Semantic Web Interest Group IRC Chat Logs for 2002-10-09

09:10:12 <danb_lap>http://www.ilrt.bristol.ac.uk/discovery/rdf-dev/rudolf/js-rdf/index.html

09:10:12 <dc_rdfig> A: http://www.ilrt.bristol.ac.uk/discovery/rdf-dev/rudolf/js-rdf/index.html from danb_lap

09:10:26 <danb_lap> A:|Now ancient RDF query / javascript prolog demo

09:10:27 <dc_rdfig> Titled item A.

09:10:59 <danb_lap> A:Nearby: [http://snowball.ilrt.bris.ac.uk/cgi-bin/rdf-dot/xrdf2tinyprolog?URI=http://www.w3.org/XML/2000/04soccer-sched/team-2.rdf|sample data] in triple-y format to paste into query form alongside rules.

09:11:00 <dc_rdfig> Added comment A1.

09:12:21 <Snakker> Snakker is now known as libby

09:13:09 <DanC> anybody taking notes in the calendaring thingy?

09:13:42 * DanC sees his question projected on the screen

09:14:04 <DanC> Libby invites participants to introduce themselves...

09:15:27 * DanC waves

09:15:53 <DanC> BLURB: Calendaring Workshop notes

09:15:54 <dc_rdfig> B: Calendaring Workshop notes from DanC

09:15:56 <DanC> logger, pointer?

09:15:56 <DanC> See http://ilrt.org/discovery/chatlogs/rdfig/2002-10-09#T09-15-56

09:16:07 <DanC> B:see [discussion|http://ilrt.org/discovery/chatlogs/rdfig/2002-10-09#T09-15-56]

09:16:07 <dc_rdfig> Added comment B1.

09:16:10 * danb_lap waves

09:16:22 <danb_lap> ABCDEFGHIJKLMNOP <- legible?

09:16:47 <mdean> nick mdean_

09:16:52 <mdean> mdean is now known as mdean_

09:17:24 <danb_lap>http://www.w3.org/2001/sw/Europe/events/200210-cal/

09:17:24 <dc_rdfig> C: http://www.w3.org/2001/sw/Europe/events/200210-cal/ from danb_lap

09:17:40 <danb_lap> C:|SWAD-Europe RDF/SW calendar workshop

09:17:40 <dc_rdfig> Titled item C.

09:17:53 <danb_lap> C:We're borrowing the #rdfig channel today...

09:17:53 <dc_rdfig> Added comment C1.

09:18:45 <shellac_> shellac_ is now known as shellac

09:19:17 <danb_lap>http://www.icalshare.com/

09:19:17 <dc_rdfig> D: http://www.icalshare.com/ from danb_lap

09:19:45 <danb_lap> D:|iCalshare "Share your iCalendars!"

09:19:45 <dc_rdfig> Titled item D.

09:19:53 <danb_lap> D:calendar sharing site from the Mac community

09:19:53 <dc_rdfig> Added comment D1.

09:21:23 <deltab> D: still using 'http'-like 'webcal' scheme

09:21:23 <dc_rdfig> Added comment D2.

09:22:09 <danb_lap>http://www.mozilla.org/projects/calendar/

09:22:09 <dc_rdfig> E: http://www.mozilla.org/projects/calendar/ from danb_lap

09:22:19 <danb_lap> E:|Mozilla Calendar project

09:22:19 <dc_rdfig> Titled item E.

09:22:48 <danb_lap> D:Example [webcal://ical.mac.com/jlquinn/CART32Schedule.ics|calendar] feed

09:22:48 <dc_rdfig> Added comment D3.

09:23:37 <danb_lap> D3:Example [calendar|webcal://ical.mac.com/jlquinn/CART32Schedule.ics] feed

09:23:37 <dc_rdfig> Replaced comment D3.

09:25:50 <danb_lap> Damian demos Mac OS X calendar tools.

09:26:04 <danb_lap> Creates a cal entry. Publishes it to the Web via WebDAV to mac.com.

09:26:10 <danb_lap> DanC: 'Zope does WebDAV too'

09:26:23 <DanC> mnot reported (somewhere?) that the protocol doesn't require anything outside HTTP 1.1 PUT

09:26:32 <gk> Thinks: does anyone have code that runs on Windows to dump Palm desktop software calendar as RDF?

09:26:38 <DanC> webcal: :-{

09:27:55 <danb_lap> gk: if you have Perl etc installed, DanC has some code. Maybe needs cygwin; not sure about the requirements of the Perl Palm library.

09:29:03 <danb_lap> Mike: 'how does it interact with the palm?'

09:29:12 <maxf> the Perl Palm library reads from PDB files. I don't think you need any requirement

09:29:20 <danb_lap> damian: ok i think. Uses sync-ml(?)

09:29:30 <danb_lap> ...can sync for eg with an ericson phone

09:29:46 <danb_lap> Mike: with Palm interacts with the Palm desktop

09:29:59 <danb_lap> damian: yup. Also can interact w/ Outlook via icalendar

09:30:26 <maxf> "Palm desktop does icalendar"?

09:30:51 <danb_lap> webcal://ical.mac.com/pldms/test.ics

09:31:08 <danb_lap> BLURB:Damian's test calendar file

09:31:09 <dc_rdfig> F: Damian's test calendar file from danb_lap

09:31:21 <danb_lap> F:See [test.ics|http://homepage.mac.com/pldms/.calendars/test.ics]

09:31:22 <dc_rdfig> Added comment F1.

09:33:23 <danb_lap> F:Or via 'webcal:' uri scheme, [test.ics|webcal://homepage.mac.com/pldms/.calendars/test.ics]

09:33:23 <dc_rdfig> Added comment F2.

09:36:19 <danb_lap> DanC 'Semantic Web Travel Tools'

09:36:38 <danb_lap> (slides have file:// urls, expect we'll see them here later...)

09:36:51 <libby>http://www.w3.org/People/Connolly/

09:36:52 <dc_rdfig> G: http://www.w3.org/People/Connolly/ from libby

09:36:54 <danb_lap> DanC: "I almost didn't make it here. Used the tools but screwed up anyway..."

09:37:06 <danb_lap> DanC: "RDF useful, you get merging for free across data formats

09:37:18 <danb_lap> ...shows diagram of some tools he'll demo

09:37:45 <libby> G:|Dan Connolly's homepage including travel picture

09:37:45 <dc_rdfig> Titled item G.

09:38:08 <danb_lap> "itin3 on homepage generated via xplanet, which takes its own config files generated via RDF tools

09:38:20 <libby>http://www.w3.org/2000/10/swap/pim/travel.html

09:38:20 <dc_rdfig> H: http://www.w3.org/2000/10/swap/pim/travel.html from libby

09:38:34 <danb_lap> ...using Cwm, usually does RDF in and out, but in this case outputs strings in Xplanet marker file format

09:38:37 <libby> H:| Dan Connolly's semantic web travel tools page

09:38:37 <dc_rdfig> Titled item H.

09:38:52 <danb_lap> ...we use .n3 'notation 3' syntax, which is another notation for RDF that has special support for rules

09:39:01 <libby> H:'The bane of my existence is doing things I know the computer could do for me'

09:39:01 <dc_rdfig> Added comment H1.

09:39:02 <danb_lap> ...so generate markers to overlay on travel map.

09:39:27 <danb_lap> ...the itinary comes in a structured email format to mailbox.

09:39:37 <danb_lap> ...I routinely misread these things!

09:39:43 <danb_lap> ...don't trust myself to decode this stuff.

09:39:52 <danb_lap> Mike: 'what we get is a PDF image of a fax!'

09:40:12 <danb_lap> DanC: I also get these in HTML, can scrape those. Use nasty perl script to generate an RDF view.

09:40:44 <danb_lap> ...run a standard check to generate a view of an itinary

09:40:52 <danb_lap> ...for final (flawed) human check

09:41:11 <danb_lap> ...The tool I *didn't* run checks that my return flight is after the end of the meeting.

09:41:27 <danb_lap> ...so Perl for import, a rules file for munging it.

09:42:33 <danb_lap> (runs Cwm rules engine, with inputs, and --think, fetching some data from web)

09:44:32 <danb_lap> DanC: I also have tools that take this and generates RDF version of iCalendar. And other tool from there to generate IETF iCalendar text format.

09:44:49 <danb_lap> ...notes: Evolution tools are case sensitive (but shouldn't be)

09:45:10 <danb_lap> ...iCal extensibility: things like x-lic-location: America/Chicago

09:45:18 <libby>http://swordfish.rdfweb.org/discovery/2002/10/calwk/calwk.jsp?pics=+calendar+workshop+images+

09:45:18 <dc_rdfig> I: http://swordfish.rdfweb.org/discovery/2002/10/calwk/calwk.jsp?pics=+calendar+workshop+images+ from libby

09:45:33 <libby> I:| people at the workshop

09:45:33 <dc_rdfig> Titled item I.

09:45:34 <danb_lap> ...todo: map such extensions into URI-named namespaces

09:45:48 <libby> I:| people at the SWAD-Europe calendar workshop

09:45:48 <dc_rdfig> Titled item I.

