i-sako.com

I recently started working on a long-term project to improve the architecture of the Web site at the company where I work. Looking at this project, I cannot help but wish that the principles of good information architecture were better understood at my company. Unfortunately, as with many companies in Japan, there are often very few guidelines in place to keep projects moving in a consistent direction. All too often, how projects are run is a decision left up to the people who staff them. Given that staff changes are almost inevitably an annual event, this can be quite a gamble as projects grow. Without a set of guidelines to follow, the architecture begins to crumble under the strain of multiple ways of doing essentially the same thing.

It reminds me of a book review I wrote for the Tokyo PC Users Group not long ago, which I have reprinted below. I wish I could find a copy of this book in Japanese. It would be just the thing to hand to my coworkers to explain how we need to structure this project as we move forward.

Update: August 5

I stand corrected. It seems that the company finalized its 200-page “Global Web Style Guide” just last month.  I am pleased to report that nearly a quarter of the guide itself is made up of architectural information.  This makes me feel a lot better about the long-term project I mentioned earlier.

Information Architecture: Blueprints for the Web

By Christina Wodtke

Published by New Riders Publishing

ISBN: 0-7357-1250-6

Price: USD$29.99

With Information Architecture: Blueprints for the Web, Christina Wodtke has done what few information architects before her have been able to accomplish: Write an IA book that is simultaneously accessible to newcomers and yet still informative even for more experienced
readers.

Over the course of 12 easy-to-read and often lighthearted chapters, Wodtke lays out the case for good information architecture, how to go about it, and where you can go from here (both in the sense of where the field of information architecture is going and in the sense of what other books you might like to read if you want to learn more about the field itself).

Wodtke starts out with a simple, straightforward question: Why blueprint a Web site? (Note how easily this question could be rephrased to make its answer more interesting to freelancers: Why would a client pay you to blueprint a Web site?) The answer, of course, is all about maintaining usability in spite of ever-increasing complexity. It is entirely possible to manage a small site without a pre-planned architecture, but as the site grows it will inevitably become exponentially more difficult to maintain and downright impossible to use and if there is no underlying architecture to guide its growth. Just as you need a blueprint to build a skyscraper, you also need one to build a thriving Web site.

One of the truly nice things about Wodtke’s book is that it avoids the “guru syndrome” that afflicts many books of this sort. (Guru syndrome is what happens when Web-savvy people become too enamored with one way of building sites—so much so that they feel it is the only correct way.) In fact, Wodtke steadfastly refuses to recommend any particular methodology or advocate the use of any particular software. Instead, she advises her readers to develop their own “toolbox” of skills to apply to any situation. Some of what she writes contradicts the “rules” espoused by some experts, but she offers no rules of her own in their place, which is precisely her point: Information Architecture is not about figuring out the One True Formula that can be applied without fail to all usability projects. Each site will have its own needs and will need its own rules. Rather than precise rules, use rules of thumb. Rely on principles, certainly, but avoid the urge to boil them down into axioms. To make sure that the right principles are being used for the right purposes, Wodtke recommends applying proven techniques (or “tools” in the toolbox metaphor), which make up much of the content of the book.

One of the earliest steps in designing the site is the creation and testing of a prototype, but even before that, Wodtke offers practical advice on how to conduct effective, probing interviews that will help the prototype stage yield meaningful information about what your users need and how to help them get it. Armed with this information, Wodtke then offers a few choice tools with which to make sense of this information.

Stuck on how to categorize your content? A card sort could help you discover the most logical organization scheme. Does your content fit into multiple categories? Faceted classification can help your users find what they want in more than one place. Not sure how to reveal the breadth of your content without overwhelming your users? You might want to rethink your labels. Have content that different people are likely to describe in different ways? Careful, consistent use of metadata and controlled vocabularies can help keep everyone on the same page. Finding out that your user demographics are too vague? Design some user personas and walk them through a number of actual-use scenarios. With each of these techniques, the usability of the site improves as the categories, navigation, and workflow are refined.  The overall results will inform your choices when it comes time to build the interface and make decisions about issues like global and local navigation and pagination.

In the last quarter of the book, Wodtke focuses on putting those techniques into practice. Whereas the middle section of the book was all about what tools to use and when, this section is about how to deliver the goods to the client—and how to do so with style and professionalism. As you might expect, this is where the heavy-duty “blueprinting” comes in.

Contrary to what you might expect, Wodtke offers compelling reasons to avoid the use of any software for the first steps, the sitepath diagramming or topic mapping, because it may actually inhibit the creative process. Software comes back into play later, when putting the finishing touches on your storyboards, but even those are best written in pencil in the beginning, Wodtke believes. Projects naturally become more software-intensive as you move toward expanding your storyboards into full-fledged wall diagrams with functional specifications. As your project evolves even further still, you will need to rely more on software to manage your content inventory, to arrange your sitemap, and to draft your page schematics (often called “wireframes").

