WWW2003 Conference IRC Chat Logs for 2003-05-22

This web site provides the chat logs for the WWW2003 Community Hub for attendee-generated notes and coverage of the Twelfth International World Wide Web Conference.

You are here: Logs Home / 2003-05-22

00:21:56 <uche|away> uche|away is now known as uche
05:55:02 <amy> amy is now known as amyv
06:52:12 <amy> amy is now known as amyb
06:52:20 <amyb> amyb is now known as amyvdh
06:53:00 <dajobe>http://www2003.org/schedule.htm#22
06:53:00 <dc_www2003> A: http://www2003.org/schedule.htm#22 from dajobe
06:53:16 <dajobe> A:|WWW2003 Day 2 Schedule - Thursday 22 May
06:53:16 <dc_www2003> Titled item A.
06:53:38 <dajobe> dajobe has changed the topic to: WWW2003, logged and blogged at http://www2003.xmlhack.com/ - Budapest, Hungary - May 20-24 - Second day of main conference
06:57:13 <dajobe>http://searchenginewatch.com/searchday/article.php/2208981
06:57:13 <dc_www2003> B: http://searchenginewatch.com/searchday/article.php/2208981 from dajobe
06:57:35 <dajobe> B:|Virtually Attending the 12th IW3C2 Conference (SearchEngineWatch)
06:57:36 <dc_www2003> Titled item B.
06:58:58 <dajobe>http://www.internetnews.com/dev-news/article.php/2210201
06:58:58 <dc_www2003> C: http://www.internetnews.com/dev-news/article.php/2210201 from dajobe
06:59:15 <dajobe> C:|W3C Adopts Patent Policy (InternetNews)
06:59:15 <dc_www2003> Titled item C.
07:00:46 <dajobe> "Page Rank" session
07:00:55 <amy> amy is now known as amyvdh
07:01:50 <dajobe> Sep Kamvar - Accelerating PageRank Computations by Quadratic Extrapolation
07:01:56 <dajobe> ... some of our recent work is getting media attention
07:01:57 <edd> morning.
07:02:00 <dajobe> ... some caveats
07:02:07 * edd in W3C XML track.
07:02:09 <dajobe> mayber ead NSF press release of 5x speedup
07:02:25 <dajobe> ... has been achieved with this and other work, this work has done 30% speedup, not 5x from this work
07:02:40 <dajobe> ... other thing is that we're speeding up google 5x, not the case
07:02:53 <dajobe> ... just one component of google; speeding up PageRank
07:03:09 <dajobe> PageRank (PR) is computed offline
07:03:16 <dajobe> distributed computation
07:03:24 <dajobe> why speed up PR?
07:03:28 <dajobe> pretty fast already
07:03:38 <dajobe> motivation is personalization and "freshness"
07:03:40 <edd> CMSMCQ up, talking about XML Core specs
07:04:02 <dajobe> so for example, "Giants" might be for different sports teams
07:04:13 <dajobe> want to compute 100s of PR vectors in a reasonable time
07:04:30 <dajobe> other reasons are to keep the PR fresh for sites that change a lot - news sites
07:05:13 <dajobe> he defines PR
07:06:35 <edd>http://www.w3.org/2003/Talks/0522-xml-core/
07:06:43 <dc_www2003> D: http://www.w3.org/2003/Talks/0522-xml-core/ from edd
07:06:49 <edd> D:|XML Core Specifications
07:06:50 <dc_www2003> Titled item D.
07:07:01 <edd> D:by C.M. Sperberg-McQueen
07:07:01 <dc_www2003> Added comment D1.
07:07:01 <dajobe> PR .. still describing PR and estimating it
07:07:15 <dajobe> ... propogating PR estimates
07:07:28 <edd> (xml talk) CMSMCQ running over XML origins, describing what XML is.
07:07:36 <mattb> guy from stanford makes note that their pagerank algorithm acceleration is 1) not accelerating google 2) only a (possible) 5-times speedup when combined with other algorithms - this one gives 30-50% improvement by itself
07:07:53 <dajobe> I think I typed that already :)
07:08:07 <mattb> heh, k
07:09:59 <dajobe> PR is the first eigenvector of the matrix A (described in the paper)
07:10:08 <zool> hehe
07:10:12 <dajobe> power method of finding the PR
07:10:15 <dajobe> hi zool
07:10:24 * zool smiles hey dajobe
07:10:40 <dajobe> lots of maths that don't transcribe to IRC ...
07:11:26 <edd> Projector malfunctioning in XML track.
07:11:38 <dajobe> not powered by XML then?
07:12:10 <edd> I think it's a moral reaction, I didn't see "XML Schema" preceded by the magic "W3C" :-)
07:12:25 <dajobe> :)
07:12:56 <edd> Bring on the layer cakes: http://www.w3.org/2003/Talks/0522-xml-core/slide8-0.html
07:13:47 <edd> Except being SVG I can't see it on my browser, as SVG is effectively a one-vendor standard :(
07:14:03 <dajobe> PR he describes how convergance on the required PR result works with the Power Method of computing PR
07:14:33 <dajobe> ... approach is to estimate and eliminate the components of the non-principle eigenvectors (2 onwarsd)
07:15:16 <dajobe> ... assume that the result can be estimated by the first 3 eigenvectors of A
07:17:41 <dajobe> ... describes how the estimate transforms the problem to solving a simpler goal
07:17:46 <dajobe> and refers to the paper
07:18:23 <dajobe> summary - make an assumption, solve for the dominate eigenvector from the first three terms, use power method to clean it up
07:18:35 <edd> CMSMCQ giving a good introduction to the XML spec family. Nothing new, but would be good for those who've never come across XML much.
07:18:45 <dajobe> empirical results - it works well, but not well enough for personalized pagerank
07:19:21 <dajobe> for 15% random (surfer) jumps, we get about 30% speedup
07:19:29 <dajobe> for larger random jumps, bigger speedups
07:19:36 <dajobe> these are rather unrealistic
07:19:37 <dajobe> s/larger/smaller/
07:19:53 <dajobe> .. 1% random jumps aren't realistic
07:20:19 <dajobe> take-home method: does speedup PR but initial goals were for personalization speed, it doesn't do it for that
07:20:22 <edd> Liam rocks. "I need to know how to get to a URL -- this is both a browser and an operating system I don't use" (to CMSMCQ, asking him how to get IE running on his Windows laptop).
07:20:26 <dajobe> ... it is powerful ideas for further speedup
07:20:37 * edd starts a conference quotefile
07:20:41 <dajobe> lol
07:20:59 <dajobe> ... quadratic extrapolation as used her ecan be used for a variety of problems
07:21:19 <dajobe> ... this is a classic example of what turns up in numeric algorithms
07:21:32 <dajobe> ... what we are doing at present is natural extensions
07:21:43 <edd> (Liam really is struggling with Windows, poor fellow)
07:21:46 <dajobe> ... trying to use 2nd-4th
07:22:02 <dajobe> ... would also be better if the 2nd eigenvalue onwards were known
07:22:09 <dajobe> ... Q&A
07:22:15 <edd> Liam now talking about XML Query.
07:22:49 <amy> amy is now known as amyvdh
07:22:57 <edd> The URI wasn't on screen though, so I can't find the talk. Same monitor issues making it difficult to concentrate on screen.
07:23:56 * _Yves notes that presentation URI is http://www.w3.org/2003/Talks/05-05-liam-xquery/
07:24:17 <dom> linked from http://www.w3.org/2003/03/w3c-track03.html
07:26:15 <edd> thanks both
07:26:20 <edd>http://www.w3.org/2003/Talks/05-05-liam-xquery/
07:26:20 <dc_www2003> E: http://www.w3.org/2003/Talks/05-05-liam-xquery/ from edd
07:26:45 <edd> E:|Liam Quin -- XQuery
07:26:46 <dc_www2003> Titled item E.
07:27:13 <edd> "The best thing about XQuery is that it feels like writing scheme" TV Raman (I believe) in the audience
07:27:34 * dajobe leaves this session
07:28:16 <edd> Liam asks for public feedback on XQuery's decision to omit some axes from XPath
07:29:14 <edd>http://www.w3.org/2003/03/w3c-track03.html
07:29:14 <dc_www2003> F: http://www.w3.org/2003/03/w3c-track03.html from edd
07:29:19 <edd> F:|W3C track programme
07:29:19 <dc_www2003> Titled item F.
07:29:22 <edd> F:Handy links to slides
07:29:22 <dc_www2003> Added comment F1.