09:46:10 <danb_lap> ....also we're merging with RDF-dumped SQL stores

09:46:18 <danb_lap> ...Sync-ML would be holy grail of rdf apps

09:46:36 <danb_lap> ...eg. I had wrong itinerary (in my cell, palm, ...), so what to do in case of an update

09:46:51 <danb_lap> ...tricky to put changes back and have them propagate

09:48:07 <danb_lap>http://lxr.mozilla.org/mozilla/source/calendar/resources/content/converters/

09:48:08 <dc_rdfig> J: http://lxr.mozilla.org/mozilla/source/calendar/resources/content/converters/ from danb_lap

09:48:20 <danb_lap> J:|Mozilla Calendar: XSLT convertors

09:48:20 <dc_rdfig> Titled item J.

09:48:26 <DanC>http://www.w3.org/2000/10/swap/pim/ical2rdf.pl

09:48:26 <dc_rdfig> K: http://www.w3.org/2000/10/swap/pim/ical2rdf.pl from DanC

09:49:35 <libby> K:|DanC's perl script for ical->RDF and back

09:49:35 <dc_rdfig> Titled item K.

09:49:57 <libby> K:|DanC's perl script for ical->RDF

09:49:57 <dc_rdfig> Titled item K.

09:51:53 <libby> danbri: which schema to use - is this a problem?

09:52:31 <libby> DanC: don't expect to have one vocab for everything, e.g. palm etc - syntax translations and rule transformations

09:52:42 <libby> ...never been a problem

09:52:53 <libby> DanBri demoing

09:53:17 <libby> ...Mozilla stuff

09:56:01 <libby> Danbri shows http://www.icalshare.com

09:56:17 <libby> ...grass roots - lots of files up here

09:56:48 <libby> ...RDF is good for describing lots of different things - things like sports, movies, complex things

09:57:04 <libby>http://www.icalshare.com/

09:57:05 <dc_rdfig> L: http://www.icalshare.com/ from libby

09:57:25 <libby> L:|iCalShare - lots of calebndar feeds

09:57:25 <dc_rdfig> Titled item L.

09:57:30 <libby> L:|iCalShare - lots of calendar feeds

09:57:30 <dc_rdfig> Titled item L.

09:57:47 <libby> ...location-based feeds.

09:58:00 <libby> ...can import calendars into Mozilla

09:58:15 <libby> ...you can switch them on and off within mozilla

09:58:57 <libby> ...does it work with an ical mac file?

09:59:37 <libby> ...yes it does!

10:00:02 <libby> greg: how does it handle the xproperties+

10:00:03 <libby> ?

10:00:17 <libby> Danc: does it have an internal rdf model of icalendar?

10:00:32 <libby> danbri: no - its an external software

10:01:56 <libby> danC; not sure a schema will help reachh critical mass: data or UI better

10:03:02 <libby> danbri: people are enthused and want answers about what to use in RDF calendar

10:03:25 <libby> [discussion about timezone ids - non standard property is used in the mac file]

10:03:58 <libby> [mozilla file doesnt recognise the duration]

10:04:11 <libby> [mozilla file doesnt recognise the duration in damian's mac file]

10:05:21 <libby> ...location - ical? - location is text; there is a 'geo' option which can have long and lat

10:05:35 <libby> greg: no more will be doen on this in icalendar

10:05:48 <DanC> the ideal recipe for critical mass is UI+schema+data, of course.

10:05:58 <DanC> (multiple UIs, in fact)

10:05:58 <libby> danbri: gary frederick's xslt transformations

10:06:22 <libby> ...for mozilla: does moz->danc's icalendar format

10:08:00 <libby> danbri: jang's calendar demo using javascript prolog

10:09:15 <libby>http://www.w3.org/2000/10/swap/pim/ical.rdf

10:09:15 <dc_rdfig> M: http://www.w3.org/2000/10/swap/pim/ical.rdf from libby

10:09:31 <libby> M:| DanC's icalendar format

10:09:31 <dc_rdfig> Titled item M.

10:09:41 <libby> M:| DanC's icalendar schema for rdf

10:09:41 <dc_rdfig> Titled item M.

10:10:28 <libby> danC has semantic web tools that make him a better father! (football schdules for his son)

10:11:40 <libby> danbri talks about http://www.ilrt.bristol.ac.uk/discovery/rdf-dev/rudolf/js-rdf/index.html

10:12:00 <libby> ....it worked!

10:12:51 <libby> A: see also [http://www.w3.org/XML/2000/04soccer-sched/pre-k.html|ble valley soccer club schedule] - raw materials

10:12:51 <dc_rdfig> Added comment A2.

10:13:07 <libby> ....will rules the web getr's smarter and we get mmore answers

10:13:35 <libby> jos: we have been merging, munging rdf together - and getting the same answers

10:14:14 <libby> [discussion of rdf - assumes everythign has a uri?]

10:15:21 <libby> terry payne demoing retsima calendar agent

10:15:39 <libby> terry: came from a challenge at SWWS2001 - to get rdf into outlook

10:15:44 <mdean> nick mdean_

10:15:55 <mdean> mdean is now known as mdean_

10:16:21 <danb_lap>http://www.daml.ri.cmu.edu/Cal/

10:16:21 <dc_rdfig> N: http://www.daml.ri.cmu.edu/Cal/ from danb_lap

10:16:40 <libby> N: see also [http://www.daml.ri.cmu.edu/site/projects/RDFCalendar/|another retsina page]

10:16:40 <dc_rdfig> Added comment N1.

10:16:42 <danb_lap> N:|Retsina Semantic Web Calendar Agent homepage

10:16:42 <dc_rdfig> Titled item N.

10:16:55 * danb_lap looks for photo of libby and terry

10:18:03 <libby> here: http://swordfish.rdfweb.org/photos/2001/08/01/000586.JPG

10:19:01 <libby> N:+[Terry and libby at SWWS|http://swordfish.rdfweb.org/photos/2001/08/01/thumb-000586.JPG]

10:19:01 <dc_rdfig> Added comment N2.

10:19:56 <libby> terry: proper rdf parser, doesnt depend on ordering of the information. usus foaf to describe people

10:20:32 <libby> ....in the program, if finds lat/long, find via an agent e.g. restaurants locally

10:20:35 <libby> [neat!]

10:20:55 <libby> N:terry and libby at SWWS 2001

10:20:56 <dc_rdfig> Added comment N3.

10:21:16 <libby> terry....except! rthat the agent was a wrappewr around google, and google changed their pages!

10:21:18 <danb_lap>http://chefmoz.org/

10:21:18 <dc_rdfig> O: http://chefmoz.org/ from danb_lap

10:21:31 <danb_lap> O:|Open Directory's ChefMoz project

10:21:31 <dc_rdfig> Titled item O.

10:21:56 <libby> greg: webserice on xmethods which translates any US address to long/lat

10:22:11 <danb_lap> O:Useful fodder (sorry) for restaurant agents. Data available under liberal license [in rdf|http://chefmoz.org/rdf.html].

10:22:11 <dc_rdfig> Added comment O1.

10:22:29 <danb_lap> O:See also: [open directory in rdf|http://dmoz.org/rdf.html]

10:22:29 <dc_rdfig> Added comment O2.

10:22:30 <libby> danC can also do this for any city in the US and any in the world over 100,000, from UN data

10:22:45 <DanC>http://www.w3.org/2000/10/swap/pim/usPlace2LatLong

10:22:46 <dc_rdfig> P: http://www.w3.org/2000/10/swap/pim/usPlace2LatLong from DanC

10:23:00 <DanC> P:|U.S. Gazetteer rules: place names/zips to lat/long

10:23:00 <dc_rdfig> Titled item P.

10:23:25 <libby> terry: can also find contacts from the data

10:23:46 <JibberJim> btw you can also do it for the UK, but I can't find a site with nice licence...

10:24:16 <libby> terry shows importing rdf to outlook. can pick the ones you are interested in. will also store urls of the events (and soon - scan every 24 hours say for changes)

10:24:24 <libby> cheers jim

10:24:55 <libby> danbri: can you strip out corrupt sources if somethign goes wrong/ can;t get it out of palms say

10:25:15 <libby> terry thinks a good idea, but not yet done.

10:25:43 <libby> danbri: peopel are pestering for provenance in rdf tools

10:26:15 <libby> terry: more hooks for working within an agent architecture, e.g. alarms.

10:26:21 <DanC>http://www.w3.org/2000/04/mem-news/teamToGlobe

10:26:21 <dc_rdfig> Q: http://www.w3.org/2000/04/mem-news/teamToGlobe from DanC

10:26:52 <DanC> Q:|rules to plot W3C team members on a globe

10:26:52 <dc_rdfig> Titled item Q.

10:27:17 <DanC> Q:in progress; the interesting bit is the part that scrapes from [un lat/long stuff|http://www.un.org/Depts/unsd/demog/ctry.htm]

10:27:18 <dc_rdfig> Added comment Q1.

10:27:45 <libby> terry: mike dean had already published schedule for conference in a diff format last year: so - created a translation service between teh two.

10:27:48 <JibberJim> O: RDF appears to be invalid XML :-(

10:27:48 <dc_rdfig> Added comment O3.

10:28:06 <libby> ..and can do this on the fly

10:29:33 <libby> ...it's 'just' cgi script, but it works!

10:30:02 <libby> ...would lie to be able to automate discovery of translation services

10:31:31 <libby> danc:problems with 'hybrid' schema?

10:31:43 <libby> terry: yes: geo, alarms

10:31:53 <libby> ...no export at the moment in retsina

10:32:25 <libby> danC: extracted all the properties to RDF in the data he had created - not all of the icalnedr schema

10:33:41 <libby> danbri: have you looked at repeating events?

