIFOO to BayCHI. It was an honor to be invited to speak there because BayCHI is such an institution in the HCI world and because the talk was in the PARC auditorium, feet from where the core concepts of ubiquitous computing were first formulated. Thank you, BayCHI and Rashmi!

The presentation is available on Slideshare:

And as a 710K PDF where you can see a complete transcript of my talk in the notes.

With more than 500 BlinkM's sold in less than two weeks, we're already seeing a number of really fascinating projects. We've made a project gallery on the site to highlight some of these. If you have a BlinkM project that you'd like featured on our site, please send a note to blinkm@thingm.com (or just blog about it, we have a Google Alert that'll probably pick it up in a day or so).

Also, a bit of total self-indulgence. About a year and a half ago, SMITH magazine held a Twitter-based contest for six-word memoirs. I entered on a whim and mine got picked up (along with this image and the memoirs of some 850 others). SMITH got a book deal based out of the contest and now the book is out. It makes entertaining bathroom reading and I congratulate SMITH and Twitter for making a cultural product before pretty much anyone knew who they were.