07:34:32 * dajobe leaves this session
07:40:09 <dajobe> Semantic Web Services - Massimo & Toshiba
07:40:12 <shab> The panel session is cool at the mo, guy talking about providing different views onto digital cultural content - v.big data sets
07:40:26 <dajobe> nice
07:41:07 <dajobe> SWS - ontologies to describe WS
07:42:27 <dajobe> SWS - discover and select the provider (phase1), interact with WS provider (phase2)
07:42:35 <dajobe> shab: do tell us more :)
07:42:39 <edd>http://www.w3.org/2003/Talks/0522-dom-plh/
07:42:39 <dc_www2003> G: http://www.w3.org/2003/Talks/0522-dom-plh/ from edd
07:42:44 <edd> G:|DOM Level 3
07:42:45 <dc_www2003> Titled item G.
07:43:01 <shab> Panel: good point being made about augmented reality - providing different cultural views/stories on real objects
07:43:01 <edd> G:by Philippe Le Hégaret
07:43:01 <dc_www2003> Added comment G1.
07:43:09 <dajobe> SWS - capavility based sicovery using daml-s, uddi; automonomous invocating using daml-s process model and vm
07:43:32 <mattb> p2cast - p2p video on demand
07:43:42 <mattb> sounds something like a streaming bittorrent
07:43:51 <dajobe> heh:)
07:44:43 <shab> panel - showing reconstructions over time and use as educational tool e.g. visting historical places and compare over time
07:45:36 <dajobe> sws - daml-s profile, process model and grounding+WSDL match the WS infrastructure roughly for discovery, choreograph and invocation of services
07:46:09 <edd> Groan. Laptops changed, but the projector still malfunctioning.
07:46:59 <shab> panel - talking about development and need for Giga-bit and Peta-bit/s rates to deal with these very high res re-creations? is that really necessary
07:47:48 <dajobe> sws - daml-s profile repreents cabailities of the WS
07:48:26 <dajobe> sws - requester describes it's ideal WS as a daml-s profile, the matching engine then hunts for close matches of that description
07:49:05 <shab> panel - talking about different cultures' views of objects and place of cultural objecgts and environments in cultures - being a bit sterotypical but interesting
07:50:10 <shab> panel - talking about national cultural herritage networks e.g. canadian, russian, china, US cultural grid...
07:50:49 <shab> panel - EU network (E-Culture Net)
07:51:24 * edd tries to get folk in GNOME talking about building scriptable SVG component
07:51:38 <dajobe> sws - using t-models to represent daml-s within uddi. Lots of mappingsbetween multiple tmodels
07:51:45 <shab> panel - ends by suggesting need to look at World level
07:52:29 <dajobe> sws - already got Q&A in the talk...
07:52:40 <shab> panel - last presenter was Kim Veltman
07:52:47 <dajobe> ah, Kim :)
07:52:59 <shab> ?
07:53:15 <dajobe> sws - architecture for daml-s/uddi
07:53:41 <dajobe> shab: i've heard it before from him
07:54:00 <shab> panel - new presenter - Susumu Sawai
07:54:25 <dajobe> sws - daml-s into the system are matched, translated to uddi and then talk to the uddi registrty.
07:54:57 <dajobe> sws - adv of WS, UDDI such as security infrastructure, linked to SW tech
07:55:13 <dajobe> sws - value fromd aml-s capavillities, matching
07:55:31 <shab> panel - visual archive for education and culutre - GENES Project in Japan(?)
07:55:39 <dajobe> sws - limitations iinc preconditions and effects for I//O, need tighter integration within UDDI
07:55:53 <edd> Philippe shows a handy diagram of how DOM events work.
07:56:03 * dajobe invites bijan to comment on "sws - " track
07:56:49 <dajobe> sws - invocation (after selecting a provider)
07:57:08 <dajobe> sws - interaction requires shared protocols, content messages, ports and low level detail
07:57:28 <shab> panel - GENES project gov funded and seems very big - 1200 school teachers, 22 CDs, 3 DVD and 300 Web pages...
07:58:02 * bijan comments that dajobe is doing a fine job of scribing this track
07:58:41 <dajobe> sws - process model - a detailed view of the WS and an interaction protocol
07:59:01 <shab> panel - seems a bit like NOF Digi project in the UK
07:59:38 <dajobe> yeah, I think I've heard of GENES
07:59:51 <shab> panel - not sure but seems to be saying that they have a central repositary and metadata standards...
08:00:11 <dajobe> sws - grounding specifies the mapping to WSDL
08:01:47 <dajobe> sws - daml-s vm, interacts with the requestor - contains a ws invocation part and daml-s processor and inference engine (Jess kB, Jess)
08:02:12 <shab> panel - can't find any web sites on the project(s)?
08:02:15 <dajobe> sws - vm takes wsdl in at ws side, daml-s on the daml-s side. The daml-s side talks to the reasoning system although that is still under development
08:02:52 <dajobe> sws - vm then talks to the WS via SOAP
08:03:30 <edd> ralph up now,presenting reagle's slides
08:03:58 <edd>http://www.w3.org/2003/Talks/0513-xmldsig-xenc-xkms/
08:03:59 <dc_www2003> H: http://www.w3.org/2003/Talks/0513-xmldsig-xenc-xkms/ from edd
08:04:10 <shab> panel - ubiqutious e-learning any where, any... ra ra
08:04:17 <dajobe> sws - DAMLzon - daml-s for amazon.com
08:04:18 <edd> H:|XML Security: XML Signature; XML Encryption; XKMS
08:04:19 <dc_www2003> Titled item H.
08:04:25 <dajobe> shab: lol
08:04:32 <edd> H:by Joseph Reagle, presented by Ralph Swick
08:04:32 <dc_www2003> Added comment H1.
08:05:22 <dajobe> sws - performance, mostly time was amazon ws invocation
08:05:23 <shab> panel - XML as basis of elearning infrastructure...
08:05:33 <dajobe> sws - results in milliseconds
08:06:13 <shab> panel - privacy & legal issues... of wide scale implimnetation
08:06:17 <dajobe> sws - B2B demo with Toshiba to organize a supply chain for computer manufacturing
08:06:40 <shab> panel - presenter ends
08:07:55 <shab> panel - still can't find anything about the GENES project...
08:08:36 <shab> panel - Liddy Nevile starts
08:10:32 <dajobe> sws - conclusion - daml-s supports diff parts of WS aspects, does not result in a (large) performance penalty
08:13:25 <dajobe> sws - q&a
08:19:19 <dajobe> sws - timbl asks have you considered making a query used with extra knowledge from other DBs that weren't related to WS specifically such as geo info, weather,...
08:19:58 <dajobe> sws - ... most interesting things wouldn't be about WS, but from other sources
08:20:14 <dajobe> sws - MM replies - main point was to express capabilities
08:21:19 <dajobe> sws - requires you express things in terms of capabilities, then we can do hte inferenceing
08:23:00 <dajobe> sws - MM mentions YA "the huge problem with the SW is"
08:23:25 <dajobe> sws - in this case, having lots of ontologies describing the same thing(s)
08:23:50 <dajobe> sws - haven't looked much at the ontology translation problem; it would be somewhere in the daml+oil reasoner, skeptical putting it in the matcher
08:23:55 <dajobe> sws - END
08:24:46 <shab> panel - semantic web as way of linking cultural infomratoin and persepctives
08:26:01 <shab> panel - SVG as means of represenign, describing and analysing images e.g. comparisons
08:28:16 <shab> panel - people in communities adding photos (in context of family conections, important in Abroiginal peoples)... represening family structures perhaps use FOAF and build up
08:30:07 <shab> panel - in aboriginal communities the relationships don't map easily to simple family tree approach
08:31:54 <shab> panel - members of community tend to use graphical means of representing relationships, RDF is a very good map to this and hence they can easily work with developers
08:31:54 <shab> panel - Liddy ends
08:33:32 <shab> panel - discussion... poetic net not semantic net - I can't really understand what he is saying but I like the concept of a 'poetic net'
08:35:15 <shab> panel - discussion - high density of sematnics in poetry about art adn cultural objects...
08:38:21 <shab> panel - new presentation about UNESCO projects based a about World heritage... new projects on endangered cultures...
08:39:08 <amy> amy is now known as amyvdh
08:50:19 * DanC waves, wonders what session anybody's going to at 11
08:52:20 <nmg> web services for me
08:53:51 <shab> education
09:01:01 * DanC waves to edd, looks around for sandro and EricM
09:01:18 <DanC>http://geourl.org/near/?lat=47.51&lon=19.08&dist=40
09:01:18 <dc_www2003> I: http://geourl.org/near/?lat=47.51&lon=19.08&dist=40 from DanC
09:01:32 <DanC> I:|sites near 47.51, 19.08, and within 40 miles
09:01:32 <dc_www2003> Titled item I.
09:01:41 <DanC> I:i.e. Budapest, Hungary
09:01:41 <dc_www2003> Added comment I1.