10:33:59 <libby> terry: we looked at it but we didnt implement it - a real pain

10:34:05 <libby> ...shoudl add it in

10:34:29 <libby> ...workign on it inthe next few months - using for teaching - class schedules from the web

10:35:18 <libby> danbri: it's not clear when you bail on general languages and do app-specific ones. can webont to do it.

11:04:54 <DanC> libby, I'd like to queue some question to the group... like whether to use 20001009 or 2000-10-09, and whether to pick a common namespace name for ical-in-RDF (and pick some ownership/maintenance policy)

11:05:22 <libby> ok. those are good questions :)

11:05:55 <libby> maybe we can discuss it at lunch, or perhaps we will ahve some time before lunch

11:06:09 <danb_lap> yes, good questions

11:06:31 <libby> Greg fritzpatrick talking about 'orlando' repeating events and icalendar

11:07:06 * DanC chuckles as Greg F switches into PC mode

11:07:21 <libby> greg: the calsch group is moving towards the close of it's work and is mostly doing protocol stuff

11:08:47 <libby> ..frank dawson is the leading light: the groups started with versit, then with microsoft to imc - doing vcard as well.

11:09:13 <danb_lap>http://www.w3.org/TR/vcard-rdf

11:09:13 <dc_rdfig> R: http://www.w3.org/TR/vcard-rdf from danb_lap

11:09:18 <DanC> I've thought, a couple times, about generazlizing my ical<->RDF stuff to mime directory; I wonder if that's feasible.

11:09:26 <danb_lap> R:|Renato Ianella's RDF representation of vCard

11:09:27 <dc_rdfig> Titled item R.

11:09:45 <DanC> vCal isn't totally dead, btw; my Nokia cellphone groks it

11:09:51 <libby>http://lists.w3.org/Archives/Public/www-rdf-calendar/2001May/0043.html

11:09:51 <dc_rdfig> S: http://lists.w3.org/Archives/Public/www-rdf-calendar/2001May/0043.html from libby

11:10:04 <danb_lap> heh, just got it too: http://lists.w3.org/Archives/Public/www-rdf-calendar/2001May/0043.html

11:10:14 <libby> S:|Greg talks about mime-directory and icalendar on RDF calendar list

11:10:14 <dc_rdfig> Titled item S.

11:10:17 <libby> heh

11:10:23 <libby> danC: mine too - and my palm

11:10:44 <libby> danbri: expectations of vcard and icalendra mixed together?

11:10:46 * DanC enjoys getting the background on iCalendar, Dawson, and all that

11:11:07 <libby> greg: ship them around, add them to each other, combine them etc

11:11:40 <libby> S: good background on the icalendar calsched working group

11:11:40 <dc_rdfig> Added comment S1.

11:11:42 <danb_lap>http://www.yell.com/

11:11:42 <dc_rdfig> T: http://www.yell.com/ from danb_lap

11:11:49 <danb_lap> T:|Yell directory

11:11:49 <dc_rdfig> Titled item T.

11:11:52 <DanC> "I call it 'state of the fart'" -- GregF

11:12:10 <danb_lap> T:Greg mentioned its treatment of opening times, and contextual assumptions eg of 'is open 5 days...'

11:12:10 <dc_rdfig> Added comment T1.

11:12:20 <libby> greg: e.g. yell.com assumes olots of background information about what open 5,6,7 doays means

11:13:17 <libby> ...some issues were never solved: lat/long or long/lat? alt(itude)

11:13:37 <libby> danbri were decisions documented?

11:14:09 * danb_lap looks for issues lists

11:14:10 <libby> greg no, only in new drafts, and 2445 and 2446 have never been updated. all docs in the rfcs. there is an issues list though

11:14:18 <danb_lap> Greg: issue lists are IMC.org site somewhere

11:14:23 <libby> ...no formal resolution

11:14:56 <libby> danbri: I think that rdf people are wary of reinventing the wheel

11:15:06 <libby> danC: hack first, then showit to people

11:15:47 <libby> greg: proposed a joint w3/ietf wg on calendaring - didnt happen

11:16:19 <libby> ...much stuff not used: 'is your dog a vegetarian'

11:16:44 <danb_lap> hack'n'show++ #what we're doing here...

11:16:59 <libby> ...diffeence between descriptions of things and the things themselves was an issue in the group

11:17:43 <danb_lap> greg: argued w.r.t. re-occurances vs recurrences

11:17:44 <DanC> re hack first: there's a tension between trying to solve all the world's problems, doing a totally general solution, on the one hand, and writing a 100 line perl script on the other. I'm recommending starting with the 100 line perl script. Or at least: don't spend so much time surveying the literature that you never get started.

11:18:04 * danb_lap completely agrees; standards can be paralysing

11:18:25 <libby> ....30th day of the particular calendar or in exactly 30 days: exact or falling on a particular calendar date

11:19:32 <libby> ...inclusive or exclusive 9-5 - go home at 5, or mon-fri - until end fri

11:19:52 <DanC> BLURB: iCalendar/RDF questions

11:19:52 <dc_rdfig> U: iCalendar/RDF questions from DanC

11:20:02 <libby> ...jst that can cause problems with repetitions etc

11:20:34 <DanC> U:ical uses 20021009, while XML Schema part 2 uses 2002-10-09. When we convert ical->RDF, which should we use?

11:20:34 <dc_rdfig> Added comment U1.

11:21:20 <DanC> MikeD: there's a daml time working group. James A. participating

11:22:12 <DanC> . http://www.daml.org/listarchive/daml-time/

11:22:16 <libby> greg: problems: iso 8601 is not public

11:22:16 <mdean_> DAML Time WG: http://www.daml.org/listarchive/dalml-time/

11:22:28 <libby> ...though soon they are doing recurrences

11:22:30 <mdean_> DanC typed it correctly

11:23:09 <DanC> note that the interesting bits of ISO8601 are re-capitulated with XML Schema part 2

11:24:16 <libby> greg: problems: comma-separation (000s or decimal separator); dash and hyphen distinuishing

11:24:35 <libby>http://skical.org/orlando/

11:24:36 <dc_rdfig> V: http://skical.org/orlando/ from libby

11:25:10 <libby> V:|Orlando A Logical Mnemonic Model for Dates and Times

11:25:10 <libby> - Greg Fitzpatrick

11:25:10 <dc_rdfig> Titled item V.

11:25:53 <libby> jos: cyc time model

11:26:00 <libby> patH: daml time is based on it

11:26:09 <libby> danbri: testing in icalendar?

11:26:20 <libby> greg: results are secret

11:26:30 <DanC>http://www.cyc.com/cycdoc/vocab/time-vocab.html

11:26:30 <dc_rdfig> W: http://www.cyc.com/cycdoc/vocab/time-vocab.html from DanC

11:26:47 <DanC> W:|Time and Dates, from the OpenCyc vocabulary

11:26:47 <dc_rdfig> Titled item W.

11:27:51 <danb_lap> ----

11:28:00 <danb_lap> Per talking about use cases, ski cal

11:28:14 <danb_lap> "example earlier w.r.t. movie theatres was enlightening re SkiCal

11:28:22 <danb_lap> consider a movie, ten movies a night, you could subscribe

11:28:36 <danb_lap> ...if you consider an entire city, eg tourist offices in many cities maintain event databases

11:28:42 <danb_lap> ...usually have 100s of events each night

11:28:53 <danb_lap> ...if you want that kind of calendar, your cal software fills up pretty quickly

11:29:04 <danb_lap> ....so need stronger structure to be able to filter, choose amongst these 100s of events.

11:29:20 <danb_lap> ...That's where skical comes in. iCalendar is good for office-oriented events liek this

11:29:26 <danb_lap> ...but not so for public events, eg concerts

11:29:35 <danb_lap> ...since you want to sort by category, price, who can attend...

11:29:42 <danb_lap> ...lots of interesting metadata when working with public events

11:29:50 <danb_lap> ...that's what we did w/ skical project

11:29:55 * DanC wonders why my stuff is labelled "Dan Connolly's foo" when everybody elses' stuff is just called 'foo'

11:30:04 <danb_lap>http://galway.metamatrix.se/skical/

11:30:04 <dc_rdfig> X: http://galway.metamatrix.se/skical/ from danb_lap

11:30:15 <danb_lap> X:Per's presentation

11:30:15 <dc_rdfig> Added comment X1.

11:30:48 <danb_lap> X:See those pages for detail; I won't scribe so verbosely. -- dan-scribe

11:30:48 <dc_rdfig> Added comment X2.

11:31:27 <danb_lap> per: many implementations don't try to implement recurring events

11:31:38 <danb_lap> danc: evolution and kde seem to manage ok

11:31:56 <danb_lap> per: public events don't have list of attendees, but often you want to capture notion of intended audience (eg. age etc)

11:32:04 <danb_lap> ...we tried to solve this in the skical spec

11:32:13 <danb_lap> ...we also have a minimalistic http-based sync specification

11:32:26 <danb_lap> ...a small thing. Similar to the approach Apple adopted

11:32:40 <danb_lap> ...we have a clear distinction between the information object and the event being described by the info object

11:32:48 <danb_lap> ...this distinction not so clear in iCalendar spec.

11:32:56 <danb_lap> ...We've worked on this since 1998. To what use?

11:33:15 <danb_lap> ...In Sweden, a number of newspapers, tourist offices etc that use SkiCal format, exchanging calendar databases

11:33:40 <DanC> U:how to represent iCal's rrule in RDF? a proposal/example: [ics in|http://www.w3.org/2000/10/swap/test/pim/testCal.ics], [rdf out|http://www.w3.org/2000/10/swap/test/pim/testCal.rdf]

11:33:40 <dc_rdfig> Added comment U2.

