Tuesday, October 08, 2002

Damn this pervasive liberal hegemony

This Modern World. Conservatives with an attitude! [Salon.com]


Satirical cartoon by Tom Tomorrow

With the kind permission of Tom Tomorrow.

08/10/2002 08:05 by Matt Mower | Permalink | comments:
More about:

A personal manifesto for growth

Manifesto for Growth

via Absolute One:

  1. Allow events to change you
  2. Forget about good
  3. process is more important than outcome
  4. love your experiments like ugly children
  5. go deep
  6. capture accidents
  7. study
  8. drift
  9. begin anywhere
  10. everyone is a leader
  11. harvest ideas, edit applications
  12. keep moving
  13. slow down
  14. don't be cool (cool is conservative fear, dressed in black)
  15. ask stupid questions
  16. collaborate
  17. an image which email won't replicate
  18. Allow space for ideas you haven't had yet
  19. Stay up late
  20. Work the metaphor
  21. time is genetic
  22. repeat yourself
  23. make your own tools
  24. stand on someone's shoulders
  25. avoid software (everyone has it)
  26. don't clean your desk
  27. don't enter awards (its bad for you)
  28. creativity is not device dependent
  29. organisation is liberty
  30. don't borrow money
  31. listen carefully
  32. take field trips
  33. imitate
  34. make mistakes faster
  35. scat (break it, stretch it, crack it, fold it)
  36. explore the other edge
  37. coffee breaks, cab rides, ream (?) rooms
  38. avoid fields, jump fences
  39. laugh
  40. remember
  41. power to the people

[via NotExactly] [via Sebs Open Research] [via The Universal Church Of Cosmic Uncertainty]

» Thought provoking list.  I would add:

  • always write it down
  • listen to lots of good music
  • seek first to understand, then to be understood
  • review often
  • ride change
  • go do something different instead
  • ...


08/10/2002 09:58 by Matt Mower | Permalink | comments:
More about:

Let's all not sing along

In his famous speech at the O'Reilly Open Source Convention in July, Lawrence Lessig framed the same point as a four-stanza refrain for a song:

  • Creativity and innovation always builds on the past.
  • The past always tries to control the creativity that builds upon it.
  • Free societies enable the future by limiting this power of the past.
  • Ours is less and less a free society.

[Via LinuxJournal]

08/10/2002 10:18 by Matt Mower | Permalink | comments:
More about:

Water - Slush - Ice, and a slice

The metaphor describes how innovations go from ideas to implemented projects. Here's a diagram that illustrates this process:

water to ice:

Weblogs are an excellent example of a "highly networked community that encourages innovation." The water to ice metaphor describes a way to move these ideas from interesting conversation to successful projects.

Interesting piece from back in July by Andy Chen [Kumquat's musings]

» I'm not sure yet whether idea management is a leaky pipe, but if it is then knowledge logs are, as Andy says, a very good way to handle the water-slush-ice transition.

Andy's weblog is new to me, but chock full of interesting ideas and insights.  I'll be reading more.

08/10/2002 10:33 by Matt Mower | Permalink | comments:

So angry

Sharon hails 'successful' Gaza raid. Israel's prime minister defends an army operation which killed 14 Palestinians¸ shrugging off international condemnation. [BBC News | WORLD]

» Okay, I don't understand all the issues. I can't claim to have fully appreciated all the claims, counter-claims, greviances and disputes.

But just get the fuck outta their country!


[Addendum to the above rant: I see a Jewish settler has been killed in an apparent reprisal.  I really should stop reading about the Middle East, it makes my head spin.  I should probably stop commenting on it too.]

08/10/2002 12:25 by Matt Mower | Permalink | comments:
More about:

XFML 1.0 (CORE) Published today

Peter Van Dijck has published the XFML 1.0 Core specification today.

Important points about XFML:

  • XFML lets you exchange hierarchical faceted metadata.
  • It also lets you indicate topics in different published XFML documents are equal, thus allowing you to reuse indexing efforts.
  • Finally, XFML lets you build connections between different XFML maps, by indicating that a topic in one map is equal to a topic in another map: we call this connecting topics, or that a topic is described on a certain resource (a webpage usually), we call this published subject indicators.


