What if you looked at a document and only saw squiggles? What if you couldn't tell an invoice from a love letter from an old shopping list? Life would be pretty tough. But that's exactly the problem with web pages because of Hypertext Markup Language (HTML)-the language of the web.
Not that HTML isn't useful. HTML puts up a good front. It does a wonderful job of telling browsers how to display information. Put this text here. Make that text bold. Put a picture over there. This is a link. But what HTML doesn't do is recognize what the document or the document contents are. And that's about as limiting for a browser language as it would be for you.
The fix seems simple. Just figure out a way for the language of web pages to "know" what kinds of documents and contents are being displayed. It's a great idea and it's idea behind the development of a new web language that you may have heard about.
That language is called Extensible Markup Language or XML. Extensible means "extendable." In this case, the idea is to extend good old HTML by adding things that make it smarter. That should produce some big benefits for business and for every web user. Here's what XML will do.
XML will make the web work better.
Let's say that you want to find books that contain the poetry of Robert Frost. If you do a standard search on the Web today, you're likely to come up with those books, but also books about Frost and books that are literary criticisms of his poetry. You might even get some stuff on David Frost or the effect of frost on citrus crops. If the Web were just a little bit smarter, you could sort those out quickly and up front. XML gives us the opportunity to do that.
That, by itself, should speed up some parts of the Web since it will take less searches to find what we want, but it will speed the Web up in other ways as well. Right now, because HTML just displays information and doesn't understand it, a lot of information gathering on the Web is essentially a back-and-forth process.
Let's say that you've decided to go for a trip, and you've searched for airline tickets. You have a number of flights displayed on your screen, and you want to click the option that lets you narrow down those flights by time-of-day. When you do that in today's world, you hit a button and your browser sends a message back to the server where it got the first set of data. That server, then, performs the calculation for you and sends the results back to display on your screen. All that back and forth wouldn't be necessary with XML.
Here's how it would work with XML. The servers would send you the first set of results, along with a simple program (such as Java). If you want to narrow things down, everything is done on your computer, using the tiny programs sent with the results. You only to go back to the original server if you change things entirely. It's a lot faster and more effective
XML will help businesses share information.
Today's business software programs would probably get a bad mark in the Citizenship section of a third grade report card. That's because they don't work and play well with others.
Even inside the same organization, there are lots of different programs that essentially don't share information well. Accounting systems don't share information well with marketing systems, which in turn don't share information well with calendar and scheduling systems or production systems. XML gives us a way to pass information from one program to another without having to follow the "Three Rs of Electronic Commerce" - Rip, Read, and Re-key.
Take the example of a bank in the Midwest. They got a huge improvement in loan officer productivity by simply laying a browser over the top of a collection of legacy systems. Suddenly, instead of having to go through a series of six or seven menus to bring up results from separate databases, all the loan officer could see all the results in one place with only six mouse clicks.
XML could take that farther, by letting all the different systems share information with each other. It could allow the loan officer to see important information flagged. It might allow the information to flow into a credit scoring system without a lot of background programming.
XML does this by tagging the elements of information with some kind of agreed-upon meaning. That agreement is essential. Inside a company, you can decide what those tags are like so that you can move information easily between operations. Once you go outside the organization, things get a bit more complicated.
Let's take our example one step farther. The bank might be partnered with an auto dealership to make auto loans. A potential buyer might be seeking loan approval through the dealer's website. Or a salesperson, might be using the bank's extranet to look for financing, while at the same time querying other dealers for availability of the model the customer wants in a particular color.
This scenario would require agreement on a number of information tags that might come through a system of leads on automobile buyers. That's under development. It's called Auto-lead Data Format or, inevitably, ADF. It uses XML.
There are lots of sets of standards under development right now. The Newspaper Association of America is developing standards for classified ads. Folks in Human Resources are developing one for their types of applications, and the insurance and banking industries are creating some for themselves.
It's not just business applications. There's science, where researchers are developing markup languages to make it easier to share their results. There's already a math and a chemical and an astronomical instrument markup language in the works.
Groups of enthusiasts are developing standards to help share information with other folks who share their passion. There's a chess markup language (ChessML) for chess players. There's MusicML for sheet music. There's even one under development for astrology.
