Re: A bit of bomb throwing....

From: CoC Mail <>
Date: Wed, 18 Jan 2006 08:45:37 -0500

Some good points, but as someone studying usability and HCI (so of
course it's my hammer for this nail :) ) it seems to me that a good
chunk of the problem is also that we haven't quite managed to
demonstrate what interaction sequences for a system that uses 'semantic
web' concepts look like. In other words, we just don't know how to
design user interfaces for software that incorporates re-usability of
data - even in highly structured domains like calendaring/PIM, we hardly
have any mature applications, and the people who're trying to figure
that out are having a hard time of it: take a gander at the 325 Project
(, the PlatypusWiki
( and the much awaited Chandler
mail client to get an idea of the complexities involved. For all sorts
of reasons - lack of good cognitive models, assumptions about the
structure and nature of information, the lack of sufficiently intuitive
yet expressive formalisms, a lack of rigourous understanding of the
social use of information - the only way an average user can tell the
computer to "take this bit of data from here, mash it up in this
(completely predictable) way, and throw it there" is to write lots of
messy code and hack tools together (even the tools are themselves most
excellent, like the SIMILE stuff). And I'm not even going to talk about
intelligent agent-based architectures.

Real-world uses? We all keep talking about them, and we claim we know
they exist, but the only concrete real world uses I have heard of are
from computing systems designers and architects who want to make their
life -- when manipulating data -- easier. Enthusiastic hobbyists may
also be very well meaning, but as the crowd
shows, a deep look reveals that they often don't really know what they
are talking about... and that no one else does, for that matter. Give me
a 'real world use' and I will probably be able to show you why the
average google user won't be able to do a thing with it.

At the end of the day, however, I'm glad I was (and am continuing to be)
able to use the SIMILE tools to do parts of my Masters' thesis
(ironically on something as semanticweb-laden as social bookmarking
systems). It did take some work, and messing around with code and
endless hours of fiddling with the data I was generating/capturing. But
I know it enabled me to do certain things I could otherwise not have
done, and at this stage, I'd much rather have a tool I can adapt than be
handed someone else's idea of how I should be manipulating my data. This
way, I'll at least get some insight into how I can - in ways specific to
my life - interact with, create, and otherwise manipulate 'semantic web'
data, and perhaps contribute to the pool of ideas.

-- arvind

<long reply train cut>
Received on Wed Jan 18 2006 - 13:45:22 EST

This archive was generated by hypermail 2.3.0 : Thu Aug 09 2012 - 16:39:18 EDT