I found an interesting article in the new issue of the journal Ariadne:
Paul Miller, Towards a typology for portals. An excerpt:
Look around you. They're everywhere. Everyone who is anyone has a portal, even if in reality that means no more than paying some graphic designer an insane amount of money to knock up a new graphic for your old Web site that says 'Portal'. We can only assume that, for the corporate sector at least, someone at PwC, Accenture, KPMG, McKinsey or one of the others decided that portals were cool and started sticking them into their PowerPoint as another thing you could pay them to build for you, in order to turn your company around, and make it world-beating. Before we knew it, the portal plague was out of the lab, replicating and mutating like anything.
James Robertson pointed to William Denton's nice annotated bibliography about faceted classification and the web.
Almost every article or book listed here begins with an explanation of what a faceted classification system is, so I won't (but see Steckel in Background below if you don't already know). They all agree that faceted systems are very appropriate for the web. Even pre-web articles (such as Duncan's in Background, below) assert that hypertext and facets will go together well. Combined, it is possible to take a set of documents and classify them or apply subject headings to describe what they are about, then build a navigational structure so that any user, no matter how he or she approaches the material, no matter what his or her goals, can move and search in a way that makes sense to them, but still get to the same useful results as someone else following a different path to the same goal. There is no one way that everyone will always use when looking for information. The more flexible the organization of the information, the more accommodating it is. Facets are more flexible for hypertext browsing than any enumerative or hierarchical system.
The Online Development Center has just launched an online standards-based elementary report card tool in the Enumclaw School District. Here's a brief description of what the tool does and how we did it.
Many districts are moving to standards-based report cards. Instead of giving students a single grade for "Reading," "Math," and other grades, the standards-based system lists specific items a student is expected to know or do. For example, one of the items under "Writing" is "Writes coherent paragraphs." The result is that both students and parents have a clear idea of how each student is doing in school.
The district had been using a largely paper-based system. They decided to move to an electronic system to help reduce the workload on teachers and other staff, and to ensure consistency across the district. The existing systems they evaluated did not have the flexibility the district needed, so we created a very customizable system for them.
When teachers log into the system (which is integrated into Enumclaw's intranet, and tied to the district's Active Directory), they see their classlist. From here, they can enter grades for individual students or choose to enter grades for some subjects all at once. The grade entry screen was set up to encourage quick data entry. Teachers are also encouraged to enter comments for some subjects. At the district's request, the comment feature allows teachers to write as much as they want.
Specialists, like Physical Education and Music teachers, enter grades through a specific interface. They can enter grades for their subject one class at a time, speeding up their process considerably.
When it comes time to send report cards home, the system produces printable report cards as PDF files. The printed report cards are both attractive and flexible, able to accommodate as much or as little information as provided on each student. As the year progresses, the report card will show a student's grades from the entire year on a single easy-to-read printout. This allows parents and teachers to quickly assess student performance over time. Another notable feature is that report cards can be printed in either English or Spanish.
District administrators can define report card templates for each grade level, providing a good deal of flexibility for the district over the long run. As standards change and evolve, the system can be changed without making programming changes. District administrators can also modify and add students to the system. Most of the student data is imported from the district's student information system, but changes often need to made during the grading period.
On the technical side, the system uses Macromedia ColdFusion as the middleware, and Microsoft SQL Server 2000 for data storage. Report card data is generated as XML, then is transformed (using XSLT) into XSL-FO, a XML-based formatting language. The XSL-FO is then transformed into PDF using the Apache Foundation's FOP render. Additional PDF processing is provided by PDFLib. PDFs are cached to the disk to speed operations.
Usability.gov's Research-Based Web Design & Usability Guidelines looks worth a look...
D. Keith Robertson has posted the first in a series of lessons learned from his recent re-design of the Seattle Children's Hospital website. This first post focuses on why they used web standards:
Since it went live a few weeks ago we’ve already began to reap the benefits of Web standard development, specifically XHTML 1.0 and CSS, down at the hospital. I’ve also had quite a few questions about some of the things we did, where we are seeing benefits and how it was all done.
Update: The second installment, on Usability, is now up:
If you are involved in designing, developing, writing for or managing Web sites or applications you owe it to yourself to get involved in some kind of usability process. It’s a real eye opening experience. No matter how much you feel you’ve mastered you craft, or how much you know you “get” the Web, watching a few real people use your sites will teach you loads and could really open your eyes to some things you’d never before noticed.
Very true.
Like many other fields, Information Management is packed full of vocabulary. The terminology gets even more thick when we wander into the more traditional Library/Information Science areas like Controlled Vocabularies. But don't work, because the folks at Boxes and Arrows have come to the rescue with a thesaurus of terms about controlled vocabularies. Yes, they're actually demonstrating many of the concepts they're defining. Cute, huh?