11:33:41 <danb_lap> ...number of uses not that large, not reached critical mass. Not reached international standards status yet, just drafts that need further work.

11:33:47 <JosD> W: some slight attempt to DanC's code to accomodate cn:temporallySubsumes usling rdf list arguments

11:33:47 <danb_lap> ...but have lots of deployment experience.

11:33:47 <dc_rdfig> Added comment W1.

11:34:08 <danb_lap> ...There are still lots of repetition in data creation, re-keying of data when should be able to sync

11:34:19 <JosD> W: ... in http://www.agfa.com/w3c/2000/10/swap/test/pathCross.n3

11:34:20 <dc_rdfig> Added comment W2.

11:34:27 <danb_lap> ...so big need for interchange formats. The tourism sector realises this, want this tech., ready to invest time and some money

11:34:42 <danb_lap> ...usually event producers are small orgs. Don't have a lot of resources to invest in computer systems.

11:34:49 <danb_lap> ...we have sold skical-based systems to some of these.

11:35:07 <danb_lap> ...Most organisers are very small orgs. For those, the Apple approach is better: just use personal PIM software to publish

11:35:16 <danb_lap> ...rather than maintain a (separate) database of events.

11:35:30 <danb_lap> ...Most tourist offices should be willing to exchange info, since their ultimate goal is to promote tourism.

11:35:43 <danb_lap> ...but sometimes it seems they forget this objective when can sell that info instead.

11:35:56 <danb_lap> ...so get some small $$$ but tourist info exchange suffers.

11:36:22 <danb_lap> ...Also tourist agencies seem happy with quite simple timing models. The 5,6,7 day open-ness approach might be good enough for next five years in tourist sector.

11:36:32 <danb_lap> ...since a real recurring event language hard to implement/design.

11:36:50 <danb_lap> ...So skical is based on the ical syntax, for compatibility, but easy to migrate the semantic work into another syntax

11:36:58 <danb_lap> ...A few weeks ago I tried this, using RSS+events

11:37:22 <danb_lap> ...entire events database of Stockholm is available (see www-rdf-calendar for urls) @@@urls please

11:38:06 <danb_lap> ...Am workign 1/2 time as a consultant on these issues. Other 1/2 is finishing thesis on this project., so am writing, doing analysis on this project., so want to hear questions etc on these issues

11:38:08 <danb_lap> questions

11:38:13 <danb_lap> danc: how do you get this data?

11:38:24 <danb_lap> per: many tourist offices have employees hired to type in the events

11:38:32 <danb_lap> danc: what UI do they use?

11:38:43 <danb_lap> per: various. lots use MS Access based. Some newspapers use MS Word.

11:39:09 <danb_lap> ...we've been building skical based systems that sit on top of the various SQL stores; sometimes thru local UI sometimes via Web.

11:39:27 <danb_lap> greg: considering Stockholm, lots of events, but folk try to sell events

11:39:39 <danb_lap> ...We preach: museums create their info directly and export it

11:40:03 <danb_lap> ...but stockholm info services like having a central role.

11:40:19 <DanC> events database of stockholm, msg to www-rdf-calendar: http://lists.w3.org/Archives/Public/www-rdf-calendar/2002Aug/0005.html Aug 2002

11:40:22 * maxf wonders: SKiCal, SKICal or SkiCal?

11:40:30 <danb_lap> SKIcaL :)

11:40:43 <danb_lap> greg: telling yellow pages that in a few years they'll give opening times

11:40:51 <danb_lap> ...service times, eg. when can you call your dentist

11:41:19 <danb_lap>http://www.zetlanddental.co.uk/images/home1.gif

11:41:19 <dc_rdfig> Y: http://www.zetlanddental.co.uk/images/home1.gif from danb_lap

11:41:29 <danb_lap> Y:How my dentist shares opening hours in the Web

11:41:29 <dc_rdfig> Added comment Y1.

11:42:02 <danb_lap> danc: re unstable marketplaces, centralisation can be routed around re google, web page indexing

11:42:21 <danb_lap> ...plenty roles for real estate agents, but many of their offerings becoming redundant

11:42:52 <danb_lap> ...so they're seeing need to add value (3d fly thrus, search by crime index etc)

11:43:02 <danb_lap> ...contrast w/ the music industry

11:43:28 <danb_lap> ...also see CDDB, who tried to go proprietary. So people create freedb instead.

11:43:33 <danb_lap> ...million monkeys effect

11:43:50 <danb_lap> ...interesting to see people in unstable marketplaces

11:43:58 <danb_lap> greg: next thing we did was job adverts

11:44:10 <danb_lap> ...people said 'hey thats a face to face app', but we can still do preselection

11:44:25 <danb_lap> ...in sweden, there was a monopoly on job info, so we're closer to exposing all that data

11:44:30 <danb_lap> ...huge interoperable labour thing

11:44:41 <danb_lap> ...in a way we're close

11:44:58 <danb_lap> DanC: in the usa, everyone with a resume mails it to everyone on the internet!

11:46:06 <DanC> BLURB:Pat Hayes summarizes survey of Knowledge Represenation formalizations of time

11:46:06 <dc_rdfig> Z: Pat Hayes summarizes survey of Knowledge Represenation formalizations of time from DanC

11:46:41 <DanC> Z:timePoint, timeInterval, timeDuration were key concepts

11:46:41 <dc_rdfig> Added comment Z1.

11:46:42 <danb_lap> Z:See [pat's research page|http://www.coginst.uwf.edu/~phayes/research.html] for links to postscript docs and overview.

11:46:42 <dc_rdfig> Added comment Z2.

11:47:06 <danb_lap> pat... "tried to identify areas of disagreement

11:47:16 <danb_lap> ...eg points at the end of intervals, are they included?

11:47:25 <danb_lap> ...at first point of light being on, is the light on? (etc)

11:47:36 <danb_lap> ...people had written papers on half-open intervals

11:47:47 <danb_lap> ...with other folks help, we got some of this cleaned up

11:47:54 <danb_lap> ...the answer is "you don't ask the question!"

11:48:19 <danb_lap> ...think of the endpoints not as being at the ends., instead "there are these things called intervals, the endpoints are these markers that start and stop"

11:48:27 <danb_lap> ...you don't even make it possible to ask the question

11:48:36 <danb_lap> ...being an interval means that theres an interval with endpoints that surround you

11:49:06 <danb_lap> ...i called this the 'glass continuum'; if you break a glass rod in half w/ the two ends, which half owns the break-point? ie silly question

11:49:24 <danb_lap> ...also (per greg's q), one month from now issue (@@detail)

11:49:39 <danb_lap> ...I used term 'calendar' for an interval that has a pattern of sub-intervals, which self-repeat.

11:49:43 <libby> X:|P&#228;r Lanner&#246;'s presentation

11:49:43 <danb_lap> ...eg months of the year

11:49:43 <dc_rdfig> Titled item X.

11:49:51 * danb_lap thanks libby

11:49:56 <danb_lap> ...leap year is tricky!

11:50:03 <libby> b***ger, doidnt work

11:50:06 <danb_lap> ....once you've got your pattern, this repeats. You have a hierachy of these

11:50:15 <danb_lap> ...with each, possibly sub-patterns, eg hours of the day

11:50:21 <DanC> cf http://www.cyc.com/cycdoc/vocab/time-vocab.html#CyclicalIntervalGroupType

11:50:37 <libby> X:|Par Lannero's presentation (sorry about lack of interenationalization Par!)

11:50:37 <dc_rdfig> Titled item X.

11:50:42 <danb_lap> ...eg. consider hours of sunlight which shrink and strech thru the year

11:50:57 <libby> X1:""

11:50:57 <dc_rdfig> Deleted comment X1.

11:51:05 <danb_lap> ...also in cultures that are largely devoted to eg cattle; cattle don't care about clocks! history of clocks fascinating...

11:51:11 <DanC> (note pat restores old clocks)

11:51:25 <danb_lap> ...With that one idea, you could do just about everything. Hierachy of repeating patterns which might all be same

11:51:49 <danb_lap> ...following the Cyc trick, now embodied in DAML, you identify times by hierarchies, eg the 5th second of the 3rd minute of the 19th hour of the whatever whatever

11:51:54 <danb_lap> ...stop at years, for eg.

11:52:08 <danb_lap> ...you think of that as a nested term; each stage selects within the enclosing

11:52:38 <DanC> e.g from cyc docs: (#$holdsIn (#$QuarterFn 2 (#$YearFn 1929)) (#$age AlbertEinstein (#$YearsDuration 50)), w

11:52:45 <danb_lap> ...hierarchy of temporal shapes; a very general purpose mechanism. Handles just about any naming convention. Perhaps over expressive for many apps.

11:53:02 <danb_lap> ...Other questisons: is time fundamentally discrete, continuous.

11:53:17 <danb_lap> ...for calendaring, can assume discrete

11:53:33 * danb_lap missed detail of point; @@scribe help?

11:53:42 <danb_lap> ...eg seconds since birth of unix etc

11:53:47 <libby> U: in icalendar for timezones you can use: a 'floating' date-time - with no timezone. (This means: this time in all timezones - and therefore different actual times); or a date-time with a timezone property, e.g. DTSTART;TZID=US-Eastern:19970714T133000 (I'm unclear if there is a uniquely specified list of TZs: guess there must be.) ; or a date-time with a UTC timezone: DTSTART:19980119T070000Z (no timezone property allowed)

11:53:48 <dc_rdfig> Added comment U3.

