An Interview with XML

by Dan Zambonini

I pulled a few strings and managed to get an exclusive interview with XML,
who's really hot right now. Sitting in a café
just off Venice Beach, this is what she had to say:

Me: Hi XML, thanks for talking
to me. I know you're not usually very flexible with these kinds of

XML: No problem, I am:

  • Glad to be here

  • Pleased to meet you

Me: I wanted to ask about your
relationship with HTML, which has been in all the papers. It's on,
it's off, it's back on again... What's the story? In the photos I've
seen, you don't seem like a very happy couple.

XML: Well:

  • The media

    • Have made a big deal out of

    • Always take photos of me
      looking bloated

  • Me and HTML

    • Are trying to agree to a
      long-term settlement

Me: To be honest, I think it's a
bit sick anyway. You're pretty much each other's parent, child and
partner all at the same time. You make me want to throw up. What
does your ex-husband think about this?

XML: I'd rather not talk about:

  • My relationships

  • My husband (type = "previous"), is:

    • A nasty little schema

    • A recovering alcoholic

      • With a bad case of the DTDs

Me: Fine. So let's talk about
your new single, "No REST For The Wicked". You were supposed to
be performing it live last night, but couldn't; what happened?

XML: I had to cancel:

  • My guitar was OK

  • One of my &s

    • Broke, and ruined the whole

Right, that's enough of that

I'll tell you what got me started on
this: the essence of web sites hasn't really changed at all in the
last 10 years. We're still using a similar data format, with similar
navigational devices.

And I'm not sure that the hierarchical
organisation of information - on which these devices are built - is
the best solution. I'm guessing that it developed from the structure
of books; chapters, sub-chapters and table of contents acting like
a hierarchical menu system and site-map. And the book's index
functioning very much like a typical keyword search ("you can
find instances of this word on these pages/resources

Books were obviously far more limited
in the ways in which we could record, interrogate and browse for
information. Is there no better modern way to structure and explore
web sites than through these rigid hierarchies? And without
resorting to a cumbersome 'advanced' search or wizard?

I'm not really sure what the answer is,
or if there is one. It may involve metadata, the semantic web, topic
maps, graphical interfaces, or none of these. Maybe the combination
of hierarchical browsing and free text search is near to the
optimum solution, and is partly responsible
for the ubiquity of the web today.

Has anyone seen anything on the web
that demonstrates a viable new process for exploring web sites?


2005-09-05 06:56:30
Other ways of navigation
There's delicious ( , which lets you define keywords ('tags') for pages, and lets you view pages that were tagged by other people, and StumbleUpon ( which lets you vote pages as thumbs up or thumbs down, and attempts to match your votes with other people's, so as you rate more websites, your 'stumbles' will be more and more likely to be interesting.
2005-09-05 14:52:20
Hierarchies and free text may well cover the majority of document indexing needs, but I reckon the challenge really starts when you want to find/navigate/use arbitrary data on the Web, not just documents.

There are various good demos of facetted kind of views knocking around, and then there's that hyperanimated thesaurus thing, but the only thing I can name right now is Longwell (sorry, it's late...)