09:04:12 <shab> education - Thorsten Hampel - Experience With Teaching and Learning in Cooperative Knowledge Areas
09:07:21 * DanC wonders how widely this channel/blog has been announced
09:08:18 <shab> edu - talkign about their system STeam - sounds like a VLE
09:11:50 <shab> edu - learnt - need a natural metaphor, create and keep real and virutal learning groups linked
09:12:48 <shab> edu - 'blended' learning used as term for this matching of real/virtual
09:14:36 <edd-lap_> edd-lap_ is now known as edd
09:14:55 <shab> edu - use fairly standard metaphor of 'learning rooms'
09:15:54 <edd>http://www.w3.org/2003/Talks/0521-hh-wsa/
09:15:55 <dc_www2003> J: http://www.w3.org/2003/Talks/0521-hh-wsa/ from edd
09:16:05 <edd> J:|Web services architecture
09:16:05 <dc_www2003> Titled item J.
09:16:13 <edd> J:by Hugo Haas.
09:16:14 <dc_www2003> Added comment J1.
09:16:32 <shab> edu - have a cms type thing 'media functions' inc access rights to help users manage resousces
09:18:03 <edd> J:An excellent presentation setting the scene for web services.
09:18:03 <dc_www2003> Added comment J2.
09:18:49 <edd> J:Touches on the relationship between [web services and REST|http://www.w3.org/2003/Talks/0521-hh-wsa/slide14-0.html].
09:18:49 <dc_www2003> Added comment J3.
09:19:51 <shab> edu - live demo of system
09:22:04 <shab> edu - showing navigation & features
09:22:50 <shab> edu - feels a bit odd using a 'rooms' metaphor but a file/tab interface...
09:24:06 <shab> edu- quite like firstclass... can annotate content
09:24:21 <shab> edu - gives url http://open-steam.org/index.html?lang=de
09:26:11 <shab> edu - questions. didn't hear question - answer that users can put things in their 'backpack' and then take them to another 'learnign room' and drop them - inc. their own room
09:28:26 <shab> edu - discussion. about culture of use/users of the system and user behaviour. how to encourage participation?
09:28:58 <edd> w3c-track - Yves Lafon talking about SOAP 1.2
09:29:49 <shab> edu - grading of students in competion or cooperative - point that if assessed individually then little motivation ot cooperate
09:30:07 <edd> slides aren't linked from w3c page.
09:30:10 <bijan> "as flexible and extensible as possible"
09:30:27 <hugoh> edd, they are if you reload
09:30:29 * bijan notes that flexible and extensible tend to fight with simple
09:30:36 <bijan> Pick any two
09:30:41 <hugoh> edd, http://www.w3.org/2003/Talks/0522-yl-soap/
09:30:48 <edd>http://www.w3.org/2003/Talks/0522-yl-soap/
09:30:48 <dc_www2003> K: http://www.w3.org/2003/Talks/0522-yl-soap/ from edd
09:30:55 <edd> K:|SOAP 1.2
09:30:55 <dc_www2003> Titled item K.
09:30:58 <edd> K:by Yves Lafon
09:30:59 <dc_www2003> Added comment K1.
09:31:35 <shab> edu - point from audience - assessment needs to be based on cooperation if this is to work - ben schneiderman
09:32:47 <bijan> SOAP == plenty of systems to toy with
09:32:54 <bijan> Toy systems? :)
09:33:05 <edd> now on SOAP/Web slide
09:33:18 <shab> edu - new presentaton Almut Rudolph - goal complexity ...
09:33:18 <edd> but not SOAP/Semantic Web :)
09:34:23 <DanC> BLURB: Social Meaning of RDF
09:34:23 <dc_www2003> L: Social Meaning of RDF from DanC
09:34:40 <shab> edu - correction presenter Angela Brunstein(?)
09:35:24 <DanC> L:perhaps get together to continue [http://www.w3.org/2001/sw/meetings/tech-200303/#ses-soc|formal meaning meets social context] discussion in Cambridge in March?
09:35:25 <dc_www2003> Added comment L1.
09:36:50 <shab> edu - about impact of goal (specific v unspecific) on navitation and knoweldge required
09:36:53 * DanC looks around for sandro and EricM
09:38:11 * darobin suggest going to http://www.mozillazine.org/weblogs/hyatt/archives/2003_05.html#003340 and voting :)
09:38:57 <dajobe> (ericm here... working with dajobe at moment)
09:39:23 <edd> darobin: would these be open source development?
09:39:55 <darobin> edd: they may, they're working on the KHTML codebase
09:40:09 <mattb>http://weblogs.cs.cornell.edu/AllThingsDistributed/archives/000058.html
09:40:09 <dc_www2003> M: http://weblogs.cs.cornell.edu/AllThingsDistributed/archives/000058.html from mattb
09:40:20 <darobin> I suggested in my comment that if they did SVG, they should really give it back to the community
09:40:28 <edd> *absolutely*
09:40:30 <mattb> M:|day 1 of www2003 blogged by Werner Vogels
09:40:30 <dc_www2003> Titled item M.
09:40:32 * edd dying for lack of SVG
09:40:37 <edd> it's a scandal
09:40:37 <darobin> :)
09:40:58 <darobin> don't hesitate to add your comment there, it could have an influence
09:41:01 <darobin> no kidding...
09:44:31 <shab> edu - ... major issues about the nature of specific v unspecific tasks for learners on their ability to actually apply their learning...
09:45:04 <DanC> L:hmm... I've been noodling on [http://meta2.stanford.edu/kif/Hypertext/node14.html|indexical entailment], using "the web" as the essential indexical, in the sense of Guha's thesis
09:45:04 <dc_www2003> Added comment L2.
10:15:49 * alexbrn slaps eamonn__ around a bit with a large trout
10:23:15 <edd> so wireless has gone down here :-)
10:23:28 <edd> connectivity starvation results.
10:23:44 * edd listening to Martin Dürst presenting on I18N and web services
10:24:09 <edd>http://www.w3.org/2003/Talks/0522-ws-i18n-mjd/
10:24:10 <dc_www2003> N: http://www.w3.org/2003/Talks/0522-ws-i18n-mjd/ from edd
10:24:24 <edd> N:|Internationalization and Web Services
10:24:24 <dc_www2003> Titled item N.
10:24:33 <edd> N:by Martin Dúrst
10:24:34 <dc_www2003> Added comment N1.
10:25:18 <edd> N:There are a lot of potential problems with web services and I18N. I personally doubt that people will bother to properly internationalize their web services in the vast majority of cases.
10:25:18 <dc_www2003> Added comment N2.
10:25:47 <edd> N:Maybe folk like Amazon will -- but they're the atypical use case. I don't even know how far large EAI systems go with I18N -- anyone any idea?
10:25:47 <dc_www2003> Added comment N3.
10:26:01 * edd praises the merits of bluetooth and gprs
10:26:49 <edd> N:by Martin Dürst
10:26:49 <dc_www2003> Added comment N4.
10:26:53 <edd> N1:by Martin Dürst
10:26:53 <dc_www2003> Replaced comment N1.
10:26:55 <edd> N4:""
10:26:55 <dc_www2003> Deleted comment N4.
10:28:56 <edd> Time for lunch. Hopefully they'll fix the wireless before we return for the plenary session.
10:57:38 * DanC waves from RDF Query testing BOF
10:58:37 <DanC>http://esw.w3.org/topic/RdfQueryTestingBudapestMeeting
10:58:37 <dc_www2003> O: http://esw.w3.org/topic/RdfQueryTestingBudapestMeeting from DanC
10:58:45 <DanC> O:|RdfQueryTestingBudapestMeeting
10:58:46 <dc_www2003> Titled item O.
10:59:00 <DanC> O:participants introducing themselves (EricP taking notes)
10:59:00 <dc_www2003> Added comment O1.
11:12:25 * DanC waves
11:47:34 <edd-lap_> edd-lap_ is now known as edd
11:50:04 <edd>http://www2003.org/speakers.htm#Shneiderman
11:50:05 <dc_www2003> P: http://www2003.org/speakers.htm#Shneiderman from edd
11:50:21 <edd> P:|Leonardo's Laptop: Human Needs and the New Computing Technologies
11:50:21 <dc_www2003> Titled item P.
11:50:35 <edd> P:Ben Shneiderman.
11:50:36 <dc_www2003> Added comment P1.
11:50:48 <edd> P:First afternoon plenary session today.
11:50:48 <dc_www2003> Added comment P2.
11:52:49 <edd> dc_www2003: view 1
11:52:49 <dc_www2003> P: Leonardo's Laptop: Human Needs and the New Computing Technologies (http://www2003.org/speakers.htm#Shneiderman)
11:53:01 <edd> P:[Ben Shneiderman's home page|http://www.cs.umd.edu/~ben/]
11:53:01 <dc_www2003> Added comment P3.
11:53:51 <edd> P:His book, [Leonardo's Laptop|http://mitpress.mit.edu/main/feature/leonardoslaptop/index.html].
11:53:52 <dc_www2003> Added comment P4.
11:54:26 * edd downloads sample chapter to read
11:56:20 <edd> P:*"The old computing was about what computers could do; the new computing is about what users can do."*
11:56:20 <dc_www2003> Added comment P5.