In addition her advice on how to handle the deliverables, Wodtke also offers some very useful suggestions on how to manage the less tangible aspects of large-scale projects, things like how to “jump start your brain” and the creative process, how to present your solutions to the client, how to peruade others, and even how to handle arguments. This section was a thoughtful bonus. In a book already brimming with good advice, it was like finding a digest version of How to Win Friends & Influence People tucked in among the information architecture techniques and how-tos. In closing, Wodtke also offers some insights into the field of information architecture and a “Recommended Reading” section that points readers to some of the best known publications (both books and Web sites) in the business.

Information Architecture: Blueprints for the Web is a breeze to read, is nicely illustrated, and offers loads of sensible advice. It can be read in a few hours and obtained for small amount of money, both well spent. Any serious site-development or project-management bookshelf should have this book on it.

Thanks to Murakami Haruki and Michael Moore, I’ve got a couple of hot dates coming up.

I’m just about to finish Murakami Haruki’s The Elephant Vanishes, which I’ve been reading at my wife’s suggestion. She’s taking me to a new play based on this book, you see, and she says I need to have some background knowledge to fully appreciate the performance.

She introduced me to translations of Murakami’s work when we lived back in the States, and I’ve always enjoyed them a lot, but it has only been more recently that I’ve started reading his work in its original Japanese (again, with the aim of understanding the Japanese in the play). It’s been good exercise for my brain, because I don’t read Japanese for pleasure as often as I should.

Although I think Murakami’s best work is found in his stand-alone stories, this collection of short stories was perfect for reading on the train during my daily commute. I can also imagine that these shorter stories would be much easier to adapt to the stage than, say, Hard-Boiled Wonderland and the End of the World. Still, the surrealistic fringes of Murakami’s stories are sure to be a great challenge for both the actors and the director. I look forward to seeing how they handle them.

In the same vein, my wife has been reading Michael Moore’s Stupid White Men in preparation to see ”Bowling for Columbine”.

Hmmn...The Elephant Vanishes and “Bowling for Columbine”. Odd choices, perhaps, but that’s what our respective cultures are producing these days.

I’ve always admired Tim O’Reilly for his sensible positions on technology issues, but this really put a smile on my face:

Tech Books to Enter Public Domain

Computer book publisher O’Reilly & Associates is taking a dramatic stand against automatic extensions of U.S. copyrights by voluntarily limiting its own copyright protection on hundreds of technical titles—and promising they’ll enter the public domain after that.

This, I feel, is a bold step in the right direction. In an era when corporations are pushing for ever more control, O’Reilly & Associates is putting its assets into the public domain within the timeframe deemed appropriate by our Founders (who I daresay understood the implications of never-ending copyrights far better than our current government). Good for them!

Companies like O’Reilly deserve our unwaivering support. They’ve certainly got mine.

Over the past few days, I read The Cathedral & the Bazaar from my Safari Bookshelf. It was only the second book I have ever completely finished reading online. (The first was Stephen King’s Riding the Bullet, which of course had to be read online. I guess that makes The Cathedral & the Bazaar the first book I’ve ever read online when a printed copy was also available.)

[Aside: I’ve purchased and read other e-books, of course, but I don’t usually finish them. I find that I still don’t like reading books online. This is no fault of the e-books themselves, but simply the technology. Computer monitors, because they project light, are not as suitable for reading long texts as printed pages, which reflect light. Online news articles and other relatively short content are no problem, but book-length material can be hard on your eyes.]

I really enjoyed the author’s approach to the subject, the benefits of open source software. Eric S. Raymond, a well known Open Source evangelist (though some apparently prefer to think of him as an Open Source Jedi), leads the reader through a very analytical investigation of what Open Source is all about, how it evolved, what social norms it observes, and why it produces results that might seem counterintuitive to business people. The book contains a lot of detailed ethnograpical analysis, much of which builds a very strong moral case for Open Source.

The author, given his background, is far from impartial, but the book itself is remarkably objective. There are situations where it makes sense to keep the source a secret, Raymond concedes, but those are not typically for the reasons most business leaders cite. 

More than anything else, I enjoyed reading this book because it provided concrete inspiration for a course of action that I have been slowly drifting toward for a long time now. It’s time for me to get serious about “becoming a hacker,” as Raymond would put it. 

Fortunately, one of the appendices provides suggestions for doing just that. Ironically, I found that I have been going about it the exact opposite of the way Raymond recommends, which is to learn how to program first, how to use Unix (or Linux) next, and then learn HTML and other Web-related technologies. In my case, I learned HTML and Web-related technologies first, have been learning Linux over the past year, and am now feeling ready to start learning some programming languages. How was I to know that this approach was backwards?