11:54:00 <libby> U:W3C's profile says that you can use

11:54:00 <libby> *floating (no timezone)

11:54:00 <dc_rdfig> Added comment U4.

11:54:00 <libby> *Z (UTC)

11:54:00 <libby> *+/- hh:mm (UTC offset)

11:54:10 <danb_lap> Greg: Michael Sperberg-McQueen in Orlando talked about transformations between different calendars

11:54:13 <danb_lap> ...Julian etc

11:54:39 <danb_lap> Pat: Consider clock in computer goes haywire, too fast, then resynchs from net

11:54:48 <libby> U:so these conflict

11:54:48 <dc_rdfig> Added comment U5.

11:54:50 <danb_lap> ...so you need to deal with another persons timeline being temporarily stretched

11:55:02 <maxf> libby, there isn't a uniquely specified list of timezones. You're supposed to define them using VTIMEZONE

11:55:07 <danb_lap> ...(hairy problem...)

11:55:14 <libby> ah, thanks maxf

11:55:44 <danb_lap> jang: problematicness depends on whether you're scheduling meetings vs cracking protons

11:56:05 <danb_lap> pat: another one re structure of the timeline. When planning ahead, thinking about the future, about things that may not happen

11:56:17 <danb_lap> ...time goes on; you go past the time when the thing that might've happened didn't

11:57:09 <danb_lap> ...some folk want the memory of the past to include all these things

11:57:15 <danb_lap> pat: many parallel timelines

11:57:29 <danb_lap> danbri: why semweb interesting; very abstract stuff <-> practicalities of exporting todo lists

11:57:34 <danb_lap> danc: timeline partially ordered

11:57:38 <danb_lap> pat: future is fluid

11:58:00 <danb_lap> danc: in Email threading you could measure 'physical' receipt time, but not when things were read etc

11:58:09 <danb_lap> pat: John Sowa's example...

11:58:22 <danb_lap> (@@see archive msg)

11:58:42 <danb_lap> ..."when communication is delayed, you have to apply something like Einsteinian relativistic thinking

11:58:47 <danb_lap> ...there isn't really a single 'now'

11:59:07 <danb_lap> ...depends what speed you're moving, eg. flying thru timezones

11:59:23 <danb_lap> ...the distance between our 'nows' changes, and from each perspective the others now changes shape

11:59:34 <danb_lap> ...people have built ontologies around this!

11:59:52 <danb_lap> danc: when you're scheduling meeting you need such reasoning!

12:00:05 <danb_lap> ..."he realises that she hasn't read his msg that..."

12:00:11 <danb_lap> (representations of representations...)

12:00:20 <danb_lap> pat: can schedule relative to getting the confirmation

12:00:42 <danb_lap> ...people like Sowa would like a temporal planner, that might say 'oh hell, he's on a plane so i need to wait until tommorrow night to schedule this'

12:00:55 <danb_lap> danc: or you might do things like 'if nobody objects within 24 hours...'

12:01:04 <danb_lap> pat: The alternate futures thing for me is central

12:01:15 <danb_lap> ...all AI needs to keep track of alternate possibilities

12:01:24 <danb_lap> ...leads into situation calculus

12:01:35 <danb_lap> ...or in Cyc, distinction between temporal things vs times

12:01:45 <danb_lap> ...stuff that lasts for a while versus times

12:02:01 <danb_lap> danc: that was very practical. there are just temporal things

12:02:22 * danb_lap missed jang's point; trying to think, talk and scribe :(

12:02:30 <danb_lap> pat: some event or process over a time interval

12:02:45 <danb_lap> ...does it mean it is happening over each subinterval? eg. driving west

12:03:19 <danb_lap> danbri: seems like an ambiguity in what we mean by 'driviing west'

12:03:32 <danb_lap> pat: yep, or others say 'there are gaps in your west driving where you're driving east'

12:05:06 <danb_lap> danbri: do people appeal to causality...?

12:05:14 <danb_lap> ADJOURN for lunch and possible laptop theft

12:05:16 <danb_lap> bbiab!

13:27:25 <AndyS>http://skical.org/orlando/

13:27:25 <dc_rdfig> AA: http://skical.org/orlando/ from AndyS

13:28:52 <AndyS> AA: Greg's Orlando paper

13:28:52 <dc_rdfig> Added comment AA1.

13:31:52 <JibberJim> Does anyone have any idea of a tool which will remove the duff characters in a 97MB RDF/XML file which stops XML processing of it?

13:32:12 <dajobe> oh not dmoz data again

13:32:24 <JibberJim> chefmoz actually!

13:32:28 <dajobe> same thing

13:32:32 <JibberJim> but presumably the same thing..

13:32:37 <danb_lap> ------ rdf-cal folk regroup after lunch --------

13:32:47 <JibberJim> do they still have laptops?

13:32:48 <dajobe> AndyS: AA:|Greg's Orlando paper ?

13:32:53 <danb_lap> demo + use case from Mike Dean, then use case discussion()

13:33:07 * danb_lap seems to have laptop still

13:34:10 <libby>http://www.daml.org/2001/10/agenda/

13:34:11 <dc_rdfig> AB: http://www.daml.org/2001/10/agenda/ from libby

13:34:17 <danb_lap>http://lists.w3.org/Archives/Public/www-archive/2002Sep/0091.html

13:34:17 <dc_rdfig> AC: http://lists.w3.org/Archives/Public/www-archive/2002Sep/0091.html from danb_lap

13:34:30 <libby> AB:|Mike Dean's daml agenda demo

13:34:30 <dc_rdfig> Titled item AB.

13:34:42 <danb_lap> AC:|Multi-namespace architecture (danbri, rss-dev discussions) -- use case fodder?

13:34:42 <dc_rdfig> Titled item AC.

13:34:54 <libby> Mike: real calculations with xml schema datatypes....

13:35:08 <libby> ...used at the dev day WWW2002

13:35:25 * danb_lap happy to see the xml schema datatypes work

13:35:44 <danb_lap> revisiting that part of the app in terms of recent RDF Core discussions/design would be useful.

13:35:50 <libby> mike: bootstrapping the sweb with lots of data

13:36:04 <libby>http://www.daml.org/data/

13:36:04 <dc_rdfig> AD: http://www.daml.org/data/ from libby

13:36:14 <libby> AD:|DAML data

13:36:14 <dc_rdfig> Titled item AD.

13:36:55 <libby> mike wishist: standard representations for timezones - this is a repository ofg ideas e.g. for student projects

13:37:13 <libby> ...public domain tz database

13:37:25 <libby> danc: also available in icalendar format

13:37:42 <libby> blurb: Mike's usecase

13:37:50 <libby> BLURB: Mike's usecase

13:37:51 <dc_rdfig> AE: Mike's usecase from libby

13:40:05 <libby> AE: see [slide 10| http://www.daml.org/2002/02/dagstuhl-rules/slide10-0.html]

13:40:06 <dc_rdfig> Added comment AE1.

13:40:21 <libby> danbri: what sands in teh way of implemnting the usecase now?

13:40:47 <libby> mike: info not on the web, e.g. openiong hours of hotel restaurant; also not what paid for

13:41:11 <libby> ...also what hotels have high-speed/wireless intenet access

13:41:23 <libby> patH: would this app be in codfe or in rules?

13:41:30 <libby> mike: start in code then rules.

13:41:55 <libby> danbri: obstacle seems to be lack of data

13:42:26 <libby> mike: hence the daml data page - try to go for things both authoritative and complete

13:42:42 <libby> danc - normally expensive

13:43:06 <libby> mike: awareness of other tools availble very useful

13:43:42 <libby> BLURB: danbri's usecase

13:43:43 <dc_rdfig> AF: danbri's usecase from libby

13:44:34 <libby> AF: [dentist's homepage | http://www.zetlanddental.co.uk/images/home1.gif] - a gif!

13:44:35 <dc_rdfig> Added comment AF1.

13:44:55 <libby> AF: would like to give them good advice: looks good, accessibility, semnatic web - everything

13:44:56 <dc_rdfig> Added comment AF2.

13:45:21 <libby> AF:terry - wouldnt it be cool to make appointments

13:45:22 <dc_rdfig> Added comment AF3.

13:45:55 <libby> BLURB:joachin's usecase

13:45:55 <dc_rdfig> AG: joachin's usecase from libby

13:45:56 * DanC isn't sure what a 'congress' is

13:46:07 <DanC> ah... conference.

13:46:12 <libby> AG: organnising a conference

13:46:12 <maxf> a conference in Spanish

13:46:12 <dc_rdfig> Added comment AG1.

13:46:35 <libby> AG: one point for distributiong information about it.

13:46:35 <dc_rdfig> Added comment AG2.

13:47:47 <danb_lap> "multi-conference tool... ISABEL application

13:48:12 <danb_lap> ...pages used for coordination

13:48:18 <danb_lap> ...static web pages, produced by hand

13:48:21 <danb_lap> (@@url?)

13:48:43 <danb_lap> ..."eg Global IPV6 test schedules... probs w/ timezones, since distributed

13:49:06 <danb_lap> ...also dependencies, need to coordinate by phone, www; expensive coordination overhead

13:49:19 <danb_lap> ...a nightmare!