11:56:41 <edd> sage advice for this conference, possibly
11:58:57 <edd> about to start
12:01:41 <edd> steven pemberton chairing this session
12:01:51 <edd> introduces ben shneiderman
12:03:44 <edd> pemberton: big difference web made was made the internet usable for people
12:04:13 <edd> <shneiderman>
12:05:14 <edd> takes a scientific approach to improving hci
12:05:21 <edd> get past "user friendly" - want to be scientific
12:05:54 <edd> goal to understand. e.g. why are experts 20x faster... why are command line interfaces good for certain tasks?
12:08:09 <edd> describes work he did in the mid-80s to replace menu driven system with "embedded menus", controlled via mouse
12:08:27 <edd> did many empirical studies to figure how people would use it
12:08:55 <edd> timbl called "embedded menus" "hot links", cited scheiders work in his manifesto for web
12:11:16 <edd> with the web... transformation from invididual resource to global resource become important theme
12:11:35 <edd> leonardo quote about eye being best way for brain to understand world around you
12:12:59 <edd> showing smartmoney market visualization
12:13:04 * bijan wonders what the accessibility group things of this quote
12:13:13 <bijan> s/things/thinks/
12:13:24 <edd> at www.smartmoney.com/moneymap
12:13:59 <edd> ... making the point that visualization methods can help you see wide-ranging information sets more clearly
12:14:05 <darobin> (404)
12:14:25 <edd> s/moneymap/marketmap/
12:14:25 <edd> sorry
12:14:37 <mattb> (java applet)
12:14:39 <sandro> spotfire -- tool used by all gene and drgu researchers to view multidimensional data
12:15:16 <edd> (/me so tired of seeing stock data as the only ever example. can't we have a non-capitalist example? global warming statistics, maybe?)
12:15:33 <edd> visualization gives you answers to questions you didn't know you had
12:15:36 <bijan> Er...
12:15:40 <bijan> Weak.
12:16:10 * bijan has no idea what ben's doing
12:16:22 <tim-budapest> I didn't call them "hot". I just called them links.
12:16:22 <tim-budapest> I tried to use the Dexter model - link, anchor.
12:16:23 <bijan> Ok, get this bit with the 3 boxes
12:16:36 <edd> Demonstrating a nifty tool for finding patterns by drawing example.
12:16:42 <edd> Visual query by example.
12:17:04 * bijan waiting for all that scientific evidence backing up the claims for these tools
12:17:30 <edd> ... end of his preamble ... that's the past, this talk is about the future.
12:17:36 <edd> gives quote I already put on P:
12:17:48 <edd> P5:
12:17:49 <dc_www2003> (edd) *"The old computing was about what computers could do; the new computing is about what users can do."*
12:18:12 <shab> timesearcher at http://www.cs.umd.edu/hcil/timesearcher/
12:18:18 <bijan> P:Er..isn't this what, oh, Alan Kay et al at PARC were all about in the 1970s?
12:18:18 <dc_www2003> Added comment P6.
12:18:22 <edd>http://www.cs.umd.edu/hcil/timesearcher/
12:18:22 <dc_www2003> Q: http://www.cs.umd.edu/hcil/timesearcher/ from edd
12:18:25 <edd> Q:|Timesearcher
12:18:25 <dc_www2003> Titled item Q.
12:18:55 <edd> Q:visual data exploration tool demo'd by Schneiderman in his talk
12:18:55 <dc_www2003> Added comment Q1.
12:19:24 <DanC> P:"rapidly, incrementally, reversibly" <- principles of direct manipulation
12:19:24 <dc_www2003> Added comment P7.
12:19:40 <edd> we need "rae
12:19:49 <edd> we need "renaissance 2.0"
12:20:05 <bijan> A "rerenaissance"?
12:20:11 <bijan> We must all be born yet again?
12:21:38 <dajobe> O:[http://ilrt.org/discovery/chatlogs/rdfig/2003-05-22.html#T11-30-15|chat transcript] (partial)
12:21:38 <edd> (lots of description of da vinci's life and skills)
12:21:38 <dc_www2003> Added comment O2.
12:21:58 <edd> emphasis on the visual
12:22:36 <edd> Why Leonardo inspires us? He integrated scientific outlook, practical technology and artistic skill.
12:23:07 <edd> Goals for the New Computing
12:23:14 <edd> (1) More usable (reliable and comphrensible)
12:23:34 * make computer usage less frustrating
12:23:36 <sandro> "enuf" ending user frustration
12:23:41 <edd> shows crazy windows error dialog
12:23:53 * edd invites others to help scribe too, i can't keep up :)
12:24:11 <edd> table of frequent frustrations
12:24:26 <edd> top ones interesting. on Internet: timeouts/ long download/ page not found
12:24:41 <edd> apps: error messages, freezes, missing/hard to find features/crashes
12:24:55 <edd> OS: crasheds, wrong response, slow response, unexpected message boxes
12:24:55 <dc_www2003> Label OS not found.
12:25:08 <edd> tada - an unexpected message :)
12:25:37 <edd> result of this: 46% of time using computers was wasted
12:26:03 * JibberJim thought that was due to IRC not error messages...
12:26:21 <shab> draft of enuf paper at http://www.cs.umd.edu/Library/TRs/CS-TR-4371/CS-TR-4371.pdf
12:27:39 <edd> talks that we should make computers more usable, get the wasted time down
12:27:46 <edd> theme (2)
12:27:56 <edd> Universality: Digital Divide Remains Troublingg
12:28:09 * Jhendler waves at Bijan
12:28:18 <edd> shows graph of education against internet usage
12:29:07 <edd> makes some wild claims about that other gaps like men/women gone, and education/income only ones remainnig
12:29:21 <edd> makes some american political point (not very universal :)
12:29:37 <shab> ref at base of slide http://www.ntia.doc.gov/ntiahome/dn/
12:29:41 <bijan> many digitial divide gaps have disappeared or reduced: men/women; old/young
12:29:41 <bijan> But not rich/poor
12:30:02 <JibberJim> so it's US specific data...
12:30:05 <edd> Back to the point. * user diversity , technology variety, gaps in user knowledge
12:30:19 <edd> systems should be universal across these spectra
12:30:50 <edd> high technology advances bring systems to the widest possible number of users
12:31:54 <edd> third challenge: (3) systems should be Useful
12:32:06 <sandro> ben: low self-eficacy folks have the hardest problems with computers (that is, people who think they cant do it, cant do it)
12:32:09 * edd at a loss to put this slide in context
12:32:40 <edd> What we are missing is generative theories that will guide our invention of new technologies.
12:33:13 <edd> onto Theories of HUman NEeds
12:33:36 <edd> from Jefferson, Roosevelt, Maslow's hiearchy, Covey
12:34:05 <tim-budapest> Ben: We have tools to analyse the present but not to analyse the future.
12:34:27 <sandro> maslow's hierarchy of needs: physiolgical, safet, love, esteem, self-actualization. (echoed in COvey, Liginv, Loving, Learning, leaving a legacy)
12:34:53 <sandro> personal computer ==> social computer
12:35:17 <sandro> comptuer folks are the most introverted of professions
12:36:02 <Jhendler> I heard this somewhere before -- ahh yes
12:36:04 <Jhendler>http://www.w3.org/DesignIssues/Fractal.html
12:36:04 <dc_www2003> R: http://www.w3.org/DesignIssues/Fractal.html from Jhendler
12:36:24 <Jhendler> oops - didn't mean to chump - didn't realize we had that on -
12:36:35 <edd> too late. best give it a title :)
12:36:45 <Jhendler> R:| Fractal nature of the web
12:36:45 <dc_www2003> Titled item R.
12:36:58 <edd> explains levels of human relationships. observes how ebay capitalised on our need for trusted relationships.
12:37:02 <sandro> the next killer apps are : EMPATHY, trust, pricavy
12:37:13 <Jhendler> R: Shneiderman talk is similar to some ideas that Tim has re: communities on the web
12:37:13 <dc_www2003> Added comment R1.
12:37:50 <sandro> Activities: collect inforation, relation: communication, create: innovation, Donate: Dissemination
12:37:59 <edd> weblogs cited as example of creative urge
12:38:07 * edd never seen them as creative, more imitative
12:38:09 * DanC recovers from net outage...
12:38:12 * Jhendler thinks we need to show him group chump/ircs some time ...
12:38:15 * bijan sick of weblog references
12:38:21 <DanC> "unusable at any bandwidth" ala Nader's "unsafe at any speed"
12:38:22 <tim-budapest> c.f. "Interpersonal computing" - Steve Jobs, on the NeXT
12:38:31 <shab> There is a slideset with similar stuff inc. some of the graphs at http://www.cs.umd.edu/hcil/pubs/presentations/enuf%5B1%5D.ppt
12:38:43 <DanC> amen! raise expectations of users about the quality of software!
