trip 16.7 KB

Raw Blame History Permalink

<?xml version="1.0"?>
<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN"
    "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd">
<html xmlns="http://www.w3.org/1999/xhtml">
<head>
  <title>Dan's trip to KT2001 in Austin, Mar 2001</title>
  <meta http-equiv="Content-Type" content="text/html" />
</head>

<body>
<p><a href="../../">W3C</a> <a href="/People/Connolly/">Connolly</a></p>

<h1>KT2001 in Austin</h1>

<p style="text-align: center"><a href="#L647">Fri 2nd</a> &#183; <a
href="#L655">Sat 3rd</a> &#183; <a href="#L665">Sun 4th</a> &#183; <a href="#L687">Mon
5th</a> &#183; <a href="#L695">Tue 6th</a> &#183; <a href="#L703">Wed 7th</a> &#183; <a
href="#L711">Thu 8th</a></p>

<p><strong>Note</strong>: I'm making this trip report
<strong>world-readable</strong>, but some of links go to confidential
materials.</p>

<h2>Prologue</h2>

<p><a href="http://www.knowledgetechnologies.net/">Knowledge Technologies
2001</a> is a new GCA conference; when I first heard about it last August
during <a href="/Team/connolly-200011#2000,1">my trip to Extreme Markup
Languages</a>, I was very interested to help get it going because</p>
<ul>
  <li>its focus is very hear the heart of the Semantic Web, and</li>
  <li>it was in Austin, where I learned most of <a
    href="/Collaboration/knowledge">what I know about knowledge
    representation</a>.</li>
</ul>

<p>Eric Miller and I served on the <a
href="http://www.gca.org/attend/2001_conferences/kt_2001/welcome.htm"><strong>conference
board of advisors</strong></a> (see also: <a
href="http://www.egroups.com/group/kt2001-board">kt2001-board</a> egroup)
along with Steve Pepper et. al. At first, I expected the board of advisors to
be involved with all aspects of planning the conference, and I was surprised
that, for example, the CFP went out without consulting the board. Topicmaps
were much more heavily represented in the conference submissions; later we
realized that the CFP had gone to all the topicmap fora and none of the RDF
fora, which at least partly explains the difference. But <strong>development
of RDF tutorial presentations</strong> remains a concern (as does
<strong>Semantic Web Evangelism</strong> in general).</p>

<p>In the end, though my contribution to the board was not very substantial,
but I benefitted considerably: I got on the program twice; once to give the
traditional W3C part of the standards update, and once to present DAML.</p>

<h2><a name="L647" id="L647">Fri, 2 Mar</a></h2>

<p>Mary and the boys came a long for the trip; we travelled by car (expense
report note: $0.34/mi up to air fare price, per <a
href="http://web.mit.edu/afs/athena.mit.edu/org/c/cao/www/travpol2.htm">MIT
policy</a>).</p>

<p>I got home from <a href="../02dc-bos/Overview.html">my trip to Cambridge
for the all-wg shin-dig</a> late Thursday night; actually, 1am Friday morning.
We were planning to leave at 1pm Friday afternoon; I spent about</p>
<ul>
  <li>2 hours trying to get tools and background materials for putting
    together my presentations onto my new laptop</li>
  <li>5 or 6 hours sleeping</li>
  <li>5 more hours doing a little office work, packing and fiddling with the
    laptop</li>
</ul>

<p>before I decided it was time to cut bait and go. We drove to Oklahoma City,
where Mary's sister and her family hosted us for the night.</p>

<h2><a name="L655" id="L655">Sat, 3 Mar</a></h2>

<p>Saturday, we bid farewell to the cousins and headed for Austin, arriving
just after 7pm; 737 miles from KC. We checked in to the Radisson Hotel &amp;
Suites Austin, <a href="tel:+1-512-478-9611">(512) 478-9611</a> and settled
down for the night.</p>

<h2><a name="L665" id="L665">Sun, 4 Mar</a></h2>

<p>Sunday we visited our old church, <a
href="http://www.northwestfellowship.com/">Northwest Fellowship</a>. We caught
up a bit with our friends there, parted with an over-optimistic "we're in town
for most of the week; maybe we can get together for dinner one evening..."</p>

<p>Then we went to a picnic at eastwoods park that was the tail end of the <a
href="http://www.cs.utexas.edu/users/cline/alumni/AlumniActivities.html">Dean's
scholars reunion</a>. The Dean's scholars is an organization that helped me
find my way among the 50,000 students when I was at <a
href="http://www.utexas.edu/">U.T. Austin</a>. Unfortunately, I got there just
as Dr. Cline was leaving.</p>