13:49:19 <maxf> http://isabel.dit.upm.es/

13:49:19 <dc_rdfig> AH: http://isabel.dit.upm.es/ from maxf

13:49:49 <danb_lap> ...another case, result of an EC research project, 'Universal Brokerage Package for Learning'

13:49:57 <danb_lap> ...broker for interchange of learning materials

13:50:14 <danb_lap> ...goal: anyone can interchange such info; not just static but dynamic delivery of new courses

13:50:25 <maxf>http://www.ist-universal.org/

13:50:25 <dc_rdfig> AI: http://www.ist-universal.org/ from maxf

13:50:31 <danb_lap> ...can offer realtime delivery of new courses

13:50:50 <danb_lap> (live demo w/ login...)

13:51:12 <danb_lap> ...Here problem is that you have to make a distributed class w/ scattered participants

13:51:47 <danb_lap> ...and people have to go to this distributed classroom. Developed an RDF format to describe how to connect to the conference

13:52:01 <danb_lap> ...so that when student gets the schedule, so gets rdf file w/ connection details

13:52:27 <danb_lap> ...with iCalendar can't add such info; with RDF, its real important is that we can add additional information (in this case connection details) to the basic iCalendar info

13:54:00 <danb_lap> lib, wanna scribe?

13:54:03 <libby> BLURB: martin's usecase

13:54:04 <dc_rdfig> AJ: martin's usecase from libby

13:54:05 <libby> sure

13:54:07 <danb_lap> ta

13:54:28 <danb_lap> LTSN -- central clearing house for subject-specific teach'n'learn info

13:54:50 <libby> AJ: needs of a specific community: LTSN - 30-odd projets catalohuing events, conferences etc - all using different underlying technologies

13:54:51 <dc_rdfig> Added comment AJ1.

13:55:12 <libby> AJ: a really big event could be cataglogued 30 times.

13:55:12 <dc_rdfig> Added comment AJ2.

13:55:37 <danb_lap> AJ:See also [LTSN website|http://www.ltsn.ac.uk/]

13:55:37 <dc_rdfig> Added comment AJ3.

13:55:52 <libby> AJ:[generate events|http://chewbacca.ilrt.bris.ac.uk/events/] - for people who are moving from static to database-generated websites

13:55:52 <dc_rdfig> Added comment AJ4.

13:56:29 * danb_lap wonders how many people know what RSS is

13:57:03 <libby> AJ:the national clearing house for information could poll the individual LTSN centres for data. A step towards interoperability is using RSS+events module

13:57:04 <dc_rdfig> Added comment AJ5.

13:58:11 <libby> AJ:the spec for rss+events is vague, with lots of ill-defined text strings. However it's very simple spec and so very easy to implement - the 5 fields available suit our needs very well.

13:58:11 <dc_rdfig> Added comment AJ6.

13:58:31 <danb_lap>http://www.w3.org/2001/12/rubyrdf/squish/service/webfetch_tests.rb

13:58:31 <dc_rdfig> AK: http://www.w3.org/2001/12/rubyrdf/squish/service/webfetch_tests.rb from danb_lap

13:58:50 <danb_lap> AK:|A bit of Ruby code that shows LTSN event data cross-queried against Google backlinks

13:58:50 <dc_rdfig> Titled item AK.

13:58:57 <libby> AJ:several centres are becoming enthusastic about this technology and we want to be able to ID events from multiple sources.

13:58:58 <dc_rdfig> Added comment AJ7.

13:59:28 <danb_lap> AJ:So my LTSN-economics use case, 'notify me of economics events associated with Bloomfield House'.

13:59:28 <dc_rdfig> Added comment AJ8.

14:01:40 <libby> AJ:jan: lots of interest in UK portals for event calendars e.g. for students

14:01:40 <dc_rdfig> Added comment AJ9.

14:02:05 <libby> AJ:martin - stndards are very important because of the diversity of the technologies used

14:02:06 <dc_rdfig> Added comment AJ10.

14:02:31 <libby> AJ:**SIMPLE** spec - this is really important

14:02:31 <dc_rdfig> Added comment AJ11.

14:03:14 <libby> chrisc: you're nopt meant to repeat elements in rss -

14:04:01 <chrisc> the RSS 1.0 events module example has dc:subject repeated

14:04:39 <libby> martin: people are getting into RDF by this route

14:05:05 <libby> danc: yes - it's natural people want to be able to *do* something with it - or not interested

14:05:36 <libby> danbri: these things need to be simple but maybe a chunks are better than stripping it right down

14:05:47 <libby> ...or maybe just a nice writeup

14:06:04 <libby> damian: when it's that simple can you merge data?

14:07:51 <libby> martin: can aggregate but can;t identify events

14:08:46 <libby> danbri: phiilospy of identifying events patH?

14:08:56 <libby> [scribe missed it]

14:09:13 <libby> patH - things can be described in different ways - a discussion was an argument

14:09:40 <libby> jang: were they in the same room listenig to the same person

14:09:55 <libby> danbri:"the meeting whose homepage was http://ghsdjhfj/"

14:10:49 <danb_lap> danc: sw travel use case(s)

14:10:54 <libby> BLURB:danC's usecase

14:10:55 <dc_rdfig> AL: danC's usecase from libby

14:11:06 <danb_lap> travel itin visualisation; loading data into evolution; palm pilot import;

14:11:13 <libby> AL:visualizing travel information

14:11:13 <dc_rdfig> Added comment AL1.

14:11:39 <libby> AL: palm_>desktop and back. scraping schedules off webpages to RDF.

14:11:39 <dc_rdfig> Added comment AL2.

14:12:16 <libby> AL: ical from evolution to RDF, and scrape off the webpage, and see if missing off the personal calendar

14:12:17 <dc_rdfig> Added comment AL3.

14:12:47 <libby> chris: a lot of rss was initially scraped - could have cinema times for example

14:13:29 <libby> kal: making it as simple as possible for people to do teh bare minimum - so could find out where small touring theatre companies are playing locally

14:13:49 <libby> BLURB:kal's usecase

14:13:50 <dc_rdfig> AM: kal's usecase from libby

14:13:58 <libby> AM: making it as simple as possible for people to do teh bare minimum - so could find out where small touring theatre companies are playing locally

14:13:59 <dc_rdfig> Added comment AM1.

14:14:14 <libby> AM: make it easy for the little guy to publish

14:14:15 <dc_rdfig> Added comment AM2.

14:14:33 <libby> AM:danc - what about a form of html you culd publish in

14:14:34 <dc_rdfig> Added comment AM3.

14:14:46 <libby> AM:andy agrees - peopel don't read rdf

14:14:47 <dc_rdfig> Added comment AM4.

14:14:58 <libby> andy: not clear to me who the users are

14:15:18 <libby> BLURB: jan's usecase

14:15:18 <dc_rdfig> AN: jan's usecase from libby

14:15:47 <libby> AN: find two people in a house worth 200,000 pounds where they're both away for a week!

14:15:48 <dc_rdfig> Added comment AN1.

14:16:08 <libby> AN:trust: even knowing where you are is valuable e.g. for a CEO

14:16:09 <dc_rdfig> Added comment AN2.

14:16:43 <JibberJim> Jan's a Burglar?

14:16:57 <libby> greg: tried for filter oput what looked like was an event - scrpaing from web pages and then retuning it to them

14:17:05 <libby> heh, I don;t think so :)

14:17:14 <libby> he was making a serious point though

14:17:48 <JibberJim> ah so it wasn't his actual "use case" then?

14:17:49 <danb_lap> BLURB:Building on the world's investment in HTML (danbri use case)

14:17:49 <dc_rdfig> AO: Building on the world's investment in HTML (danbri use case) from danb_lap

14:18:09 <libby> AN: danc: security things is main focus of researtch project at W3C - policies of who gets to see the phone numbers of people going to a meeting etc

14:18:10 <dc_rdfig> Added comment AN3.

14:18:57 <libby> AO:danc created something danbri and eric vderl worked on - w3c main page witrgh a profile of xhtml - use xslt to scrape out a rss channel

14:18:57 <dc_rdfig> Added comment AO1.

14:19:41 <libby> AO: libby did a [events+rss version|http://chewbacca.ilrt.bris.ac.uk/events/events.xml]

14:19:42 <dc_rdfig> Added comment AO2.

14:20:18 <libby> AO: very minimal changes to html environment

14:20:18 <dc_rdfig> Added comment AO3.

14:20:29 <libby> a tutorial for making a web page web-friendly

14:20:43 <libby> AO:a tutorial for making a web page web-friendly, building on teh w3c one

14:20:43 <dc_rdfig> Added comment AO4.

14:21:13 <libby> mike: how well does this work with frony page and so on

14:21:28 <libby> danc: depends on if does class attributes - then ok

14:21:45 <libby> graham: adobe tols does work

14:22:01 <maxf> (adobe GoLive)

14:22:07 <libby> danbri: easier to teach these rules than icalendar syntax

14:24:04 <libby> greg: in orlando model can combine several calendars together - scraping e.g. countries' holidays which do over dfferent countries e.g. ramadan - whether open or not - very useful for traveling etc

14:24:15 <libby> ...but too many in the US; europe much fewer

14:24:16 <danb_lap> webcal://ical.mac.com/ical/Swedish32Holidays.ics

14:24:29 <danb_lap> BLURB:Swedish holiday calendar (from Apple)

14:24:29 <dc_rdfig> AP: Swedish holiday calendar (from Apple) from danb_lap

14:24:32 <DanC> aka http://ical.mac.com/ical/Swedish32Holidays.ics

14:24:48 <libby> AP:http://tizo.com

14:24:49 <dc_rdfig> Added comment AP1.