Catching up on a few good web-design related links:
Quirksmode.org is Peter Paul Koch's collection of tips and tricks for dealing with browser differences and quirks.
Dave Shea has a nifty new technique for creating CSS-only translucent dropdown menus.
The most recent edition of the Web Standards Project's "Ask the W3C" series addresses the question of HTML versus XHTML. The answer is...drumroll, please...XHTML. Why? XHTML is easier to maintain, can be transformed using XML tools like XSLT, is easier to teach and learn (HTML: "you need to close your tags...well...except for those you don't need to") , and it is "ready for the future."
A mini-case study on Harvard's digital archives:
I had never given much thought to long-term storage and archiving of content. Well, I did to the extent that I typically advocated XML and other standards-based technology--mainly so content and data would outlast any particular software system or application. But I had never thought much about truly long-term storage—what some in the field like to refer to as the "100 year digital object"—until I was engaged by the State of Washington to help study the feasibility of creating a digital archive. Part of my analysis was to interview other organizations that had undertaken similar initiatives. The most interesting interview was with Harvard University.
If you've been building websites for any length of time, you've probably been faced with an existing site that needs some maintenance. Sometimes changes are minor, and other times you're looking at a total re-design. Which ever situation you face, you're going to end up messing with the site's existing code. Even if you wrote it, it can be hard to find your way around in this code. Lucky for us, Zeldman has the answer. Guess what, the answer is to use standards-based code.
(Note: the link I pointed to above is the printer-friendly version. The Flash version was a big too much for me...)
Lots of folks out there want a great website for a little money. Creating a great site is tough, no matter the budget. But Carrie Bickner has some tips and advice in a new book called Web Design on a Shoestring. You can check out Chapter 2, on project planning, for free. Her advice?
1. Dare to do less. 2. Write a short project goals document. 3. Create a functional requirements document. 4. Craft a technical requirements document. 5. Keep documentation nearby.
Here's a fun article/presentation:
Why tables for layout is stupid: problems defined, solutions offered.
There isn't really much new here to those who have been following the standards revolution of the past few years. But the nifty pictures are worth a look.
I know how Tim Bray feels when he writes about his difficulties installing some fairly standard open source software on his Mac. While Tim's battles have been with perl and mysql, I've been fighting with python. More accurately, I've had trouble installing a variety of third-party python modules on Mac OS X.
So I gave up.
I decided to use my Windows box as a python development environment. I actually already had a couple of flavors of python installed (2.1 and 2.2...yes, I know I'm at least .1 out of date). 4Suite installed without a hitch (I never could get it compile on my Mac). So, I finally had an XSLT processor to use with python.
Next, I went off in search of an editor. On the Mac, I've been very happy with BBEdit. I decided to check out some python-specific editors. Boa Constructor, and open-source product, looked cool. The install went fine, but it was very buggy during use. Every time I tried to save it would throw an error. And it was slow. Not fun. Next, I tried a trial version of ActiveState's Komodo. It turned out to be a very nice little IDE for a number of languages. I'm considering buying a copy when my trial runs out. Before I do that, I'd like to try out the Wing IDE.
So, things have installed nicely, and I have a decent programming environment. But, all was not rosy. I blew a fair amount of time trying to figure out why what looked like perfectly good code kept throwing syntax errors. Turned out it was a problem with some code I'd copied and pasted from my Mac. The Windows version of python didn't like the end-of-line characters. Once I figured it out and fixed, the code worked fine.
Eric Meyer has an interesting post on some new Microsoft technology (XAML) and what it could mean for the future of the web. Oh, and Meyer doesn't see a happy future:
Permit me to repeat myself: "When one company owns the medium, everyone else loses." Everyone from design firms to tool vendors to browser makers will have to dance to Microsoft's tune. We have until about 2007, maybe 2008, to prevent that from happening. Can it be done? How? By whom? If XAML lives up to its potential, Microsoft won't need the W3C any more. Why should they play by the open community's rules when they can create their own very lucrative and highly controlled gated community?
Update: More commentary from Dave Shea.
Another Update: Jon Udell chimes in.
The HITS (Humans, Interaction, Technology, Strategy) looks pretty interesting. IASlash has links to the event slides and some commentaries.
