February 2005 Archives

As I showed in a recent blog post, I carry a paper notebook around. I've been doing it for years. From Genevieve, I learned to paste stuff into it--like business cards, notes on napkins, and clippings from the Economist. That gives me an artifact and context (the notes around it), rather than having the artifact in one place and the notes in another.

I also, of course, note the prevalence of index cards. They're all over agile programming techniques, geek task organization and the current personal productivity trend.

And then there's the reinterpretation of digital artifacts in meatspace (blinkenlights, Space Invader's digital tile graffiti, and the "Connect-Four" version of Tetris).

Anyway, this has led me to conclude (of course) that

Analog is the new digital.

What does that mean? Got me, but it's a slogan. ;-)

As part of the Design Engaged conference last November there was a day when we were broken up into groups and encouraged to walk around Amsterdam and brainstorm ideas about design, technology, cities and human interaction. It was a pretty broad mandate, but we shared a similar set of interests, so many of the ideas resonates and we quickly assembled a bunch of interesting ideas about new technologies. During that exercise, I started thinking about heat maps as ways of representing things other than heat.

Heat maps are maps like the heat weather maps that appear in newspapers. They represent shifting gradients of temperature, as mapped to geography. In the newspaper they consolidate a lot of potential information (think of the listing of temperatures and cities that scrolls on CNN International) by keeping one variable (geography) stable and mapping another one to it. I thought "I wonder what else can be mapped like that?" and, of course, WiFi was the immediate choice, since I'm constantly checking for WiFi strength when sitting in cafes. Mapping WiFi to heat maps has been done by several groups, with the following map taken from Chris Lentz's 2003 Dartmouth Senior Thesis as one example:

My next thought was "OK, assuming that someone is going to do some semi-automated way of doing this based on the Netstumbler database [link[, what to do with this information?"
Then, I thought of all of the other information that's already bombarding me as I walk down the street looking at potential places to sit with my laptop: I often have headphones in my ears, or I'm talking on the phone, I'm looking out for traffic and reading menus, I'm trying to find a place with a table near the window. Finding the strength of WiFi connectivity is a secondary consideration at that point, and is always going to be, but the information is there, so how to get to it?
OK, I thought, so our visual sense is totally overloaded, and our auditory sense is pretty close to it, but we have a bunch of other senses that are perfectly good and relatively underused. We can map information to these other senses. One of these is heat, so why not reverse the polarity, so to speak, and take the WiFi heat map and map it back to actual heat.

That's the long windup for the pitch of the following idea:

The WiFi heat sensor

This type of shifting of data to a non-visual/non-auditory sense could have all kinds of other possibilities. Connected to a GPS (or some other way to find your place) with a database, it could tell you where there was an ATM, or a point of historic interest. Connected to a Lovegetty-like (or "Familiar Stranger"-like) device, it could tell you when there was an acquaintance nearby. It would give you a sense of the data space around you in a way that only you would know about.

My next step is to make a working one. No. Really. I have the Peltier Junctions and a hackable WiFi detector. Spidey sense here we come!

I've been playing a lot of World of Warcraft the last couple of weeks. Ben has been talking about it, and I have heard so much about it, that I felt it needed to be tried. It's been years since the last time I seriously played video games, so I was behind on all this MMOG stuff and wanted to see what was up.

To write a simple review: Warcraft is fantastic. The way that the experience is constructed is very clever and there are great little touches everywhere. The problem the designers face are similar to many other kinds of experience designs (whether for theme parks, web sites or functional consumer products): they have to keep people's attention, make it fun regardless how long someone's been playing, guide people deeper without alienating or boring them, and constrain the behavior of many hundreds of players while giving the perception of complete freedom.

It's a difficult task and they pull it off very well. I may be writing more about it as I force myself to tear away from the game, but suffice it to say I've been dreaming in Warcraft for about a week, and although that's a little disturbing, it's a testament to the power of the experience (or maybe I just have too much time on my hands these days, but I don't think that's it).

The observation for today is on feedback. One of the ways that Warcraft differs from everyday life, and one of the primary ways that it stays so addictive, is through continual quantitative feedback. There are a lot of different progress indicators, and one or another is always going up. There are many, many ways to get "better" and, unlike the real world, these are quantified and observable. Rather than the abstract notion that I'm getting smarter by reading books, or getting a little more fit by walking, I know immediately when I get better and by how much. This is powerful positive feedback and although it's certainly possible to go too far with obsessing about quantifiable achievement, it's one of the things that drives the interest, and continual use, of the game.

Why would someone--namely, me--spend hours killing and skinning virtual animals just to get his shooting, skinning and cooking skills up?