12:38:53 * DanC noodles on volunteering in my local community re use of internet
12:38:57 <DanC> anybody else give classes on how to use and understand the internet at their local library/church/school/etc?
12:39:03 <DanC> self/family+friends(2-50)/colleagues+neightbors(50-5000)/citizens+markets(5000+)
12:40:04 <bijan> DanC: nope. Often wanted to
12:40:34 <Jhendler> danC: yes - not even including my local university where i give lots of talks about it :->
12:40:35 <bijan> Done some ad hoc, one on one stuff
12:41:45 <tim-budapest> Thta's funny. Measured by the amound or time they spend on irc, they seem to do quite well ;-)
12:41:45 * tim-budapest as extroverts ...
12:42:08 <DanC> ooh! nifty photo thingy. (open source???)
12:42:45 <Jhendler> ahh hah, Brian McB asked me to prove I had a tie! Done.
12:42:59 <maxf>http://www.cs.umd.edu/hcil/photolib/
12:42:59 <dc_www2003> S: http://www.cs.umd.edu/hcil/photolib/ from maxf
12:43:11 <bijan> DanC: visual basic
12:43:20 * bijan was looking at it for region based markup
12:43:21 <maxf> R: Schneiderman's Photofinder
12:43:21 <dc_www2003> Added comment R2.
12:43:26 <maxf> oops.
12:43:33 <maxf> S: Schneiderman's Photofinder
12:43:34 <dc_www2003> Added comment S1.
12:43:42 <Jhendler> open source, wouldn't bet on it - guess what is one of the things Ben and I debate about :-<
12:43:51 <sandro> R2:
12:43:51 <dc_www2003> (maxf) Schneiderman's Photofinder
12:44:12 <edd>http://www.cs.umd.edu/hcil/newcomputing/
12:44:12 <dc_www2003> T: http://www.cs.umd.edu/hcil/newcomputing/ from edd
12:44:19 <edd> T:|The New Computing
12:44:19 <dc_www2003> Titled item T.
12:44:30 <edd> T:HCI Lab, University of Maryland
12:44:30 <sandro> World-Wide Med
12:44:31 <dc_www2003> Added comment T1.
12:44:51 <edd> R2:""
12:44:51 <dc_www2003> Deleted comment R2.
12:44:59 <maxf> it's downloadable
12:45:28 <maxf> but you have to register
12:45:37 * DanC starting course notes for presentation at local church/library at http://dm93.org/z2001/WebMasterTraining ; jimh, if you have any course notes, I'd like to add a link
12:45:44 <edd> recap of that third aim (3) Useful: In Harmony with Human Needs
12:46:20 <Jhendler> danC - licensing is NONCOMMERCIAL USE LICENSE AGREEMENT (http://www.cs.umd.edu/hcil/photolib/licensing.shtml)
12:46:42 <JibberJim> and source isn't available, seems not to do regions, just label x,y points.
12:47:00 <Jhendler> Microsoft only, no Mac
12:47:11 <bijan> It's in VB I'm pretty sure.
12:47:22 <tim-budapest> Ben has a *lot* of photos.
12:47:22 <tim-budapest> That you *had* a tie.
12:47:27 <JibberJim> yep VB6
12:48:11 <DanC> not open source: oh well. at least one can find a PC and run it and see the ideas.
12:48:22 <edd> --------ends-------
12:48:34 <edd>http://www2003.org/speakers.htm#Tamas
12:48:34 <dc_www2003> U: http://www2003.org/speakers.htm#Tamas from edd
12:48:48 <edd> U:|Globalities, Spatialities and IT Strategies: New Web Politics and Social Development
12:48:49 <dc_www2003> Titled item U.
12:49:06 <edd> U:Pál Tamás
12:49:06 <dc_www2003> Added comment U1.
12:49:12 <edd> U:second plenary presentation today
12:49:13 <dc_www2003> Added comment U2.
12:50:56 <maxf> edd's characters came out wrong here. How's this: Pál Tamás ?
12:50:59 * bijan has fairly high hopes for this talk
12:51:08 <edd> maxf: they come out OK on the chump: they're in UTF-8
12:51:12 * DanC sees accents from max
12:51:12 <bijan> At least he's addressing some key issues I care about
12:51:34 <edd> maxf: this new chump is my experimental UTF-8 aware version. so it understands me, even if you don't. :)
12:51:45 * maxf 's IRC not in UTF-8 :-(
12:52:12 <edd> maxf: that's ok, the chump falls back to latin-1 if it can't understand what you say as utf8
12:52:31 <sandro> I saw accents from both edd and maxf; max's looked better.
12:52:43 <edd> "looked better"!
12:52:49 <maxf> :)
12:53:15 <maxf> different fonts for different encodings, that's an interesting idea.
12:53:22 <sandro> sorry edd! your accents were all squigly. (that's the technical term.)
12:53:32 * tim-budapest tries r‰Lj
12:53:39 <sandro> eek
12:53:53 <DanC> anybody know if batik groks the SVG slides from the W3C slidemaker?
12:54:07 <DanC> or if batik is supported in/by debian?
12:54:10 <edd> (i found 日本語 to work just fine)
12:54:24 <edd> DanC: does batik have any interaction?
12:54:31 * DanC finds "libbatik-java - xml.apache.org SVG Library" debian package
12:54:34 * maxf sees r211Lj from Tim and æ227¥æ234¬èª236 from edd
12:54:44 <maxf> DanC: batik does Javascript
12:54:50 * DanC sees garbage from tim and edd
12:55:02 <JibberJim> batik has very good javascript support, but no SMIL anim.
12:55:02 <edd> DanC: plus ca change :)
12:55:18 <edd> Hmm. Could batik be wrapped as a Mozilla plugin?
12:56:18 <mattb> batik package in debian is lacking the gui viewer iirc
12:56:22 <mattb> it's just the libs
12:56:29 <maxf> Can plugins run on top of Java in mozilla?
12:56:30 <mattb> no squiggle
12:56:31 * dom just got caught by user frustation on his computer not warning it it was missing power...
12:56:35 <JibberJim> and presumably not the beta5 libs?
12:56:54 <mattb> 1.5beta2 currently
12:57:12 <maxf> batik/squiggle very easy to install w/o debian. Just a tar file.
12:57:23 <maxf> s/tar/jar/
12:57:27 <JibberJim> so get the beta5!
12:58:32 * edd wonders if anyone would care to scribe?
12:58:59 <Jhendler> arrgh - network here a pain ...
12:59:31 * DanC tries apt-get install libbatik-java ... sees kaffe installed as a consequence... hmm...
12:59:44 * DanC finds it more useful for folks to echo the points they find interesting/useful; the fact that nobody's writing down what the guy is saying is interesting information, to me, that would be lost if somebody was transcribing everything
13:00:24 * DanC hesitates to download 8.7MB from http://xml.apache.org/batik/dist/
13:00:54 * bijan can't follow the talk if he does *anything* else
13:00:55 <mattb> java package dependencies are always a bit tricky in debian
13:01:05 <mattb> java policy quite newly-minted
13:01:10 <bijan> Especially with the littl emoving arrows
13:01:27 <JibberJim> Can't you get it of someone else at the conference DanC ?
13:01:29 <Jhendler> ditto to Bijan - but drop the "anything else"
13:01:33 <edd> DanC: yeah. i guess that nobody's taken up the scribe means it's not really reaching the consitutency assembled here
13:01:54 <DanC> yup, one feels the impact of java's licensing status when relying on debian community for support
13:01:57 <dajobe> or, they can't type as fast as me ;)
13:02:04 <nmg_> nmg_ is now known as nmg
13:03:03 * dajobe gazes at baffling diagram
13:03:04 <edd> fans of steven pemberton should check out today's XML.com :)
13:03:41 <maxf> lol @ pemberton's photo!
13:03:59 <edd> maxf: I believe he chose that one.
13:04:53 <DanC> pemberton's photo?
13:05:10 <DanC> oh... xml.com...
13:05:56 <hugoh_> hugoh_ is now known as hugoh
13:08:40 <edd> moment of humor as speaker's laptop malfunctions
13:08:49 <edd> continues with unfortunate error dialog on screen
13:08:54 <edd> making the point of the previous speaker
13:11:46 <JibberJim> Why aren't they using a Show machine....
13:14:17 <sandro> "Show" machine?
13:14:23 * bijan wonders which was the *non* horrible table
13:15:28 <edd> interesting table of internet access costs by country
13:15:41 <edd> hungary near the most expensive (of OECD countries)
13:15:57 <edd> US the cheaprest. Turkey, Canda, Mexico, NZ, Aus next
13:16:09 <edd> UK near the middle
13:19:05 <darobin> XHTML 2.0 in Last Call this summer! cool :)
13:20:03 <maxf> plead for government-funded R&D
13:20:52 <libby_> libby_ is now known as libby
13:22:39 * Jhendler sometimes regrets presence of wireless in conf rooms, makes it to easy for audience not to listen...