14:25:22 <danb_lap> AP: see [swedish32Holidays.ics|webcal://ical.mac.com/ical/Swedish32Holidays.ics] (webcal url, aka [Swedish32Holidays.ics via HTTP|http://ical.mac.com/ical/Swedish32Holidays.ics].

14:25:22 <dc_rdfig> Added comment AP2.

14:36:21 <mdean> mdean is now known as mdean_

14:36:28 <mdean_>http://www.cs.rochester.edu/~ferguson/daml/

14:36:29 <maxf>http://www.ercim.org/publication/Ercim_News/enw51/payne.html

14:36:48 <maxf> BLURB: Calendars, Schedules and the Semantic Web, by Terry R. Payne and Libby Miller

14:36:59 <mdean_>http://www.cs.rochester.edu/~ferguson/daml/

14:37:12 <danb_lap> dc_rdfig:view

14:37:12 <JibberJim> eek, you've killed the chump...

14:37:20 <danb_lap> it's exploded! too much data...

14:37:41 <dc_rdfig> AQ: http://www.cs.rochester.edu/~ferguson/daml/ from mdean_

14:37:42 <dc_rdfig> AR: http://www.ercim.org/publication/Ercim_News/enw51/payne.html from maxf

14:37:43 <dc_rdfig> AS: Calendars, Schedules and the Semantic Web, by Terry R. Payne and Libby Miller from maxf

14:37:44 <dc_rdfig> AT: http://www.cs.rochester.edu/~ferguson/daml/ from mdean_

14:37:45 <dc_rdfig> AP: Swedish holiday calendar (from Apple) (blurb)

14:37:46 <dc_rdfig> AQ: http://www.cs.rochester.edu/~ferguson/daml/

14:37:47 <dc_rdfig> AR: http://www.ercim.org/publication/Ercim_News/enw51/payne.html

14:37:48 <dc_rdfig> AS: Calendars, Schedules and the Semantic Web, by Terry R. Payne and Libby Miller (blurb)

14:37:49 <dc_rdfig> AT: http://www.cs.rochester.edu/~ferguson/daml/

14:39:26 <libby> is it dead?

14:44:43 <libby> ouch

14:45:13 <libby> U:I'd be interested in talking about modelling issues - e.g. person whose homepage is x

14:45:14 <dc_rdfig> Added comment U6.

14:49:40 <danb_lap> dc_rdfig:view

14:49:41 <dc_rdfig> AP: Swedish holiday calendar (from Apple) (blurb)

14:49:42 <dc_rdfig> AQ: http://www.cs.rochester.edu/~ferguson/daml/

14:49:43 <dc_rdfig> AR: http://www.ercim.org/publication/Ercim_News/enw51/payne.html

14:49:45 <dc_rdfig> AS: Calendars, Schedules and the Semantic Web, by Terry R. Payne and Libby Miller (blurb)

14:49:45 <dc_rdfig> AT: http://www.cs.rochester.edu/~ferguson/daml/

14:50:04 <mdean_>http://www.cs.rochester.edu/~ferguson/daml/

14:50:06 <dc_rdfig> AU: http://www.cs.rochester.edu/~ferguson/daml/ from mdean_

14:50:23 <mdean_> AU:|DAML-Time Homepage

14:50:28 <dc_rdfig> Titled item AU.

14:55:08 <danb_lap> danc: in the case where icalendar data looks just like '20021009' (no TZ qualifiers etc), who cares about the '-' ?

14:55:35 <danb_lap> jang: which way did xcal go?

14:55:48 <danb_lap> greg: went the no-hyphens way

14:56:13 <danb_lap> grek: prob ical specs won't go much further

14:56:37 <danb_lap> greg: there is a proposed std RFC which takes some of the earlier cal work, and now is hyphenated

14:56:40 <danb_lap> @@url

14:56:49 <gk>http://www.rfc-editor.org/rfc/rfc3339.txt

14:56:51 <dc_rdfig> AV: http://www.rfc-editor.org/rfc/rfc3339.txt from gk

14:57:24 <gk> AV| IETF RFC proposed standard profile of ISO 8601 for date/time formats

14:58:06 <gk> AV:|Date and Time on the Internet: Timestamps

14:58:08 <dc_rdfig> Titled item AV.

14:58:29 <danb_lap> danc: would need to detect which fields were dates

14:58:33 <gk> AV:IETF RFC proposed standard profile of ISO 8601 for date/time formats

14:58:34 <dc_rdfig> Added comment AV1.

14:58:38 <danb_lap> danc: i'm going to write my code with dashes

14:58:43 <danb_lap> danbri: me too

14:59:02 <danb_lap> danbri: does anyone want to speak into favour of not using dashes

14:59:06 <danb_lap> (silence)

14:59:14 <danb_lap> jang: (mutters about rdf datatyping)

14:59:30 <danb_lap> greg: something to be careful about; dashes versus hyphens

14:59:32 <gk> Me too ... that's what we agreed for the IETF standard

15:00:11 <gk> The IETF stuff doesn't distinguish between dash and hyphen; nor in ASCII

15:00:15 <danb_lap> moving on...

15:00:19 <danb_lap> Timezones

15:00:40 <danb_lap> Libby: in icalendar you can only have 3 different formats for datetimes

15:01:03 <danb_lap> ...one with no timezone; this means a floating datetime, ie. differs

15:01:11 <danb_lap> ...eg (from DanC) 'brush your teeth at 7am'

15:01:27 <danb_lap> greg: there is a cal component which defines a timezone

15:01:37 <danb_lap> danc: in icalendar syntax, if you omit timezone it means local

15:01:56 <danb_lap> libby: you can have a timezone property which you can 'make up yourself'

15:02:05 <danb_lap> greg: this discussion is documented (@@url)...

15:02:26 <danb_lap> danc: I can tell you what to put there if you want Evolution to read it (@@@details)

15:03:16 <danb_lap> (see testCal.ics url above)

15:03:36 <danb_lap> danc: a lot of tools use 'America/New_York'

15:03:50 <danb_lap> ...thats pretty widely deployed

15:04:41 * danb_lap misses detail; we look at DanC's testCal.rdf sample

15:05:46 <danb_lap> danc: smart thing to do is use both

15:06:00 <danb_lap> eg 'America/New_York' plus also declare it as icalendar spec requires

15:06:09 <danb_lap> ...ie same timezone database that unix/linux/etc all use

15:06:31 <JibberJim> "America/New_York" is problematical as it's a literal, therefore a pure RDF tool can't discover what the actual offset is...

15:07:31 <danb_lap> Libby: Issue is that you're allowed to use timezone offsets, eg. 1998-01-19 and -1 to show the offfset

15:07:39 <danb_lap> (sorry, I don't knwo the exact syntax being discussed)

15:08:29 * JibberJim does and has had problems with it... was going to use it in my airports thingy but ran into problems...

15:08:47 <danb_lap> greg: data point... when we did that rfc, we picked up some earlier work that went into cal stuff

15:09:13 <danb_lap> ...a great wadge of comments and complaints on those specs; which we didn't address, and just focussed solely on timestamps

15:11:47 * danb_lap looks for scribe help

15:13:54 <gk> Discussion about whether RDF-encoding of iCalendar, distinct from any other RDF schema for other timestamps:

15:15:01 <danb_lap> danbri: I don't like special string oriented syntax in rdf data formats, so suggest not using tz offsets

15:15:12 <gk> Should there be support for offset-from-UTC in the literal time value? iCalandar (apparently) does this differently, so for the strict purpose of encodign iCalendar, it's not strictly needed

15:15:18 <danb_lap> chris: Dublin Core recommend using W3C's datetime

15:15:42 <danb_lap> danc: I just want to talk about iCalendar

15:15:50 <danb_lap> ...I'm gonna change newlines into angle brackets

15:15:56 <danb_lap> ...I can easily insert hyphens etc

15:16:19 <danb_lap> ....Once I get into RDF, I'm doing symbolic processing; it's a pain to do string oriented processing in an RDF environment

15:16:40 <danb_lap> danbri: I ran into same thing in RSS+events context

15:18:40 <danb_lap>http://www.w3.org/2000/01/foo

15:18:41 <dc_rdfig> AW: http://www.w3.org/2000/01/foo from danb_lap

15:19:01 <danb_lap> AW:|A quick look at iCalendar, tim berners-lee, January 2000.

15:19:01 <dc_rdfig> Titled item AW.

15:20:49 <danb_lap> danc: (to libby) One problem with comparing the work I did, and you did... is that I don't know how to do it

15:21:06 <danb_lap> Libby: they (ietf) used capital letters, for eg.; so issues like that

15:21:41 <danb_lap> terry: I can make a list of the bits of the hybrid vocab which i'm making use of

15:21:59 <danb_lap> danc: I'm looking for things to regression test; know how to do that.

15:22:38 <danb_lap> danbri: how do we find out which bits of icalendar are being (successfully) used

15:22:40 <danb_lap> ?

15:22:54 <danb_lap> greg: bruce (?) runs a site, organising a view into mailing list archives

15:23:30 <danb_lap> Libby: I found a couple of sites that talk about this; and one interop site.