Web browsers have become a common interface for lots of applications. XML will make our browsers more useful by making them more powerful and able to share information inside and outside our organizations. There's one particular benefit of XML that will affect lots of businesses.
XML will make it possible to develop a document in one form and publish it in many forms.
Here's how things work now for something like a company newsletter. You probably prepare the document initially in a word processing program and then move it over to a publishing program of some kind. You may go from there to the Web version or you may go from your word processing to an HTML template. At each conversion point, things are either done manually or with an individual bit of conversion software.
It gets even more complicated if you're talking about different kinds of Web displays. Amazon.com, for example, expects to sell books over the Web as we know it, but also over the Wireless Web on tiny little screens on cell phones. That takes a totally different kind of display. XML can help in both of these cases.
XML will let you create a document with content tags that will work with different applications to determine what information is displayed and how.
We're not there yet.
This is big stuff. In fact, it may about the only thing that I can think of, related to the Web, where the hype might actually be justified by the ultimate performance. But we've got both business process and human beings involved in this, so there are bound to be a couple of bumps in the road.
The biggest bump is that we need to have agreed-upon standards for what different tags mean in order to exchange information. Within industries like insurance, banking, and automotive, this can work pretty well. There are bound to be some retailing standards and retailing sub-standards for different kinds of stores. This is a business issue, not a technological one.
That leads us to stumbling block number two. Because of its power and complexity, XML is more difficult to do well than HTML. With HTML, just about anyone with a web design tool could create decent Web page layouts. XML will be more difficult to use, because the standards will be more complicated. Right now, there a whole bunch of "techies" ramping up for this challenge, but it won't happen overnight.
So what are the predictions? Some of them are decidedly unrealistic. The folks at Gartner predicted in late 2000 that by the end of 2001, 70% of electronic transactions between businesses will use XML. That's simply not likely. Even if we had all of the agreements in place by the end of 2001 (and we won't) we would still have to re-work our systems and develop the XML documents. Work like that takes a while.
Here's what I think is going to happen. We're in for somewhere between two to three years of standards development and politicking. At the end of that period of time, you can expect stuff to start to shake out; and the heavy lifting of development to begin. It will take another couple of years beyond that for this stuff to really spread throughout the Web. That's about five years' time.
That's long from the perspective of people like Gartner, but it's the blink-of-an-eye in terms of business development time. Here's what you should be doing right now.
Find out what's happening in your industry and the industries you work with. What standards are under development?
Next, step back and look at your business. When I work on efficiency projects with clients, we chart the information flows inside and outside the business. That will help you identify that places where information sharing can have the biggest impact on your effectiveness.
Put these two pieces of information together to develop a plan for how your business will use XML to improve your profit. Then get started, with some simple project to get you familiar with the possibilities and tools.
This is really big stuff. What we're looking at is the potential to transform the experiences we have on the Net and the Web, making them easier, faster, more effective, and more profitable. Stay tuned . . .
Here are some resources to help you get a handle on XML and what it means to business.
Start with John Bosak and Tim Bray's excellent article in Scientific American by pointing your browser to http://www.sciam.com/1999/0599issue/0599bosak.html This will give you a good overview of what XML is and how it developed.
Then review the business implications in an excellent overview article from Business Week at
http://www.businessweek.com/1999/99_24/b3633183.htm
There are a couple of articles in CIO Magazine that will help you, too. One is called "The X Factor" and you'll find it at http://www2.cio.com/archive/031500_et_content.html and the other is "Tag, You're It" at http://www.cio.com/archive/031500_tag.html The articles are cross-linked, so getting to one will get you to the other easily.
By the time you've read all the above, you may figure that XML will cure world hunger and usher in a reign of peace, all by three PM. Time for a dose of realism about the difficulties of implementation. So read Clay Shirky's down-to-earth "XML: No Magic Problem Solver" at http://www.business2.com/articles/mag/0,1640,14004,FF.html
If your web developer or IT folks are looking for something they can use, suggest they start at http://www.xml.com/ which comes to you from the folks at O'Reilly who publish all kinds of great technical stuff.
This feature appeared on 23 July 2001