13:22:55 * Jhendler thinks this talks is helping me change my mind.
13:23:29 <edd> Major message summary: "I was trying to show you the differences in the localities in the IT sector"
13:23:36 <mattb> final slide references http://www.colorbar.hu/ which is a 403
13:23:40 * edd wonders if slides available. some of the graphs looked interesting
13:24:06 <edd> There were some interesting, contradictory and nonintuitive stats in there, I thought
13:24:11 <dajobe> "made by colorbar - www.colorbar.hu"
13:24:50 <edd> ---ends---
13:24:59 <edd> time for questions
13:25:42 * DanC considers asking about open-source follow-on from software from schneider's lab
13:26:05 * DanC wonders where anybody's going next
13:26:23 <maxf> DanC, if they have a commercial licence, I doubt they will open
13:26:26 <edd> Questioner asks for a concrete response from speakers, then asks the most rambling and incomprehnsible question.
13:26:41 * edd sighs, doesn't envy session chair
13:28:11 <edd> will the "W3C-6 Semantic Web Update" talk contain anything of news?
13:28:53 <DanC> R:now spell-checked. nice fonts on linux make typos look so much more out of place ;-)
13:28:53 <dc_www2003> Added comment R2.
13:29:00 * edd will prolly go to that: emiller, bwm, jhendler, guha, rrs, danbri
13:29:38 <DanC> R:I showed this to somebody who picked this point up from timbl's talk and wanted to read more
13:29:38 <dc_www2003> Added comment R3.
13:30:23 * Jhendler wonders if edd would like to give my talk :->
13:30:33 * Jhendler would like to go to the Web Service Chor panel ...
13:30:47 <edd> Jhendler: no chance. i will just critique from afar :)
13:33:49 <Frans> Frans is now known as Jimbo
14:04:41 <shab> education - dist learning objects about to start
14:07:12 <shab> edu - erik duval... learnding objects are not 'objects' in the OOP meaning of the word
14:08:21 <shab> edu - general point that the most asked faq is what is a learning object? - basically no body knows...but
14:08:43 <tim-budapest> PN-10 = http://www2003.org/t_panels.htm#PN-10
14:10:08 <shab> edu - tagging materials is a problem for real people after initial enthusiasm - people jsut don't do it inc. erik himself at times
14:10:53 <shab> edu - ok so lets make the assumtion that people will not do it - 'electronic forms must die!' they stink, they are not good!
14:12:11 <shab> edu - sooo... can get some basic admin, lang, course context... etc and the actual powerpoint and html as basis for automatic tagging
14:12:43 <shab> edu - can then enhance this if required
14:13:17 * Jhendler thinks we can do a lot better than that, will discuss w/Ed after the panel...
14:13:25 <Jhendler> s/ed/edd
14:13:35 <shab> edu - challenge to all tools developers - make tools to make users lives more pleasent or at least not more miserable
14:14:27 <Jhendler> yes! Make tools that fit into their normal workflows, not create extra work!
14:14:45 <shab> edu - basically people don't get motivated by the promice of helping out future users e.g. their children's children's children - Erik ends
14:16:00 <shab> edu - new speaker missed introduction, talking about eduSource infrastructure
14:16:37 <Jimbo_> Jimbo_ is now known as Jimbo
14:17:18 <shab> edu - ... Tools called SPLASH, POOL and POND to assist with tagging in SCORM etc.
14:18:06 <shab> edu - more details at http://www.edusplash.net/
14:19:01 <shab> edu - basic schemas are too generic and community specicif stuff is necessary but doesn't map into the more generic standards... problem esp. accross communities
14:20:23 <shab> edu - learnign object repositries are natural places for people to 'connect' with eachother... finding eachother via peer to peer tools... can then to more persistant support to their activites?
14:21:17 <maxf> w3c - Brian shows Tower of Babel painting saying "I found a picture of W3C" ;)
14:21:40 <shab> edu - their environment is based on components that can be used as requried... speaker ends
14:21:49 <Jhendler> RDF points at little person begging at bottom of the tower :->
14:23:17 <shab> edu - next speaker Paul B? learning object is not just a file but can be a service
14:24:32 <shab> edu - stuff integrated via portals... add in a student model and the services can then use that model to adapt to user...
14:25:32 * bijan wonders what DAML-L is
14:25:42 <shab> edu - multiple adaptative services - need to be kept in sync... i.e. different teachers need to make sure that they are aware of what students know so not duplicate...
14:25:46 <bijan> I know there used to be DAML-L as DAML-Logic
14:25:47 <DanC> "consensus on datatype"; no, we reached a decision on datatypes *with recorded objections outstanding*
14:26:17 <DanC> ugh... in-your-face URLs.
14:26:40 <shab> edu - security issues over data - more advanced than user profiles
14:27:38 <sandro> DanC, why are you so opposed to in-your-face URLs? Transcribablility is the essential element of URIs according to RFC 2396, and that is surely true. THere are cases, like in human presentations, where URLs are necessary and appropriate.
14:27:48 <shab> edu - multiple portals on content will exit - tutor can choose portals that are based on particualr views of learning...
14:28:07 <sandro> of course a URI for the slide set is all that's needed or appropriate here.
14:28:12 <DanC> yes, but this wasn't one of those cases. http://www.w3.org/TR/ would have done nicely.
14:28:26 <shab> edu - but not only service adaptive but the portal too - seems very haystack like idea?
14:28:58 <sandro> okay, yep.
14:29:40 <shab> edu - speaker ends
14:29:41 <DanC> yes, a URI for the slide set is worthwhile.
14:30:11 <shab> edu - next speaker Wolfgang Nejdl
14:30:37 * DanC looks forward to the day when a URI for the conference itself could be relied on for finding all materials presented at a conference.
14:31:43 <libby> oh wolfgang. interesting stuff
14:31:45 <mattb> how far has the semweb update session got now?
14:31:50 <shab> edu - need better search engines... for learning objects esp. for conditional requiremtns of users
14:31:50 <libby> jim's up
14:32:32 <DanC> Jim: forgive me for wearing a hat; (1) lights are v. bright. (2) BenG just showed photos of me and I discovered I don't have hair. ;-)
14:32:40 <DanC> BenS
14:32:58 <shab> edu - can we build the 'adaptive web?' - RDF to represent - content, relationships, learner...
14:33:54 * DanC noodles on enhancements to the slidemaker tool ala "you've got 7zillioin words on that slide. try again with 5-7 bullets max. consider a figure"
14:34:08 <shab> edu - need to be interoperability... combine with SW inferencing (logic) and rules - adapt content, relationships and visualise them
14:35:19 <edd> DanC: my wish for slidemaker: an XSLT that makes it go into OpenOffice Presenter format. Slidermaker gives a *horrible* audience experrience.\
14:35:20 <dajobe> ah shab, the edutella people - interesting. Pity about the clash with the semweb w3c track. I hope they have a good audience too :)
14:35:43 <DanC> edd, have you seen the SVG output from slidemaker? I think it looks much better
14:35:46 <shab> edu - SW metadata - standardised discriptions... builds on Learning Materials Discriptions LOM/DC and learner disctiptsion IEEE/IMS
14:35:53 <sandro> edd, what makes it "horrible"?
14:36:11 <edd> DanC: i refer you to my woes getting SVG to render on my machine (powerpc linux)
14:36:42 <edd> sandro: umm, too small or too big text, needing to scroll, presenter fiddling to find the "next" button each time, distracting browser/operating system chrome...
14:37:16 <shab> edu - provide logic discriptions/programs/rules... exchange usign RDF-QEL and RuleML build on infernecing engines e.g. TRIPLE,Prolog, Minerva..
14:37:55 <DanC> the need to scroll problem isn't introduced by the slidemaker tool but by the slidemaker user
14:38:14 <shab> edu - e.g. using user models, content, learning goals etc... combined... but not easy to do
14:38:45 <DanC> there's a key for the next button, but I'm not sure what it is. I guess everybody knows how to go to the next slide in powerpoint; what's the key?
14:38:58 <edd> space bar, i believe.
14:39:00 <sandro> edd: that are all presenter problems not slidemaker problems. I know when I hit F11 opera goes full screen, without chrome. THe font is scalable - if it's too small, ask the presenter to increase it. (I;ve seen lots of ppt which was unreadable.). THere are keyboard keys to move fwd/backwards in browsers (like opera) which allow that.....
14:39:01 <shab> edu - issues - cope with hetrogeneity of WWW pages, what metadata do we actually need? what metadata can we get easily
14:39:14 * bijan notes that hugo haas gave MINDSWAP a nice shout out in the choerography session
14:39:56 <DanC> hmm... this slide could be a lot more clear about which things are real OWL and which things are almost OWL
14:39:59 <edd> not all browsers support those keycuts.