This lead me to thinking about smart devices and social effects. One role that smart devices, or personal technology in general, can play in people's lives is to quantify the normally unquantifiable achievements. This quantification can help us conceptualize numerically--and we love to boil things down to numbers when obsessing, don't we? (see: engine displacement in hotrods, MHz in overclocked CPUs, heart rate, frequent flier miles, dollars in the bank, etc.)--things that we can't imagine otherwise. I think that there's an immense potential in devices that help us understand how much we do x or y. Then the whole world becomes a little like Warcraft and we can take our goals--"Level 14" becomes "a bit closer to my 43Things goal"--and maybe make them happen in the small steps that Warcrafts shows us and real life doesn't.

[The screenshots are, in order: places Irving--my dwarf hunter--has explored in the land of Westfall (the UI exposes more map as Irving travels), Irving's reputation with various races and cities, and Irving's skills (you can see he's not much of a cook, but he's even worse a miner ;-).]

Another smart carpet project as blogged by We Make Money Not Art (who is astoundingly, shockingly, depressingly prolific in her blogging talents--I can only hope she's a grad students procrastinating from doing her dissertations; if she's not, then she can't possibly be getting anything done and if she is, then I may as well go back to streetsweeping).

Liz posted an interview she did about the class she taught last fall at SFAI. I was flattered to have been invited to participate the final crit of one of the first art classes to talk about the digial radio spectrum as a medium, anywhere, and it was very interesting. She generalizes her experience to talking about what it takes to teach art in such a heavily technological medium. Go Liz!

I'm co-organizing a workshop at the Pervasive 2005 conference in May with Lucia Terrenghi, Irma Lindt and Andreas Butz. We just extended the deadline and I figured I'd announce it again.

Here's the announcement:

Paper Submission Deadline: 1st March, 2005

WORKSHOP ON USER EXPERIENCE DESIGN FOR PERVASIVE COMPUTING

http://www.fluidum.org/events/experience05/

Associated with the Pervasive 2005 Conference (http://www.pervasive2005.org)

12 May 2005, Munich, Germany

Organized by Lucia Terrenghi, Irma Lindt, Andreas Butz, Mike Kuniavsky

---

Experience design is a design approach which focuses on the quality of the user experience during the whole period of engagement with a product: from the first approach, through its usage, to the reflection and memory of the complete relationship.

As technologies for wireless networks, image capture, storage and display get cheaper and more performing, and as the internet drives up the availability of a pervasive information and communication infrastructure, it becomes possible to embed computing capabilities into a variety of environments and bring communication in a much broader set of contexts. Thus, pervasive computing and context sensitive systems allow for the design of new stimuli from which people could create their own meaningful experiences, individual or shareable. These goals raise new challenges, suggesting the need of new methods and forms of interaction patterns between users and environments, and between different groups of users. Design can play a key role in shaping new toolkits for contextualized experiences, and enhance the natural evolutions of users' sense of place and time towards the experience of living in a mixed reality, in which physical and virtual spaces are blending together, and social relationships become fluid and distributed.

---

Submissions:

You are kindly invited to post a position paper no longer than 4 pages describing your work and interests. Submissions must be in Adobe PDF format and should conform to the Springer-Verlag LNCS style. Themes that are relevant for this workshop include, but are not limited to:

• interaction design for pervasive computing: thus addressing new scenarios for pervasive computing, interaction within instrumented environments, multi-user interaction and interfaces, multimodal interfaces, tangible user interfaces, interaction paradigms.
• design for experience management: techniques and activity theory approaches focusing on how to stimulate and support users in creating their experience within a pervasive scenario, how to engage them within the experience, how to collect, store, reflect on, share experiences.
• user experience evaluation: user experience taxonomies, understanding of users' needs and specification of requirements, evaluation methods and assessment approaches.
• design of toolkits for the authoring of blended experiences: design for ambiguity, allowing for a transparency of the infrastructure that can stimulate and support users' and designers' creativity while making a semantic relationship between physical and virtual spaces of a mixed reality.

---

Contacts:

For further information visit the page
http://www.fluidum.org/events/experience05/

for contact and submission
lucia.terrenghi@ifi.lmu.de

A couple of months ago I had the idea that it would be possible to use the OSX Finder to automatically (or "automagically," as the old Sysadmin term went) upload photos to Flickr when my camera was attached to my Mac, with no involvement from me. The logic went like this:

• Flickr can get uploads from a UNIX command line
  
• AppleScript can issue UNIX commands
  
• the Mac Finder can automatically execute AppleScripts when a new object appears in a folder (a Folder Action)
  
• Folders are created in the /Volumes directory whenever a new volume--such as my camera--is mounted
  

This chain of reasoning led me to calling on Josh Ellithorpe, who I asked to stitch together the Apple sample backup Folder Action to the Folder Action that fspiers, a Flickr user had put together.

Here's the standalone script, which does the same thing as the folder action, but without having to be invoked as a folder action, and here's the folder action.

It's still a rough proof of concept, but it works.

