So I started volunteering time working on the QUIP project at GUIR, the UC Berkeley Group for User Interface Research.

I've been wanting to work with the group for a while, and I'm looking forward to finding out more about how the group works and whether I'll be able to do research on my own computational modelling projects. At any rate, it'll be a good experience until the semester starts and I can determine what my classload is like, and what I'd ideally like to be doing. That and it's giving me some hands-on experience with Java Servlets.

I've had my Tivo for about two months now, and I've noticed a few interface pieces that could stand to be copied by other industries.

One of the cool things Tivo does is keep track of the shows you like. During a show you can hit a 'thumbs up' or 'thumbs down' button and Tivo will remember that you do or don't like that show (and will try to use that knowledge to guess what other shows you might like or dislike, but that's tangental here). When a Tivo has extra space, it'll fill it with the kind of shows it thinks you like and, other than some erroneously perceived desire for '80s Steve Gutenberg movies, it does a pretty good job.

Checking CNN this morning, I wished thier site had a similar feature. Wouldn't it be nice if you could quickly and easily customize the news, not by going to a profile page and checking 'politics' and unchecking 'baseball' (or what have you), but instead by indicating likes and dislikes in-situ?

For example, wouldn't it be cool if one click on the tiny thumbs down icon next to the lead story about the GOP convention would ensure that I wouldn't have to read about the convention, protests, and counter-protests for the next week? One click and 'boom!' no more Jon-Benet Ramsey stories, Bye-bye O.J.! Elian, I hardly knew ye.

And of course, one click on the thumbs-up would result in more features about Linux, the Space Shuttle, the Concorde, and two headed cock-fights. For example.

The web gives every information soruce the chance to become a person's ideal information source, but as much flash as CNN has and as much realvideo as ABC can stream to your door, it's still a choice of which channel to watch, instead of the ability to really make your own.

Thaks to Josh for pointing me to this one.

For those of you who think getting groceries from Peapod, NetGrocer, or Webvan is too cumbersome, and Costco just doesn't carry the bulk you need, Celebrity Prime Foods might be just what you're looking for. Ignoring for a moment the abyssmal look of the site, they offer semi-annual shopping. They send a rep to your home and you work together to figure out, in aggregate, what your family will want to eat for the next six months, then they deliver an industrial freezer with all the vacuum-packed food you need! Every six months, the big truck comes around again and restocks the freezer.

Who would have thought that I Love Lucy could be so prophetic?

Ammy and I are putting together some new Cameo cameras today. Check out the site, put your name on the list, and drop me a line if you're interested in getting one.

If you don't get one this batch (about 20 cameras), I'll be sending out a batch of about a hundred in September, so be sure to put your name on the list!

Hey, I know it's been a little quiet for the past few days, but I've been coding into the wee hours finishing the back-end of a new content publishing system using php4, mySQL, and some interesting data structures I've worked out.

The upshot is that it's going online today. You won't notice much a change at first. The HTML al looks (pretty much) the same, but the first new feature is a paged log entry system, so the home page only shows the last week's posts, and you can jump to archived months. This will drop the page size from 220K (including graphics) down to around 26K.

Now that the hard part is done, I'll be able to add a lot of site-wide functionality ('what site? It's only one page!' "Soon...") quickly and consistantly.

The best part of the new system is that the UI could be completely changed on a site-wide level with only a few changes in code. While I won't be flipping the UI around all the time on this site, it should prove to be an excellent foundation for usability experiments, especially some work I'm doing on adaptive personal user interfaces.

One of the realizations that's significantly influenced how I approach a design project is that there are, for the most part, two distinct kinds of sites or presentations out there: tools and experiences.

The two kinds of site are usually differentiated at the outset by defining the measure of success. A store, a news site, a search engine. These are epitomes of tool sites. The interface exists to fascilitate a collaboratively directed transfer of information, either from or to the visitor.

An experiential site typically has a very different measure of success. Unfortunately, that measure is often hidden because it is superceded by a more quantifiable measure. For example, a site about brand-building (www.levi.com for example), may have 600,000 hits per month as a measure of success. This is unfortunate because the real measure of success should be the impact the site has on the individual.