14:40:00 <edd> sandro: tools that allow many faults to be made are inadequate. many people who use slidemaker are not naturally gifted in laying out presentations. the tool should give maximum assistance.
14:40:00 <edd> sandro: i *could* give a stunning presentation with paper and pen
14:40:16 <shab> edu - issues cont... inferencing, what adaptive functionality can we express by rules and logic progrmas? how do we cope with necceary open world assumptions?
14:40:22 <edd> sandro: if you don't see why slidemaker looks bad to the audience, then i don't see any point in arguing with you
14:40:27 <DanC> but edd, how would batch conversion to openoffice format help?
14:40:28 <shab> edu - speaker ends - disucssion starts
14:40:55 <edd> DanC: well, personally i'd prefer it if they were written in openoffice and then converted the other way to html for the web.
14:41:41 <DanC> that makes more sense, though it introduces the issue of getting everybody to use openoffice.
14:42:05 <shab> edu - question Vincent ? (dublin) why did we start the 'learnign objects' thing in the start and has it done what it was meant to?
14:42:57 <edd> DanC: any kind of migration is slow i guess. i believe that OOo Presenter is an easier environment in which to write slides than an HTML editor, so that might be an appealing factor to the slide writer.
14:43:34 <dom> actually, I prefer the possibility of being able to scribe slides with a regular text editor
14:43:39 <shab> edu - erik - second part - no... but they have done lots of exciting stuff BUT often end up saying '...but can't shwo you jsut now buecause hte project has ended now...' - need to stop this projects leaving no trace...
14:43:52 <dom> s/scribe/scribble/
14:44:38 <edd> dom: it's a matter of perspective. I don't believe presentations should be "scribbled". I think they need consideration for the audience. They do take time to make, and the time taken to fire up a separate tool isn't necessarily a consideration.
14:44:40 <DanC> yes, dom, but what we're observing is that when presenters aren't helped/constrained by their authoring tools, the audience(s) suffer
14:45:23 * DanC wonders about OOo on mac; I've used it successfully on linux and windows
14:45:44 * tim-budapest uses it on Mac OS X X11
14:45:47 * bijan slinks guiltity: writes crappy slides no matter what; finds it hard to get into doing it
14:45:51 <dom> re scribbling, I was referring to the tools that allows to do it, rather than to the care you put in it
14:46:13 <bijan> But I make up by being dynamic and exciting!!!!!!!
14:46:58 <shab> edu - lots of people have been working on this stuff for 4 decades... will take more time... need a better understanding of hte problme space - chicken and egg, standards v understanding problem... standards there but still need to understadnt the problme better...
14:48:20 <DanC> re OWL DL: "there's a list of what those restrictions are" I don't think so. where is it?
14:48:40 <bijan> S&AS
14:48:47 <dajobe> yeah, I think so
14:48:48 <bijan> Um.
14:48:50 <mattb> i like using axpoint for slide authoring because it's completely constrained (<slide><point>...</point><point>...</point></slide>) and produces decent PDF output
14:48:55 <bijan> Not a busted out, clean, neat list
14:49:25 <bijan> Semantics in Web services getting lots of talk
14:49:31 <bijan> In the last three slides of each presentation :)
14:49:40 <shab> edu - peter - referes to project made in 1996 - Episodic Learner Model... 'elm' - good but an all or nothing thing... like mainstream VLEs... need to take content from system to system...
14:52:07 <shab> edu - new question Intellectual Prop issues?
14:52:50 <shab> edu - location services may be free but perhaps not content/LO copyrights... systmes need to deal wiht this.
14:53:17 <bijan> Steve Ross-Talbot winning the engaging talk contest
14:53:20 <bijan> humour++
14:53:29 <bijan> He deserves the extra "u"
14:54:33 * Jhendler wonders if his network is back up -- anyone see this?
14:54:38 <bijan> THere is another conference here in budapest...Wickedness.net
14:54:39 <edd> Jhendler: loud and clear :)
14:55:02 <shab> edu - erik - is it an issue? oh yes. but might also be used as an excuse too... the legal issues... peopel say shoudl be done people dont' do. e.g. just take stuff from a google seach result...
14:55:06 * Jhendler wonders who did "kick" so I couldn't read your comments :->
14:55:31 <bijan> Financial services accoutn for most e-commerce today
14:55:48 <bijan> Fixed income trading inteh US alone exceeds $32 trillion/year
14:55:50 <bijan> Wow
14:56:13 <bijan> Few Financial services members of the W3C
14:56:41 <tim-budapest> we should have a special membership fee for e'm ;-)
14:56:48 <shab> edu - the overhead (e.g. metadata) is consuming lots of resoruces pos. more than ever gets to the author... pershaps shoudl jsut drop it? look at creative commons initiative...
14:57:18 <edd> the financial services specs are at a level above that the W3C addresses, usually
14:57:47 <bijan> Choreography WG charter: Focus on process external descriptiong
14:57:55 <bijan> Build on WSDL 1.2, not 1.1
14:58:04 <bijan> Inputs to consdier, DAML-S
14:58:07 <bijan> What else is there?
14:58:14 <shab> edu - need brokers... to take care of hte transactions...
14:58:19 <bijan> (ok, WSCI, WSA, BPML, BPSS...)
14:58:29 <bijan> Must provide a sematnic mapping using RDF and OWL
14:58:42 <bijan> Deliverables
14:58:48 <bijan> (In a too light color to read well)
14:58:55 <bijan> Requirements document
14:59:02 <shab> edu - new question - 'the only tool that matters for LO is google' and 'the only bot that needs is the google bot' type comments...
14:59:05 <bijan> Usage scenarios, primer, tests
14:59:36 <bijan> One "of" more specificaitons of chore langage (I like this coinage, I get credit)
14:59:38 <DanC> hmm... it's more the log:uri doodad and log:implies than string: that provides aboutEachPrefix.
14:59:44 <shab> edu - if you think of Google as your competitor then you won't last... (erik)
14:59:46 * bijan wants to be on the chore language wg
15:00:04 <dom> both are involved in that, aren't they, Danc?
15:00:05 <bijan> <chore><wash dirtyThing="dishes"/></chore>
15:00:16 <bijan> Formal Models! Yay!
15:00:34 <tim-budapest> Fusion calculus
15:00:40 <bijan> Lots of trying to glob Robin Milner aura
15:00:42 <tim-budapest> <- developed from pi-calculus
15:00:48 * Jhendler Me too - oh wait, I am on chor :->
15:01:07 <tim-budapest> Lucien Wischik
15:01:12 <DanC> yes, dom, you need both string and log to do aboutEachPrefix; but string isn't the interesting/novel bit, if you ask me.
15:01:21 <dom> I see your point
15:01:22 * Jhendler s/chor/chore
15:01:24 <shab> edu - peter - can't find applets or interactive content, google can't deal with those... in general
15:01:42 * edd shakes head at SWAD "demonstrators"
15:01:45 * DanC is still trying to grok "pi calculus"
15:01:50 <edd> this has been more or less the same slide for 3 years
15:01:54 <DanC> shakes head?
15:02:00 <bijan> What is chore lang?
15:02:01 <edd> plenty more interesting work has been done, but doesn't make it to this slide.
15:02:07 <Jhendler> fusion calc - is that Steve Ross-Talbot speaking?
15:02:11 <edd> still annotea, algae, etc.
15:02:13 <bijan> A langauge for describign the external behavioru between two or more participatnts
15:02:15 <bijan> Yes
15:02:19 * tim-budapest wonders which sldie edd was lloking at
15:02:23 <DanC> edd, what seems like it belongs on that slide?
15:02:24 <bijan> Steve R-T seems to believe all this
15:02:25 <shab> edu - new question - if we don't knwo what LOs are, how can we be sharing and reusing them... do we have work on ontologies of LO or better defn?
15:02:33 <bijan> The chore vs. orche distinction
15:02:40 * Jhendler one of the SWAD ones - that's the title (includes annotea etc.)
15:02:45 <edd> Ralph's slide, 6 of 16. I'll get the URI later.
15:02:50 <bijan> So a Chore Desc Lang (CDL)
15:03:00 <bijan> Not need be executable or turing complete
15:03:01 <shab> edu - erik, not nec a problem that can't agree on a defn.
15:03:11 <dom> edd was refering to http://www.w3.org/2003/Talks/0522-sw-swad/slide6-0.html
15:03:21 <bijan> Again, font colore light grey on a grey background with a gray background image
15:03:47 <shab> edu - comment that can identify some charastrichs of LO (actually using an analogy about teachers...)
15:04:05 <tim-budapest> Designer colors, probably.
15:04:07 <bijan> can be used to enforce behavior
15:04:14 <bijan> can be recursively composed
15:04:27 <bijan> may work along side a chore programmign langauge
15:04:35 <bijan> Not all webser will have a CDL
15:04:40 <bijan> Dance jokes
15:04:42 <edd> Perhaps purely in the context of SWAD my sighs were unjustified, it's just that I don't see a lot new in 3 years. Annotea and algae pre-exist SWAD
15:05:10 <libby> danbri up...