Hold on to your hats, kids. You can now use the Library of Congress Subject Headings (LCSH) in a faceted classification scheme: FAST (Faceted Application of Subject Terminology): A Simplified LCSH-Based Vocabulary. According to the authors, the FAST schema is:
- A controlled vocabulary with all headings established in the authority file, with the exception of headings containing numeric values only;
- Based on the LCSH vocabulary;
- Designed for an online environment;
- A post-coordinated faceted vocabulary;
- Usable by people with minimal training and experience,
- Compatible with automated authority control.
This looks to be a very interesting option for systems that need a general controlled vocabulary.
(via Christina Wodke)
Here's a very interesting look at the wireless network setup at Dartmouth College, courtesy of VentureBlog:
In the late 1980s, Dartmouth College was the most wired campus on the planet, running 10GB Ethernet into every dorm room. Today, Dartmouth is the most unwired campus on the planet, with 560 access points covering 200 acres. At a recent conference here, Larry Levine, the head of computing services, challenged attendees to find a single spot on campus and surrounding areas that did not have 802.11 coverage. Even the boathouse, adjacent sections of the Connecticut river, the ski lodge, and sections of the ski slope are covered!
Another nice persona article to add to the collection: Alison J. Head's Personas: Setting the Stage for Building Usable Information Sites in Online magazine. Especially valuable is a complete sample persona.
(via Step Two)
After a hiatus, A List Apart is back with three new articles and a nice redesign. ALA is the one of the oldest and best independent publishers in the web design field. Glad to see it back.
Here's a long, but interesting, article on programing practices: Coding Smart: People vs. Tools.
I've recently seen a spurt of activity denouncing the concept of personalization. News.com reported on a Jupiter Research report that concluded that personalization was "not only ineffective, but surprisingly costly." A few days later, Gerry McGovern jumped in with a similar rant about personalization and portals. Meanwhile, the folks on the SIGIA-L mailing list jumped in with a thread on the subject. Of course, all of these probably owe something to Jakob Nielsen's 1998 article declaring personalization "over-rated." And he probably wasn't the first to critique the idea.
Regardless, here is one of the better quotes from the McGovern article:
The best definition I have found for a portal is as follows: "A portal costs four times more to buy and operate than a normal website. It delivers half the benefits." Yes, that big, cumbersome, complex portal of your dreams may well make you long for the days when you were running a simple but effective HTML website.
The word "portal" seems to be one of those great buzzwords that pops up among those who know just enough to be really dangerous. I had thought the portal craze had run its course in the late-1990s, but I've been hearing it more and more lately. McGovern goes on to encourage people to focus on the basics: write good content, lay the pages out well, create easy-to-use navigation, and make sure search works. He's right of course. If you do the basics, then you can move on to the bells and whistles.
But, is "personalization" a bell and/or whistle you should add to your site?
First, let's define personalization. Simply, it is any time you serve up a customized page based on an individual user's profile. The classic example has always been Amazon.com, where the products and links you're shown vary depending on the pages you've viewed and products you've purchased. In in intranet situation, the site could show you different content based on your department or job type. Or, search results could be filtered based on your past search history.
There do seem to be some nifty advantages here. But there also seem be some big traps to avoid.
The navigation scheme should not be customized. A site should provide a consistent navigation scheme thoughout the site and across multiple visits to the site. A navigation scheme that jumps around depending on the users like likely to make users think too much.
There has been a fair amount of interest in personalized search. The concept has promise, but plenty of potholes, as well. Most implementations make use of a user's search history to influence results. Of course, each time a user turns to search, he or she likely has a different goal in mind. Even when using similar search terms, the user might have a different goal in mind. I doubt the additional cost and complexity would be worth it. I'd look at adding a taxonomy-driven "best bets" system before trying to personalize search.
Many e-commerce sites have "Other users bought these..." type features. These often lead users to other similar products. But often, especially if there isn't much data to feed into the Bayesian filter system that drives these features, the other items are somewhat random. The technology that drives these features is obviously getting better and better. But, right now, it takes some fairly seriously processing and data mining to put a system like this in play. Its probably out of reach for most sites. A better approach would be to make good use of metadata to display a similar items list.
So, are there good personalization features? Sometimes. Is it worth it? Nope. Looks like I'm piling on the bandwagon.
Lou Rosenfeld points to SchemaLogic's interesting product in a recent posting. Basically, the SchemaServer is a fancy combination of a metadata crosswork and change management tool, perfect for large organizations (or groups of related organizations, like education organizations) that don't use the exact same schemas. I've sat through two or three SchemaLogic presentations, and I think it looks quite interesting.
Lou gets bonus points for mentioning Bob and Mike in his post.
I recently ran across an interesting site, Modern Home Design. The site offers stock plans for a variety of interesting looking houses. But perhaps more interesting, for our purposes, is the site's navigational scheme.