<p>Then we paid <a href="http://www.cs.utexas.edu/users/boyer/">Dr. Boyer</a>
a social call, introducing Mary to him and his home, which is one of the
oldest in Austin. I left him with yet another overly-optimistic "we're in town
for most of the week; maybe I can come by your office and chat about this web
logic stuff..."</p>

<p>We went to <a href="http://www.schlotzskys.com/">Schlotzky's</a> for
dinner: "Funny Name. Serious Sandwich.(tm)". Yumm. They had iMac's set up for
net access, but it was mostly worthless: not only could I not ssh in to get my
mail, I couldn't send mail with my yahoo account; they had it blocked,
presumably to avoid privacy risks.</p>

<p>Even though we didn't make it to the <a
href="http://www.io.com/house/introkites.html">Annual Zilker Kite
Festival</a>, it was another wonderful sunny Sunday in Austin. Too much
traffic, though; we don't miss that.</p>

<p>Sunday night I had a convenient distraction from writing my slides: <a
href="http://lists.w3.org/Archives/Team/w3t-pr/2001Mar/0015.html">formulating
our response</a> to an urgent request from Lenat to contribute a quote for
their OpenCyc release. Turned out to be something of a tempest in a teapot, in
retrospect.</p>

<h2><a name="L687" id="L687">Mon, 5 Mar</a></h2>

<p>I got up Monday and put the finishing touches on my <a
href="../Talks/0305-kt-upd/">W3C Standards Update slides</a>. I used IE5's
scheduled favorites to cache them locally -- or so I thought -- and rushed off
to the conference to catch most of...</p>

<h4>Doug Lenat's keynote</h4>

<p>Some notes I scribbled on my copy of the program:</p>
<ul>
  <li>OpenCyc became public knowledge at 15:32Z, with a press release to
    follow Tuesday at 10:30am local time. It's a series of quarterly releases,
    with increasingly liberal license terms, increasing depth and breadth of
    the knowledge base, increasing functionality of tools.</li>
  <li>Lenat contrasted Cyc's approach with the traditional "full understanding
    or bust" approach. Harks to our own principle of <a
    href="/DesignIssues/Evolution#PartialUnderstanding">Partial
    Understanding</a>.</li>
  <li>Something he said about giving up global consistency reminded me of the
    way the Web sacrificed consistency for scalability (cf <a
    href="/DesignIssues/Topology.html">Topology</a> in <a
    href="/DesignIssues/">Design Issues</a>)</li>
  <li>In response to a "how many contexts will there be?" question, Lenat
    answered that rather than treating contexts as arbitrary first class
    objects, their approach has shifted to treating them as points in a
    multi-dimensional context space. cf <cite><a
    href="http://www.cyc.com/context-space.txt">The Dimensions of Context
    Space</a></cite> among the <a
    href="http://www.cyc.com/publications.html">cyc publications</a>.</li>
</ul>

<p>Next up was...</p>

<h4>Nic Fulton keynote</h4>

<p>Nic claimed that Reuters had been doing knowledge management since 1851 and
went on to present lots of evidence of such in a wonderfully insightful
manner. Analogies like:</p>
<dl>
  <dt>Industrial Revolution</dt>
    <dd>tupperware parties</dd>
  <dt>Information Revolution</dt>
    <dd>online communities</dd>
  <dt>Tidbits:</dt>
</dl>
<ul>
  <li>Reuters has 2000 journalists/cameramen</li>
</ul>

<p>He cited Reagle's <cite><a href="/TR/md-policy-design">Eskimo Snow and
Scottish Rain</a></cite> as background [er... to what? I forget what,
exactly]. And he told a story about G. W. Bush's campaign and google where the
punch-line was: links in the web need labels (aka properties ;-) in addition
to start and end addresses. And he praised standardization, explaining how
ISO-xyz and ISO-pdq container standards had revolutionized the shipping
industry in the 1970's, eliminating the need to unpack trucks and repack the
cargo at the dock.</p>

<h4>Scott Cooper, Lotus</h4>

<p>Another nifty keynote; this time with lots of practical experience
deploying groupware. For example, they studied collaboration tools and made
some profoundly mundane observations about what makes a discussion forum
effective:</p>
<ul>
  <li>keep the mission of the forum visible ("in your face")</li>
  <li>identify a facilitator</li>
</ul>