15:05:33 <shab> edu - new question - need an overview of standards and how they fit together...
15:06:00 <DanC> I consider it a feature that SWAD maintains focus on a few things... N3 isn't new, but the N3 tutorial is new. i.e. it's gone from some scribbly python code to something we feel comfortable presenting to a tutorial audience.
15:06:45 <sandro> Ralph also left a bunch of stuff out.....
15:06:50 <DanC> like what?
15:07:07 <shab> edu - peter - :) yes we have a good stadnard http 1.1... they are evolving...
15:07:15 <sandro> good question. :-)
15:07:53 <Jhendler> DanC - think SWAD presentations would be better to list all, but talk about/show a few
15:08:03 <Jhendler> oops - I mean Sandro, not DanC on that last one.
15:08:21 <DanC> Danbri: "we don't do design by committee"... well, we have a distaste for it, but I think we still do quite a bit of it.
15:08:25 <shab> edu - there are standards... see eariler, but they do need to be extended to be useful
15:08:31 * bijan boggles
15:08:35 <bijan> danbri really said that?
15:08:39 <bijan> Maybe he meant swad-e
15:08:48 <Jhendler> bijan - in SWAD EU context :->
15:08:55 <sandro> what did danbri say?
15:08:59 <edd> DanC: Understood that focus important. Perhaps I'm a special audience: but I'm looking to see progress, relevance to user needs, demonstration of power. Perhaps relatively few people follow these things year by year.
15:09:19 <shab> edu - standards do help find stuff - but then prob need to follow up manually?
15:09:55 <shab> edu - new question - using reputation building for encouraging contributions?
15:09:57 <Jhendler> edd - I wish Ralph showed some of the W3C process stuff - very impressive, ask him to give you a demo later
15:10:16 <edd> I actually rather want to see SW succeed. I'd love to see the SWADdy stuff get into open source apps that I use daily.
15:10:26 <Jhendler> I'm not easily impressed by Sem Web stuff, but that stuff is really impressive in showing it
15:10:46 <shab> edu - erik at his university... citations go with LOs when used re-used...
15:10:53 * dajobe notes that shab has been reporting on the edu/education track interesting real semweb apps
15:11:06 <edd> e.g. I just got the developer of the new GNOME web browser, Epiphany, to have it export bookmarks in RSS 1.0 as a default (they have nonhieachical bookmarks)
15:11:21 <dom> edd, great news!
15:11:36 <shab> edu - soo.. boadly no not much...
15:11:45 <dom> I was speaking with Jose and Marja about their bookmarks work in Amaya looking a lot like the one happening in Epiphany
15:11:49 <Jhendler> ahh, I wondered who "edu" was
15:11:57 <libby>http://sw1.ilrt.org/discovery/2003/04/svggeo/
15:11:57 <dc_www2003> V: http://sw1.ilrt.org/discovery/2003/04/svggeo/ from libby
15:11:57 <dom> I hope we can find a way to have more contacts
15:12:20 <dom> (did you really mean RSS 1.0, BTW? or RDF?)
15:12:26 <shab> edu - hold on... look at the citeseer stuff - this does do that kind of recomendation and does up your citations...
15:12:49 <edd> dom: I meant RSS 1.0, which is as you know RDF.
15:12:50 <libby> V:|RDF queries for generating SVG for mapping locations of people and groups doing Semantic web stuff in Europe
15:12:51 <dc_www2003> Titled item V.
15:13:28 <shab> edu - question - what are the properties of a good learing object?
15:13:28 * Jhendler remembers Guha's resistance to URIs at the very beginning ... he tried to talk us out of using them in SHOE :->
15:13:31 <dom> I know, but I don't know how the RSS vocab fits in a bookmark framework...
15:13:41 <dom> but I may just need to have a closer look at it ;)
15:14:01 <shab> edu - 1) you learn from it... disagfeement - no need interpretation too... often by teacher
15:14:12 <edd> dom: Epiphany just has categories for bookmarks, no folder structure. So the categories are <dc:subject> properties on each item.
15:14:14 <sandro> Guha: Deep Semantic Web: intelligent agents
15:14:26 * DanC thinks he heard "without further ado" earlier.. maybe more than once ;-)
15:14:32 <sandro> ...: Shallow Semantics Web: data integration & search improvements
15:15:19 <sandro> ...: Process Integration & Data Integration
15:15:21 <shab> edu - question... do you deal with school level education... this all seems about Higher Ed
15:16:54 <amy> amy is now known as amyvdh
15:17:47 <DanC> phpht. no fair. cwm is designed to do boring stuff... it's intended to be the awk of the semantic web. We use it *exactly* to convert our contact databases from one tool to another.
15:17:52 <shab> edu - guy stands up and tries to jump question queue - chair hold line - yo chair.
15:18:26 <DanC> ^re guha's reference to cwm in contrast to prosaic "shallow semantic web" use cases
15:18:29 <shab> edu - question - open source...
15:19:10 <shab> edu - reply (not sure who from) - open source is not nec. dependable commerical products will be there and are supported
15:20:09 <sandro> While we could argue about how shallow/deep cwm is, I think Guha's dichotomy (spectrum) is important to acknowledge. I'm not sure his terms are great, though. Data Web vs. Intelligent Agents.
15:20:18 * Jhendler likes Guha's shallow v. deep, works well w/Tim's beach slide, but I think his shallow is a bit too shallow - we already have tools/techniques beyond the very bottom
15:20:39 <Jhendler> (oops - let's do that in log) Guha's shallow v. deep, works well w/Tim's beach slide, but I think his shallow is a bit too shallow - we already have tools/techniques beyond the very bottom
15:20:58 <Jhendler> Data Web v. Agent Web ?
15:21:27 <Jhendler> Not sure I like "data" web as many simple things can be done on things people don't think of as "data" per se (like Ben Shneiderman's photos)
15:21:42 <Jhendler> however, it is data in the way Tim used the term at keynote
15:22:35 <shab> edu - opensource content/systems... locked in to VLEs e.g. BlackBoard with content...
15:22:41 <dirkx> Does anyone know if the papers are on-line - Looking for "System Re-gegineering and Integration Using XML/Web Services [343]
15:22:41 <dirkx> Dr. Michael J. Hu "
15:24:00 <shab> edu - is a bad thing... open approaches fix this - left as open question...
15:24:36 <shab> edu - highschool/children level question...
15:25:47 <shab> edu - not that acutally nto HE level specs but from military... disagreement EU standard not US mil and had massive overlap with LOM from way back...
15:27:02 <shab> edu - Kieren Pitts - his project desingedd for a particualr age group, and always amazed that many other people use it... might therefore exclude users if tag with user data...
15:27:55 <DanC> "please help Eric miller" -- Guha's concluding slide
15:28:40 <shab> edu - from audience - point that HE has no set curriculum... where as schools (in UK) have a very tight curriculum and thus harder to map produce LO for schools becasue classroom contexts are restricitive...
15:29:27 <shab> edu - also teachers need training... donations fo computers no use on own, also need to know what to do with them...
15:30:25 * Jhendler come on Edd - ask us a hard one
15:30:47 <shab> edu - yes it is an important market...
15:33:36 <shab> edu - question, is it related to a curriculum... schools are returnign to 'old style' teaching - I think. Comparision done in his project with old style books and e-learning... not sure of the point/question
15:33:43 <shab> edu - session ends
15:35:43 <maxf> maxf is now known as max-out
15:40:23 <sandro> The program says this is "conference evening" Any clues what that means?
15:42:29 <dirkx> When was this dinner/boat trip planned ?
15:47:58 <DanC> re conference evening, see p46 of the glossy...
15:48:34 <DanC> thu, 22 May 17.30-22.30
15:48:48 <DanC> hungarian folk ensemble (warming up as I write)
15:49:08 <DanC> ... followed by a dinner on Boat "Europa"
15:49:34 <DanC> white conference badge required for admittance to boat
15:51:08 <DanC> several empty seats near me here..., and 3 empty power outlets! I'm in the back, 3rd-to-last row, in the middle.
16:11:09 * DanC_jam recovers from net wierdness... I think I lost my DHCP lease
16:11:32 <DanC_jam> from the back, I can see about 20 glowing LCD screens; not laptops, but digitcal cameras and camcorders.
16:15:26 <dirkx> Aye - is quite one of their better if not the best group
16:22:04 * darobin would like to recover from the net's weirdness too, some day
19:37:11 <uche>http://www.xml.com/pub/a/2003/05/21/xmleurope.html
19:37:11 <dc_www2003> W: http://www.xml.com/pub/a/2003/05/21/xmleurope.html from uche
19:37:29 <uche> D'oh! Wrong window
20:34:31 <sethl_2> sethl_2 is now known as sethl