The site's creator uses geography as the main navigation. A visitor to the site needs to pick the geographical terrain they plan to build on. The options, as of this writing, are:
Coastal (Northern ) Southern Plains Coastal (Southern) Coastal (West) Coastal (East) Mountains Foothills Forest Desert Northern Plains Tropical Island River/Lake Shore
I found this navigation scheme very interesting. I've rarely seen geography used as the primary means of organizing information. For a purpose like this, it can be a useful way to organize information on the site. But, used alone, this method has some big drawbacks. For a user like me, who isn't interested in buying plans for a specific project, the navigation seemed clumsy. Here are a few of the issues:
Fortunately, I think a faceted classification scheme would be perfectly suited to a site like this. To create a faceted scheme we'd need to identify the core attributes of the objects we're organizing, then provide a way for users to browse through these attributes (for a great example of this in action, see Epicurious' Recipe browse feature).
If I was doing this for real, I'd spend some time doing research on the site's users. I'd be looking for information about how the users view the information we're organizing and what aspects of the information they find most important. Since I'm not doing this for real, we'll just wing it.
So, what aspects of house plans are important to buyers? Certainly, we've already seen that geographic terrain can be a useful attribute, so let's leave it in. Many users likely have a size requirement in mind, so let's make square footage our next facet. I'd imagine that construction costs are important to most potential plan buyers, but the site doesn't have much information on that front (and for good reason, in that materials choices and local conditions can cause prices to fluctuate quite a bit). We're going to have to leave this out. Many users would likely care about the number of bedrooms and bathrooms, so this should be included. Finally, many would be interested in the style or genre of the house. Is it a Mies-style box or something more along the lines of Frank Gehry? The challenge with this attribute is coming up with meaningful and clear values.
So, we have a list of facets:
Its a decent list; not too long, not too short. And, it shouldn't be too hard to come up with values for each of the house plans on the site.
A fairly simple database should be able to power this site. I'm going to skip over the details of actually implementing a faceted classification scheme, and point to a nice tutorial by Berkeley's Flamenco project. Parts of the tutorial are overkill for our purposes (like the hierarchical nature of their facets), but it is still a good read.
Once a database is in place, the site just needs to provide users with a nice way to browse all of these facets. And, while they're at it, they should make sure to note the number of plans that fall under each facet ("Tropical Island (2)"). And, better yet, don't even display values that don't have any items in them. It should make it much easier for users to find their way around the site.
Christina Wodke has posted a some interesting feedback to a Boxes and Arrow's article: The Devil's in the Wireframes. The article discusses how wireframes are often overloaded with too much detail, defeating their point. Wodke, a leading Information Architect, questions the very notion of wireframes:
Is the wireframe an atrocity whose time has come to be booted out of the development cycle? I'm beginning to think yes.
Also, check out the comments on Wodke's post.
My take? I've used wireframes, and found them useful. I suppose much depends on the team you're working with.
Lou Rosenfeld has posted a couple of presentations on Enterprise Information Architecture.
Joel Spolsky has written a must-read article (and not just 'cause he says basically says so in the title):
The Absolute Minimum Every Software Developer Absolutely, Positively Must Know About Unicode and Character Sets (No Excuses!).
So I have an announcement to make: if you are a programmer working in 2003 and you don't know the basics of characters, character sets, encodings, and Unicode, and I catch you, I'm going to punish you by making you peel onions for 6 months in a submarine. I swear I will.
I realized I forgot to mention that I've posted a nice list of links and resources dealing with RSS and Weblogs. I compiled this list in early August for the LLILI conference. Lots of good starting places if you're new to the world of RSS and weblogs...
I doubt this is an original thought, but after listening to a fair amount of discussion on metadata and the Semantic Web at DC-2003 this week, I'm struck by this thought:
The Semantic Web will be built first behind firewalls.
Random thoughts to back this up:
The metadata and RDF aspects of the Semantic Web are hard enough that only professional developers will be able to tackle them without mature toolsets.
The Semantic Web is about creating environments that enable "computers and people to work in cooperation". Big companies have massive information management problems where Semantic Web solutions could help.
One concern with building systems that rely so heavily on exposed metadata is metadata spamming. Spamming isn't an issue in the intranet space. And, trust is less of an issue within a corporate environment.
Kendall Grant Clark has a nice analysis of the state of end-user XML editing tools.
SimpleBits has a very cool article on creating CSS-based accessible image tab rollovers. In a nutshell, the article describes a technique for creating a nice rollover effect without Javascript.
I'm linking to this mostly because of the name: Bare Bones, No Crap, CSS Text Control Primer. The content is good, too. (via WaSP)