<p>With this insight, for a negligible investment in software development,
they were able to increase the effectiveness of their groupware software a
zillion times over.</p>

<h4>Standards Update, 11am</h4>

<p>At the break, I discovered that IE had downloaded only the first 8 or so of
my 15 or so slides. <strong>Panic!</strong> I found an analog phone, logged
in, and discovered that I had chacled my slides after producing the first 8,
but not the rest.</p>

<p>That hurdle crossed, I went on stage to discover that I'd need to reboot to
get my machine to produce a video signal. I sat down to do so, only to
discover that Ron Daniel had moved my chair out of my way.
<strong>Plunk</strong>, down I go behind the table, recovering with
"<strong>tada!</strong>" clown dance to delight the audience a moment later. I
don't think they'll soon forget me. And yes, I was wearing my W3C shirt.
;-)</p>

<p>Dr. James Mason gave the ISO/IEC JTC-1 SC-blah-blah update while I
recovered. I noticed quite a bit of <strong>overlap betwen work in ISO and
other stuff</strong> (@@TODO: take another look at his slides, which should be
on the conference website RSN). I made a note to ask them to participate in
the <a
href="http://lists.w3.org/Archives/Team/ietf-new-work/">new-work@ietf.org</a>
thingy.</p>

<p>Then I presented the W3C part:</p>
<ul>
  <li><a href="../Talks/0305-kt-upd/">slides</a></li>
  <li><a href="http://heddley.com/edd/2001/03/07/m_austin-018.jpg">photo</a>
    by Edd D.</li>
</ul>

<p>Notes on questions from the audience:</p>
<ul>
  <li>something about whether XML Schema evolution; would templates be
    supported? I answered that there were tools to convert from DTDs, UML, and
    other more familiar notations, hoping that's what the guy meant by
    templates. Not sure I was on target at all, though.</li>
</ul>

<h4>Turner on Web Services</h4>

<p>I went to Turner's talk on Web Services: WSDL, UDDI, etc. It was aimed at
managers, so there wasn't much technical beef, save one reminder to look at pi
calculus. A tidbit:</p>
<ul>
  <li>&gt;200 businesses participate in uddi.org</li>
</ul>

<h4><a name="lxdev">Lunch with the xml-dev crowd</a></h4>

<blockquote>
  <h3>Lunchtime Insights</h3>

  <p>At conferences such as these, it's often who you sit next to at lunch
  that leads to some of the most interesting conversations and insights into
  the work of various developers and organizations. Today was no exception,
  and I found myself seated with XML.com author and O'Reilly editor Simon
  St.Laurent going head-to-head with Dan Connolly from the W3C. Anyone who
  thinks that the rages of XML-DEV can't be reproduced in the flesh should
  have been there.</p>

  <p>Beyond the flying sparks, the lunch gave some interesting insights into
  the workings of the W3C and the people behind the organization. One
  encouraging sign was that both Connolly, the W3C's XML lead, and Eric
  Miller, the new lead for the Semantic Web activity, were strongly in favor
  of as much openness as possible -- an encouraging sign for future
  development of at least Semantic Web related technology at the W3C.</p>
  <address>
    <p><a href="http://www.xml.com/pub/a/2001/03/07/kt1.html">Knowledge
    Technologies 2001: Conference Diary</a> by Edd Dumbill March 07, 2001</p>

  </address>
</blockquote>

<p>Some details... Simon St.L. mentioned his strong disagreement to
the spread of XML Schema types into XPath, but when I asked him to
send his comments on the XPath-revision requirements to the address
indicated in that document, he said it's more cost-effective to be an
outside agitator that to participate directly in the W3C process.  I
tried to explain that we're very responsive, at least in the
www-xml-schema-comments and xmlschema-dev forums, but I had to admit
that W3C as a whole W3C has a spotty track record with regards to
responsiveness.</p>

<h4>Implementor Track: Eric van der Vlist</h4>

<p>Good stuff: Topicmaps generated from an RDF DB. Also, RSS. All open source.
"One of the good guys." See <a href="http://xmlfr.org/">xmlfr.org</a> and <a
href="http://4xt.org/">4xt.org</a></p>

<h4>Implementor Track: Uche Ogbuji</h4>

<p>Another good guy, showing how <a
href="http://OpenTechnology.org/">OpenTechnology.org</a> works. EricM probably
grokked more of the technical details.</p>