One of the things Raymond recommends against, however, is learning proprietary languages, for obvious reasons. That is something I will certainly keep in mind as I delve into the Absolute Beginner’s Guide to Programming, which seems to include a lot of information about Microsoft’s Visual Basic among the more worthy programming languages like C/C++ and Java. (Hey, I can always skip over the VB-specific chapters, right?)

Raymond says that a properly motivated, disciplined would-be hacker can expect to get a firm grasp of the basics in about 18 months (which, interestingly enough, is about the same amount of time it takes to become proficient in a martial art, which is another thing that Raymond and I share an affinity for). Applying that to my situation, if I start right away and remain diligent in my studies, I should expect to be a reasonably competent geek by the time I turn 30. That is, admittedly, a late start for someone seeking to become a hacker, but better late than never. I’ve been toddling toward geekdom ever since I bought my first computer; it’s time to make a solid commitment.

For me, then, the best thing to come out of Raymond’s book was a clear understanding of the roots of Open Source, because from those roots I can see where I can fit in as Open Source branches out. I don’t ever expect to make contributions on the same level as first-rate hackers like Raymond, but I identify solidly with his tribe and will do what I can to help it along.

For as long as I have been in Japan, I have always found it difficult to find A) English-language books and B) sufficient free time in which to read them.

Shipping books from overseas has always been time-consuming and costly. Buying books locally is costly, too, and often means putting up with limited selections, especially in the area of technical books—which are the kind I crave most.

Because I work a typical salaryman’s schedule and have an active family life, my only “free” time to read traditional, hard copy books is often when I’m commuting to or from work—which is usually not quality reading time. (For books that are available online, however, I have much more reading time—I am reading something online almost every minute of every business day.) Although I do consider myself fortunate to have a direct, 30-minute train ride each way, the trains are often overcrowded in the morning (making it difficult to even hold a book, much less read one) and my brain is often worn out in the evening (making it difficult to focus on the subject matter). Also, the trains are almost always noisy, which is a significant distraction. All things considered, I am usually lucky if I can get 20 minutes’ worth of uninterrupted reading done in the hour I spend each day on trains.

Fortunately, neither of these things need to be a problem any more. I’ve just signed up for my own Safari Bookshelf, a service of the O’Reilly Network that allows me to search through the full text of technical books from a wide variety of publishers, and to add the ones that I like to an online “bookshelf” that allows me to fill a specified number of “slots.” (In general, one slot is equivalent to one title, but there are a few exceptions; in my case, I opted for a ten-slot bookshelf, but there are options for larger or smaller ones). The great thing about it is that I can replace some or all of the titles on my shelf with new ones each month. At the rate I buy books, the Safari Bookshelf represents a tremendous savings.

At the moment, my newly created bookshelf contains the following titles, several of which had been waiting on my Amazon.co.jp wishlist for quite some time:

\t
• Adobe InDesign 2.0 Classroom in a Book ($45.00)

\t
• Apache The Definitive Guide, 3rd Edition ($39.95)

\t
• Cascading Style Sheets: The Definitive Guide ($34.95)

\t
• Database Design for Mere Mortals ($39.95)

\t
• Managing Enterprise Content: A Unified Content Strategy ($39.99)

\t
• MySQL ($49.99)

\t
• Running Linux, 4th Edition ($44.95)

\t
• Sams Teach Yourself Apache 2 in 24 Hours ($24.99)

\t
• The Cathedral & the Bazaar, Revised Edition ($44.95)

\t
• Web Database Applications with PHP & MySQL ($44.95)

Now, if I were to buy all of these books at list price, it would have set me back nearly $410, but the same collection of material (assuming that I pay for an annual subscription) costs me only $160. Also, in addition to being able to draw from these books whenever I need to (so long as they remain in one of my slots), I can also search other Safari Bookshelf titles or replace titles on my bookshelf with different ones as my needs or interests change.

In my personal budget, I generally allow myself the equivalent of $50 a month to spend on books. On average, this works out to one or two new books a month. With my Safari Bookshelf subscription, however, I can get as many as ten new books each month while spending just slightly more than a quarter of my annual personal allowance for books. This is a very compelling advantage.

The Safari Bookshelf is also filled with helpful features that help me remember where I last stopped reading, allow me to bookmark specific sections of the books on my shelf, and allow me to make notes about the books I am reading.

I will be sure to write more about my experiences with Safari as I explore more of it (I’ve only just recently subscribed under the 14-day free trial, so there are many things I have yet to try out), but for now I think it’s a safe bet that I will look to O’Reilly’s Safari Bookshelf first for the technical books on my wishlist.