Some people try to incorporate subjective measurements by looking at other objective measurements, such as pageviews per visit, time spent on a page or a session, 'stickiness', and the like. This ends up doing the site and viewer a disservice because when the designer (architect, graphic designer, whoever) designs with the goal of increasing these numbers, they often do so to the exclusion of the subjective goals.

Of course, sometimes this is blatant and intentional. If you've ever been to a porn site (stumbled across it while looking for something else, of course), then you're familiar with the dozens of unrequested popups that scream, 'Don't go! There's more porn over here!' This is one example of inflating a number while deflating the subjective user impression.

Getting back to the point of this entry, the numbers are important goals for tool sites. Numbers indicating daily sales, purchase-path abandonment rates, coupon redemption rates, clickthroughs, repeat business and affiliate programs are vital to tool sites. The insights they yield can help create smoother purchases, more directed shopping, better search results, and the like. In short, they make better tools.

The opposite is true for the experiential site. 'Measures of success' are really measures of effectiveness, but when only objective, numeric qualia go in, the only effectiveness that's measured is that people see, that people read, that people think, but not what they see, what they read, and most importantly, what they think.

As indicated earlier, while the distinction between tool and experience is important when assessing the success of a site, it's much more important when designing the site in the first place.

In the time I spent working at the big firms, each had its own 'best practices' philosophy that sought, among other things, to define the process by which a site is made. Roles and responsibilities were at the center of these processes. In some companies, design begins with the graphic design (the 'creative' side, as it's called). In others, strategy and technical hammer out the site requirements. Typically the initiating side takes on their part of the pie as well as the interface design and, whether they realize it or not, the groundwork of the information architecture.

The problem is that the initiating side isn't often decided by the type of site, but by the culture of the design firm and, in some cases, the culture of the client, regardless of the type of site being created.

So, taking the long and winding road to get back to the point of what an "Information Architect" is, the information architect should be the bridge between the creative and technical arms. He or she creates a framework that a technical team constructs and a creative team furnishes. Still, the architect usually has a greater affinity to the technical side, which is suitable when creating a tool site, but not when creating an experiential site.

However, chopping the IA out of the equation on experiential projects would be severing the tie to innovative technical solutions that can augment the experience. What's needed is a different sort of information architect, an "Information Artisan" if you will.

The information artisan's responsibility would be to work with the design team, ensuring that the final architecture provides them with the framework they need to achieve the aesthetic goals of the site. They would be analogous to the information architect on a tools site, only instead of specifying a structural framework to support a site, they would design a piece where the artistic aspects extend deeper than the creative arm usually reaches, calling upon the tech side to enhance the art instead of using design to enhance a tool. Kvetch and The Fray are good examples of this.

That's all for now. Maybe I'll touch this up when I'm not in such a rambling state of mind.

"Lightning doesn't strike twice."

Where does this saying come from? It seem silly because anyone who's been in a thundershower can tell you that it can and does strike twice or more in the same place all the time. The whole idea of lightning is that it strikes a certain type of place, be it by altitude, metal conductivity, whatever, and those things usually don't change, so things that get hit once are likely to get hit again.

Okay, that was a non-sequitor. Anyhow, on tap for today, I'm writing up my views on last night's BayCHI Web Design BoaF meeting, where we took a look at eyeTracker. Also coding away as usual, though I came up with some really cool data-structures last night on how to make the site very, very modular, and cool from the back end. This'll filter down into several site enhancements on the front end as well.

Now that the Borg are now making spherical ships, is that a sign that they overtook the Imperial Empire, and added their uniqueness to their own?

It's a dead givaway if the Borg queen starts breathing heavily.

Well, my GREs are done, and I did well enough that I'm not planning on taking them again. So now I can get back to working on the site, the other projects, and some actual paying gigs again.

One-track-mindedness can suck, but it feels good when you finish the task at hand, even if it takes you a month.

Wish me luck! I'm taking my GRE this morning! (I really should get some sleep...)