15:23:41 <danb_lap> greg: ...also a site by a bob (?) at MIT (@@detail)

15:24:23 <danb_lap> libby: calshed took down list of users (as there were too many, apparently)

15:24:31 <danb_lap> libby: my phone knows about Alarms

15:24:42 <danb_lap> danc: Nokia phones use vCalendar

15:24:55 <danb_lap> ...the precursor. PErhaps at this level syntactically indistinguishable

15:25:54 <gk>http://www.calsch.org/CalConnect3/interop.html

15:25:57 <dc_rdfig> AX: http://www.calsch.org/CalConnect3/interop.html from gk

15:26:18 <gk> ax:|IETF Calendaring and Scheduling Working Group

15:26:18 <gk> Interoperability Testing

15:27:11 <gk> AX:|IETF Calendaring and Scheduling Working Group Interoperability Testing

15:27:12 <dc_rdfig> Titled item AX.

15:28:24 <DanC> ... http://www.w3.org/2002/10/calenar

15:28:34 <DanC> ... http://www.w3.org/2002/10/calendar <-

15:29:25 <DanC> ... http://www.w3.org/2002/10/calendar#Vevent

15:29:35 <danb_lap> ie xmlns:cal="http://www.w3.org/2002/10/calendar#" etc

15:29:37 * danb_lap nods

15:30:13 <danb_lap> danc: proposal expectations for cchange: 'we announce all changes to www-rdf-calendar

15:30:34 <danb_lap> ...if anyone screams, within 7 days, we'll back out the changes (for further discussion)

15:31:21 <danb_lap> ...and status says 'If the CVS date below haven't changed in the last two months, active developments have likely ceased'

15:32:06 <danb_lap> What to put there?

15:32:25 <danb_lap> DanC's thing, based on data from icalendar tools. Things migrating from hybrid to this would scream...

15:32:49 <danb_lap> Terry: my analysis should give us a bunch of examples

15:33:03 <danb_lap> danc: I gurantee expected two weeks turnaround on any test data to rdf calendar

15:33:10 <danb_lap> ...am willing to keep an index of consumers of the data.

15:33:47 <danb_lap> (noting that can give Libby write accessto this namespace relatively easily)

15:34:01 <danb_lap> danc: nothing gets in here unless used

15:34:14 <danb_lap> ...that we can do via machine

15:34:28 <danb_lap> Libby: that'd be really useful (and hard to find from ical sites)

15:34:43 <danb_lap> ...would give us a bunch of examples

15:34:53 <danb_lap> danc: I could show the 4 properties I needed to get this into evolution

15:35:26 <libby> - hopeflly we can see from this different subsets that are used from different toold

15:35:37 <danb_lap> DanC: "you can always round-trip this to iCalendar"

15:35:43 <libby> by differen t tolls rather - and we can do this automatically

15:35:43 <danb_lap> danbri: including repeating events?

15:35:46 <danb_lap> danc: I have that already

15:36:30 <danb_lap> danbri: shall we check in hybrid.rdf too?

15:36:35 <danb_lap> danc: any value in giving it a new id

15:36:37 <danb_lap> ?

15:36:51 <danb_lap> danbri: <thinks> nah... ilrt.org uris should stick around (i own the namespace...)

15:36:57 <danb_lap> ---

15:37:04 <danb_lap> Discussion of Rrules (repeating events)

15:37:31 <libby>

15:37:54 <danb_lap> greg: 'every year on the last sunday of the 10th month'

15:38:07 * danb_lap expecting danc to paste rule and url we're discussing

15:38:42 <danb_lap> BLURB:Aside -- has anyone got a convertor for Crontab syntax...?

15:38:42 <libby>

15:38:44 <dc_rdfig> AY: Aside -- has anyone got a convertor for Crontab syntax...? from danb_lap

15:38:48 <libby> RRULE:FREQ=YEARLY;INTERVAL=1;BYDAY=-1SU;BYMONTH=10

15:38:48 <dc_rdfig> Label RRULE not found.

15:39:51 <danb_lap> danbri: (to PatH) can we write OWL rules over an RDF cal schema to represent such things?

15:40:02 <danb_lap> pat: "last" might be tricky

15:40:31 <danb_lap> @@scribe help?@@

15:40:40 <libby> [[

15:40:40 * JibberJim offers

15:40:40 <libby> - <rrule rdf:parseType="Resource">

15:40:42 <libby>   <bymonth>10</bymonth>

15:40:42 <libby>   <byday>-1SU</byday>

15:40:42 <libby>   <interval>1</interval>

15:40:42 <libby>   <freq>YEARLY</freq>

15:40:42 <libby>   </rrule>

15:40:45 <libby>   </standard>

15:40:46 <libby> ]]

15:41:04 <libby> danc has done a very straightfoward translation

15:41:39 <libby> ...no knowledge enginnersin, some microparsing

15:41:39 <danb_lap> danc: Python thing that goes back to iCalendar doesn't handle this yet

15:41:41 <libby> ...want to get it from ics, process it using rules and then put it back

15:42:13 <danb_lap> danc: I'm doing symbolic processing, so things like <freq>YEARLY</freq> is an enumeration

15:42:20 <danb_lap> ...elsewhere in the schema I've changed these into URIs

15:42:37 <danb_lap> danbri: why move to uris?

15:42:44 <danb_lap> danc: I got up on the URI side of the bed?

15:43:23 <libby> danbri: doesnt see how this relates to the rest of the vocabulary

15:43:54 <danb_lap> greg: one thing about ical that's odd... there's always an initial time (which has to have a date)

15:44:04 <libby> greg: there are four rulesset that can be applied in ical.

15:44:39 <libby> danc: you cant say every tues, just every tues from a particular tues

15:44:49 <libby> danbri: is it adequate for TV listings?

15:45:11 <libby> chrisc: lots of things - gigs, cinmea etc just have start times

15:45:24 <danb_lap> greg: always big discussion when things don't have end-times

15:45:34 <danb_lap> (lib, what did you want to talk about?)

15:45:41 <libby> greg: some calendar implementations have zero time things

15:45:51 <libby> connecting icalendar to oyther vocabs

15:48:12 <libby> danbri: writing up some casese of using icalendar with other vocabs

15:48:14 <chrisc>http://dublincore.org/documents/dcmi-period/

15:48:15 <dc_rdfig> AZ: http://dublincore.org/documents/dcmi-period/ from chrisc

15:48:43 <chrisc> AZ:|

15:48:43 <dc_rdfig> Titled item AZ.

15:48:45 <chrisc> DCMI Period Encoding Scheme

15:48:45 <libby> danbri: will writeup usecase for rss for mixing ns

15:48:48 <danb_lap> danc: not clear if a vEvent is an event or a description of an event

15:49:07 <danb_lap> ...issues like what is updated? presumably not the event

15:49:23 <libby> danc: would not endorse mixing, but would endorse converting

15:49:45 <danb_lap> ...mixing this with other vocabs could be tricky

15:49:59 <chrisc> AZ:DCMI Period encoding uses this syntax: name=Perth International Arts Festival, 2000; start=2000-01-26; end=2000-02-20;

15:49:59 <dc_rdfig> Added comment AZ1.

15:49:59 <libby> danbri: my mixing strategy - inventing properties which join them up

15:50:07 <danb_lap> danbri: I would use new predicates to relate this vocab to things from other apps, indirect to avoid confusion

15:50:12 <danb_lap> danc: found Cyc vocab v useful

15:50:22 <danb_lap> ...makes some subtle distinctions carefully

15:50:28 <danb_lap> path: yes, they thought that through well

15:50:52 <danb_lap> greg: in iCal almost always taling about descriptions of the event (not the event)

15:51:06 <danb_lap> danc: intuitively the organiser is the org of the event; not of the event description

15:51:16 <danb_lap> greg: in our world, often is, eg an infomediary

15:51:25 <danb_lap> danc: you really really do use it as the org of the record of the event?

15:51:31 <danb_lap> greg: yup, though we call it publisher

15:51:43 <danb_lap> ...though we use icalendar's name for it ('organizer') in SkiCal

15:51:54 <danb_lap> libby: is that the way it works re publish/subscribe?

15:51:59 * danb_lap glances at the clock

15:52:10 <libby> yep, 8 minutes to go

15:56:14 <danb_lap> ADJOURNED ****

16:11:36 <libby> bye all!

16:52:09 <dmilezZ> hrrm has anyone ever defined conversation in daml.. like motivations and such?

16:52:23 <dmilezZ> (enough to write a conversant thing)

16:52:33 <dmilezZ> dmilezZ is now known as dmiles

20:34:50 <danbri> dcrdfig:view

20:34:53 <danbri> dc_rdfig:view

20:34:53 <dc_rdfig> AV: Date and Time on the Internet: Timestamps (http://www.rfc-editor.org/rfc/rfc3339.txt)

20:34:54 <dc_rdfig> AW: A quick look at iCalendar, tim berners-lee, January 2000. (http://www.w3.org/2000/01/foo)

20:34:55 <dc_rdfig> AX: IETF Calendaring and Scheduling Working Group Interoperability Testing (http://www.calsch.org/CalConnect3/interop.html)

20:34:56 <dc_rdfig> AY: Aside -- has anyone got a convertor for Crontab syntax...? (blurb)

20:34:57 <dc_rdfig> AZ: http://dublincore.org/documents/dcmi-period/

21:21:06 <deltab> AZ:|DCMI Period Encoding Scheme

21:21:06 <dc_rdfig> Titled item AZ.

23:46:46 <deltab> AR:|Calendars, Schedules and the Semantic Web

23:46:46 <dc_rdfig> Titled item AR.