<p>Based on his <a
href="http://www-4.ibm.com/software/developer/library/ws-trans/index.html?dwzone=ws">WSDL
processing with XSLT (First steps for Web service description processing)</a>,
I encouraged him to come to our Web Services workshop.</p>

<p>I spent Monday evening with family and friends.</p>

<h2><a name="L695" id="L695">Tue, 6 Mar</a></h2>

<p>Tuesday started with a 9:45 am meeting with Freese, Miller, Pepper re
XTM/Semantic Web coordination; see <a href="xr23">shared notes
w/Eric</a>).</p>

<p>I didn't go to the 10:30am press conference about the <a
href="http://www.cyc.com/opencycpressrelease03062001.html">Cycorp OpenCyc
press release</a>. Oops.</p>

<p>I had an interesting chat with <strong>Marc Edgar of GE</strong>; I
explained RDF to him and got a gratifying "yeah, that's pretty simple and
powerful" response. We commiserated over the over-hypedness of JINI and
confessed our shallow understanding of UML to each other, trying to puzzle
over the UML diagrams at the back of the <a
href="http://www.topicmaps.org/xtm/1.0/">topicmap spec</a>. He offered to put
me in touch with the <strong>EDI folks at GE</strong> (which is about 2/3rds
of the world's EDI expertise, I gather).</p>

<p>Over lunch, Eric M. helped me put together my <a
href="../Talks/0103daml-kt/">DAML slides</a>, based on a "briefing" by Mike
Dean. I think it came together well.</p>

<p>The audience asked <strong>great questions</strong>! Almost all of them
raised their hands when I asked if anybody had heard of OIL and SHOE. They
asked about the limits of formal systems, trust and context, and all the good
stuff.</p>

<p>John Sowa was scheduled to speak after me, but alas, the New England
weather held him captive.</p>

<p>Steve Newcomb presented the <strong>topicmap processing model</strong>
instead. Steve said he put the talk together in a hurry, but I think it was
just about my speed; I think the lightbulb finally went off. This is really
their data model, but the discussion of it in terms of processing obscured it
from my view for the longest time; their discussion of "topic merging" treats
it as a computational effect, rather than a logical inference; as if the
inference that 2+2=4 were a consequence of some physical computational process
necessarily happening over some amount of real time.</p>

<p>I spent Tuesday evening with family and friends again (different friends
this time).</p>

<h2><a name="L703" id="L703">Wed, 7 Mar</a></h2>

<p>[@@story-telling peters out at this point.]</p>

<p>Ontopia demo; <a href="http://www.techquila.com/">techquila</a> "a shot of
the hard stuff"; what was the guy's name? khaled(sp?) long black curly hair.
Is this <a href="http://www.ontopia.net/software/tmproc/">tmproc</a> thingy
what he was showing me?</p>

<p>during the demo, I learned <strong>topicmaps are undirected</strong>; all
links go both ways (scalability no-no!)</p>

<p>2p closing panel</p>

<p>3p Edd D's closing (analagous <a
href="http://www.xml.com/pub/a/2001/03/07/buildingsw.html">article</a>)</p>

<p>Friday's with EricM, Edd D, Dave B. (<a
href="http://heddley.com/edd/2001/03/07/m_austin-019.jpg">photo</a> by the
waiter)</p>

<h2><a name="L711" id="L711">Thu, 8 Mar</a></h2>

<p>drive back to KC. Long drive. left around 11am; couldn't afford to stop and
see the cousins on the return leg; Justin was bummed. Got home around 3am.</p>

<h2>Colophon: how this page was made</h2>

<p>see also: <a href="../02dc-bos/Overview.html">BOS Feb trip</a> (including
links to earlier trips)</p>

<h2>TODO</h2>
<ul>
  <li>put slides on /Talks/; on homepage?</li>
  <li>expense report</li>
  <li>read the topicmap spec again, trying to remember what I wrote on my lost
    copy (which wasn't too much: definitions out of context considered
    harmful, some character encoding issue, maybe a few other things)</li>
  <li>RDF/n3 model of topicmaps? merging rules look straightfoward</li>
</ul>
<hr />
<address>
  in progress Mar 2001<br />
  <a href="http://www.w3.org/People/Connolly/">Dan Connolly</a><br />
  <small>$Revision: 1.18 $ of $Date: 2001/03/16 04:42:29 $ by $Author:
  connolly $<br />
  <a href=",access">ACL</a></small>
</address>

<p></p>
</body>
</html>