Here are Josh's instructions for installation:

property flickrmedia : "Flickr Media"
property subdir : "DCIM"
property eject_when_done : true
property emailaddy : "quest@mac.com"
property thepass : "YOURPASS"

flickrmedia - the name a folder in your Application Support directory. This is where you put an alias to the memorysticks or drives you want to auto upload.

subdir - on my camera it puts everything in DCIM... so
/Volumes/MEMORYSTICK/DCIM/Photodir1/photo.jpg. I figured yours would
have a similar layout, if not the exact same folder name.

[in my case it's /Volumes/SANDISK/DCIM/100PENTX/photo.jpg --mk]

The other options are painfully obvious.

Ok, so for my setup this is how everything works. Create the folder "Flickr Media" in "/Users/YOURUSER/Library/Application Support/". Inside that folder put an alias to any drive you wish to upload on mount. This is exactly how the Backup script works but not in your homedir which I thought was bad design.

The standalone is simple. If the alias is in your flickrmedia folder then you mount it (hookup the camera) and double click the file. Then it searches your memorystick/drive for /Volumes/MEMORYSTICK/SUBDIR where subdir is the folder you define in the applescript. So in my case it was /Volumes/Untitled/DCIM/. Then it lists the files in all folders in that subdir, my cam has many folders. Then it uploads them, and displays a dialog saying how many files it uploaded. Then the disk gets ejected if you set eject_when_done to true.

The folder action is installed exactly like the backup script.

The script is definitely a first generation pre-alpha thing. It doesn't tell you how much has been updated, it doesn't keep track of what's been updated and what hasn't, the Finder pretty much freezes while it's running (which can be quite a while) and, in general, it's not Flickr Uploadr, but if the goal is to experiment with making transparent technology--and that's what this is--it works.

So go and use this in health.

Liz and I had a conversation yesterday based on my post about Neil Gershenfeld's book; specifically the part about the Web distracting people from ubicomp for most of the late 90s and early 00s.

Her point is that the Web had two big advantages over ubicomp in 1997 (whether or not it actually took resources away from ubicomp--it's not a zero-sum game):

• The barrier to entry was low. Hardware is hard. HTML is easy.
• The Web was about bringing people together, ubicomp was about making cool stuff.

The second point is really important, I think. Something that I've always talked about in UI design, and seem to have forgotten in my recent excitement with physical computing and the embedding of computation in everyday objects, is that people's ultimate goals in using any tool is to communicate with other people. There's little that's done solely by yourself, for yourself. That's a pretty boring closed system...and it's boring by definition since, by definition, we find purely solitary behavior kinda boring, on the whole.

Thus, forgetting that smart objects need to be in the service of social effects is a surefire way of making stuff that will be popular to a narrow niche, at best. An important lesson in design. Thanks, Liz!

Homebound with a cold in the unusually sunny Portland winter, I read Neil Gershenfled's book, "When things start to think." It's an interesting book, and still relevant to the ubicomp world, even through it was published six years ago and he was writing it eight years ago. It's an autobiography, an introduction to technological and scientific concepts, wide-eyed speculation, a polemic against everything Gershenfeld doesn't like, and a big ad for the MIT Media Lab and the students there. All in 200 pages. Sometimes, it's all of those things in a single chapter, as when he starts by talking about telemarketing, moves into a detailed history of the Protestant Reformation, slides over to the Bill of Rights and end up creating his own version of Isimov's Three Laws of Robotics. It's not as much about making deep connections as it is about associative idea surfing. It's sitting with Gershenfeld while he gives you a braindump of everything that's on his ADD mind. It's quite blog-like, actually, and if blogs had existed in 1997 that may have been a more appropriate forum than a book.

The most interesting thing about the book, which is on the whole about ubiquitous computing--even though Gershenfled doesn't like the term--is how little progress has been made since he was writing 8 years ago. Over and over I see people re-discovering the same ideas he was talking about in the book, which were old to him even then, as if they were something amazing and new. His iconic image is the shoe computer. When did the first real shoe computer come out? 2004 (the Adidas 1).

Why did this happen? I blame The Web. The Web sucked up the "best and brightest" for 10 years, and only now are people getting bored enough with it to start thinking about hardware. The pieces are there:

• Chinese manufacturing (which everyone has heard about and owns at least one product that benefitted from) is there to make it.
  
• The DIY movement has spawned a bunch of interest into what goes into the black boxes.
  
• And the ephemerality and increasing sameness of the Web is pushing smart people toward physical objects (confirming what Chicken John said years ago: "All these dotcom people, you know what they really want to be doing? They want to be working with wood.").
  

This book is an interesting reminder of where all of this came from, and how it's neither new nor revolutionary. However, reading it shows me how important momentum and timing are: Gershenfeld was first in many ways, but without the support of companies and individuals running with the ideas, they stagnated, and only now are we picking up the crumbs of 1997.

The end of the book introduces Gershenfeld's Things That Think Consortium. Gershenfeld is no longer a director of that program, though I get the feeling he was at the time.