28.06.2013 Views

Papers in PDF format

Papers in PDF format

Papers in PDF format

SHOW MORE
SHOW LESS

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

left to do? Style sheets don't let one publish mathematics, figures with captions, or dozens of other th<strong>in</strong>gs faciliated by<br />

old languages like LaTeX or newer systems like Microsoft Word.<br />

We could just add hyperl<strong>in</strong>ks to LaTeX. This is more or less what a group of people at Los Alamos National Labs did a<br />

few years ago. I don't th<strong>in</strong>k there are really sound <strong>in</strong>tellectual arguments aga<strong>in</strong>st this approach, but sentiment seems to<br />

be on the side of keep<strong>in</strong>g HTML. If we are <strong>in</strong>deed stuck with HTML, though, perhaps there is a better way to extend it.<br />

Our methodology for extend<strong>in</strong>g HTML seems to be the follow<strong>in</strong>g<br />

1. sit down with a few <strong>format</strong>t<strong>in</strong>g languages <strong>in</strong> common use<br />

2. argue about which are the most commonly needed commands<br />

3. argue about whether Web browser programmers are really up to the task of writ<strong>in</strong>g the code that implements<br />

those commands<br />

At this rate, it will be the year 2000 before HTML is really powerful enough for most people, by which time it may have<br />

been replaced with Adobe's Portable Document Format (<strong>PDF</strong>).<br />

I'd like to suggest an alternative approach:<br />

1. choose a set of 100 documents that represent the spectrum of th<strong>in</strong>gs we'd like to see on the Web<br />

2. come up with a language capable of express<strong>in</strong>g the author's <strong>in</strong>tent <strong>in</strong> 98 of those documents<br />

3. come up with a language capable of express<strong>in</strong>g the designer's <strong>in</strong>tent <strong>in</strong> 98 of those documents<br />

4. add to HTML the semantics of the languages developed <strong>in</strong> the preced<strong>in</strong>g two steps<br />

Fix<strong>in</strong>g the Structure Problem<br />

Can the same approach solve the structure problem? What if we locked a bunch of librarians and a handful of<br />

programmers <strong>in</strong> a room together and made them th<strong>in</strong>k up every possible semantic slot that any Web document could<br />

ever want to fill. They'd come out with a list of thousands of fields, each one appropriate to at least a small class of<br />

documents.<br />

An obvious reason why this wouldn't work is that the committee could never th<strong>in</strong>k of all the useful fields. Five years<br />

from now, people are go<strong>in</strong>g to want to do new, different, and unenvisioned th<strong>in</strong>gs with the Web and Web clients. Thus,<br />

a decentralized revision and extension mechanism is essential for a structure system to be useful.<br />

A deeper reason why this wouldn't work is that nobody would be able to write parsers and user <strong>in</strong>terfaces for it. If a user<br />

is develop<strong>in</strong>g a Web document, does he want to see a flat list of 10,000 fields and go through each one to decide which<br />

is relevant? If you are programm<strong>in</strong>g a parser to do someth<strong>in</strong>g <strong>in</strong>terest<strong>in</strong>g with Web documents, do you want to deal with<br />

arbitrary comb<strong>in</strong>ations of 10,000 fields?<br />

Malone's Work on Semistructured Messages<br />

Back <strong>in</strong> the early 1980s, Tom Malone and his collaborators at MIT developed the In<strong>format</strong>ion Lens, a system for<br />

shar<strong>in</strong>g <strong>in</strong><strong>format</strong>ion with<strong>in</strong> an organization. He demonstrated how classify<strong>in</strong>g messages <strong>in</strong>to a k<strong>in</strong>d-of hierarchy<br />

facilitated the development of user <strong>in</strong>terfaces. Figure 1 shows one of Malone's example hierarchies [**** <strong>in</strong>sert Figure 6<br />

from Malone's paper]. For each message type, there is an associated list of fields, some of which are <strong>in</strong>herited from<br />

superclasses. Consider the class . Fields such as and are<br />

<strong>in</strong>herited from the base class . Fields such as are associated with the class<br />

itself.<br />

Each message type also has an associated list of suggested types for a reply message. For example, the suggested reply<br />

type for is . Most importantly, the decomposition of<br />

message types <strong>in</strong>to a k<strong>in</strong>d-of hierarchy allows the automatic generation of helpful user <strong>in</strong>terfaces. For example, once the<br />

system knows that the user is writ<strong>in</strong>g a , that determ<strong>in</strong>es which fields are offered for<br />

fill<strong>in</strong>g and what defaults are presented. Fields hav<strong>in</strong>g to do with software bugs or New York Times articles are not<br />

presented and fields such as and may be helpfully defaulted with the usual room and time.<br />

What did Malone's team learn from this?

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!