08/10/2002 15:33 by Matt Mower | Permalink | comments:

Happy days are here again (for ever and ever)

New Scientist.  Skin-based networks that run at 10 Mbs may revolutionize P2P.   On a lighter note, imagine a party where people shared music and movies with physical contact. [John Robb's Radio Weblog]

» I can just see Jack Valenti on the phone to Coble right now:

"Howard pull your damn finger out!  We have to do something about these kids kissing and touching each other!  They're stealing our f**king content!"

Cue endless reruns of Happy Days!

08/10/2002 18:41 by Matt Mower | Permalink | comments:
More about:

Let's move on.

Yet another week where one union is throwing our economy in the toilet. I wanna protest these jerks, but don't know how to do it. Why don't we go out to the docks on Sunday and protest this union who doesn't give a crap about the rest of us? By the way, their Web site sucks. So typical.

[The Scobleizer Weblog]

» I really like Robert Cassidy's comment to a post on John Robb's weblog about the same topic.  Read it (along with the other comments that I like a lot less) here.

08/10/2002 19:01 by Matt Mower | Permalink | comments:
More about:

XTM and XFML: more cousins than competitors

A faceted classification standard.

XFML 1.0 (CORE) Published today. [Curiouser and curiouser!]

Does this have the same purpose as XTM (topic maps)? What are the differences?

[Seb's Open Research]

» Whilst XTM and XFML do have many similarities (and theoretically you could represent any XFML document using XTM -- I think) they are different.

XTM was designed to be a generalized format for representing arbitrary topic relationships. The upshot is that XTM, whilst expressive, is relatively complicated.  XFML is more focused and so, IMO, easier to get going with.  XTM can support arbitrary, complex, relationships among topics.  XFML supports fewer simpler relationships.  Don't go getting the idea that XFML is inferior though.

One of XFML's guiding principles is that it be focused and easy to implement.  In this I think it succeeds admirably.  The spec is only about 8 or 9 pages long.

In an XTM document everything is a topic or relationship.  This means you can model arbitrary structures, but this very power makes XTM quite complex and an individual XTM document is not necessarily easy to understand.

By contrast XFML defines just three structures:

  • topics (can belong to a single facet)
  • facets (can group many topics)
  • pages (can have topics as occurrences)

For most web-based applications these three concepts are sufficiently expressive.  Topics can have a parent (but only 1, which must be within the same facet).  A facet thereby is a hierarchy of topics.  So an XFML document contains a number of topic hierarchies which each define a seperable metadata concept.

To understand this idea imagine you define topics under the facet Date of Publication like:

  Jan 2000
    1 Jan 2000
    2 Jan 2000
  Feb 2000

Each page in the XFML document will have an occurence of a topic like "1 Jan 2000" indicating its date of publication.

Another facet could be Author with topics like:

  Jon Udel
  Bob Lewis
  Matt Mower

and again each page in the XFML document will have an occurrence of the appropriate author topic.

The first thing to notice is that it probably doesn't make sense for a topic from the Author facet to appear in the Date of Publication facet (and vice-verca)  They really are orthogonal concepts.

The other thing is that because the topics are hierarchical we can start off with a general filter and drill down.  These two facets would allow you to immediately restrict the range of pages you were looking at to:

  • only those published by InfoWorld (or Novissio)
  • only those published in a specific year

Drilling down further into either facet will filter to an even narrower (more focussed) set of results.  This is a very powerful tool if you have the right facets and appropriately defined topic hierarchies (for your application).

For a much clearer and more succint definition read  David Gammel's recent post to the xfml group here.

The other powerful concept embodied directly in XFML is the idea of connecting topics together.  This allows me to say, within my map, that:

my topic X = your topic Y

Which is a very powerful, decentralised, way of sharing your indexing efforts without requiring that everyone use the same topics/terminology.  For building real-world topic maps among groups of disconnected people (such as those in different organisations) this could be essential.

Okay I've about run out of steam for the moment.  Hopefully this was useful though.

08/10/2002 22:34 by Matt Mower | Permalink | comments: