Semantic Web Interest Group IRC Chat Logs for 2004-07-16

08:13:04 Topic now Metadata in a multilingual world - http://www.w3.org/2001/sw/Europe/events/200407-cph (hijacking the general chat: (b)logs http://rdfig.xmlhack.com/ and http://www.ilrt.bris.ac.uk/discovery/chatlogs/rdfig/ )

08:13:04 Users on #rdfig: logger caz domCPH sh1m bengtfCPH stork mcmay_CPH libby r12a evamen danbri_dna helle mortenf chaalsCPH arnarl_ anselm gnu|WRK chris-p luuk arnarl jubin|away OliB burtonator kota eikeon jordan phenny golbeck evlist bjoern_ bskahan maxallen chrisc tav|offline dngor sbp` azz herflet eaon crschmidt em Talliesin vdv WSBot wh4 deltab ericP LotR kao DanC Wack GabeW MarkB Jibbler rreck jsled mattmcc danbri xover Cloud sh1mmer Mutiny sandro workbench

08:13:04 Users on #rdfig: Kerphi dc_rdfig CaptSolo hex_ deusx Grantbow shammah lion jdougan d8uv iwaim kasei

08:13:04 <ChanServ> [#rdfig] This channel is logged and blogged: http://logicerror.com/rdfIRCWelcome

08:16:34 <chaalsCPH> liddy: You can decide where you collect information from

08:16:47 <chaalsCPH> hosein: having translations voted on is a possibility

08:16:53 <mortenf> (ah, thanks dajobe?)

08:21:05 <chaalsCPH> liddy: establishing a set of rules about who you trust is interesting.

08:21:31 <chaalsCPH> r12a: it can take years to work out who you should trust. In some things, like airline instructions, it HAS to be right...

08:22:58 <chaalsCPH> Hosein: No translation is as good as the original? Do people agree?

08:23:15 <chaalsCPH> chaals: Actually no - in some cases the original is badly written

08:23:24 <evamen> I don't agree

08:23:42 <chaalsCPH> r12a: Working in translation often we had people who were better writers than the people who did the original...

08:23:52 <chaalsCPH> Hosein: But then is it a translation?

08:24:13 <chaalsCPH> r12a: It can be - translation isn't a straight matching of words.

08:24:46 <chaalsCPH> bengt: A literal and a semantic translation are likely to be different.

08:25:02 <Liddy> and if there is a culturally specific metaphor?

08:25:34 <chaalsCPH>http://www.johnhowardlies.com

08:25:35 <dc_rdfig> D: http://www.johnhowardlies.com from chaalsCPH

08:25:50 <chaalsCPH> D:|A site that may or may not describe the truth

08:25:50 <dc_rdfig> Titled item D.

08:28:09 <chaalsCPH> Leif: Is the translation a legal version?

08:29:25 <evamen> I am wondering what is better to be faithful to the original text or to make something useful...

08:29:31 <chaalsCPH> chaals: In the case of WCAG, the english "authorative version" is a recommendation by a group of engineers, who don't formally recognise any translation as official. But the german translation is actually used in law. Which raises important questions about what the focus of translation is - do you "improve the clarity?" or...

08:32:03 <chaalsCPH> r12a: the model of a single normative language is difficult, because people reading a different translation don't know what the original version says except by reading the translation.

08:32:59 <chaalsCPH> Leif: You have some organisation working on their text, and they pick a language. If you want a translation that is recognised, you find a body who can read the translation, and they recognise it.

08:33:38 <chaalsCPH> Hosein: And that means that people decide if they trust a local authority recognising a translation, as well as the original authors of the translation

08:36:20 <chaalsCPH> Hosein: So how do you know that there is a translation at all?

08:36:39 <chaalsCPH> Dom: By visibility - we would expect to get a lot of mail if we publish something badly

08:37:02 <chaalsCPH> Liddy: What I think we do in Dublin Core is recognise somebody that we think is an authority in a local language.

08:37:11 <chaalsCPH> ...we don't claim that the results are normative documents

08:37:46 <chaalsCPH> Helle: There is a question of trust. Anyone can do a translation. We had problems because the person doing it isn't trusted by the audience.

08:38:22 <chaalsCPH> ... which means that if I promote it people are going to react negatively. And that is beside the question of translation quality

08:38:37 <chaalsCPH> Liddy: Library catalogues have been doing this for a while

08:38:45 <chaalsCPH> Helle: classification is similar

08:40:16 <MattSEA_> MattSEA_ is now known as mcmay_CPH

09:01:45 * danbri_dna mails Tomcat folks asking for a .rdf mime association in MimeMap.java

09:01:57 <danbri_dna> seems we're already in Apache httpd :)

09:02:33 * sh1m is wondering if using golbeck's trust stuff is the best way to allow people to 'rate' resource assesments

09:02:57 <bengtfCPH> golbeck ?

09:03:10 <bengtfCPH> link ?

09:03:22 <sh1m> ' http://trust.mindswap.org/trustOnt.shtml

09:04:08 <sh1m> I am thinking an earl statement assessing an earl statement would be better but it needs some kind of schema to rate against

09:05:06 <sh1m> 'trust' about a person's creditibility on a subject could be implied by how people rate them but that seems to be a different thing.

09:05:10 <sh1m> Thoughts anyone?

09:06:04 <bengtfCPH> so the trust is domain oriented

09:07:43 <chaalsCPH> we're baaack..

09:08:28 <sh1m> I think the trust is implied rather than stated. So for example I might trust danbri to a level of 8 but that doesn't mean his accessibility assesments are any good. Also if he does an accessiblity assement if I give it a trust level that would be a 'guess' on my experience of danbri rather than a test

09:08:34 <chaalsCPH> Question: is xtm vs RDF an interesting question?

09:08:35 <sh1m> at least thats my current feeling

09:10:56 <chaalsCPH> Morten: The models aren't quite the same - for that matter french and english don't ahve quite the same models. But I don't think the question of techology language and natural language is parallel

09:11:12 <chaalsCPH> seems nobody knows enough about xtm to sustain a discussion here.

09:13:37 <danja> chaals, given that TMs and RDF can be (more or less) mapped on each other, is it significant?

09:13:55 <chaalsCPH> Eva Méndez talks about her experience running the Dublin Core mirror in Spanish

09:14:53 <chaalsCPH> danja, I don't know enough about TM's to answer and be very sure. But I don't think it is very significant in the long run. (the model differences might imply slightly different implementation paths...)

09:16:00 <chaalsCPH> Eva: I run the mirror of the Dublin Core in Spanish, and cannot keep up with the work. I translate things for that, and then I translate them again for AENOR (Spanish National Standards Organisation) and people think that the translations are better if they are the AENOR one.

09:16:41 <chaalsCPH> Liddy: Money makes a difference. Because there is money around ISO it seems to give it more authority

09:16:52 <chaalsCPH> r12a and ISO has a reputation that they want to maintain.

09:17:05 <danja> chaals, me neither, but lars' paper points that way : http://www.ontopia.net/topicmaps/materials/tmrdf.html

09:17:53 <danbri_dna> The TM / RDF relationship is awkward, because TMs address different parts of the RDF stack.

09:18:01 <evamen> I think "trust" in translations is a question of opinion or a social convention. People need to delegate the trust on an institution...

09:18:14 <danbri_dna> In some ways, SKOS + dc:subject + foaf:topic address the same space as TMs; in some ways, RDF itself does.

09:19:34 <chaalsCPH> Eva: when a Dublin Core specification gets to a stable state I try to translate it. This is important because there are a lot of spanish speakers, who need a translation to work on metadata

09:20:41 <chaalsCPH> Liddy: I don't think we cater well in DCMI for non-English speakers. Maybe it would be interesting to look at having a spanish Dublin Core working group - it seems that some of the different applications are based on things that are relevant to speakers of particular languages.

09:21:43 <chaalsCPH> Eva: It is important to be able to speak about work in my own language. It is easier to discuss, describe ideas, and explore, than in english.

09:22:05 <chaalsCPH> ... I am afraid that we can end up with language islands. We may federate too much...

09:24:08 <chaalsCPH> chaals: English speakers don't actually have perfect interoperability either... sometimes non-english speakers commmunicate better

09:24:11 <chaalsCPH> liddy

09:24:25 <chaalsCPH> liddy: and there are things that arise in particular language communities

09:24:36 <bengtfCPH> sh1m: trust level ? 1-10 ? shouldnt it be 0-10 to achieve medium trust=5, and what about ppl that you can trust to lie, should that be negative trust ?

09:24:37 <dom_laptop> dom_laptop is now known as domCPH

09:25:08 <sh1m> bengtfCPH, hey it aint my system ;)

09:25:25 <sh1m> bengtfCPH, they do explain about why no -ive numbers though

09:25:49 <chaalsCPH> Liddy: Tom is keen to find a way to solve some problems. How semantic web technology helps with the kind of task Tom has is an interesting problem.

09:26:10 <chaalsCPH> ... we are hoping to discuss this in the Dublin Core conference.

09:26:22 <domCPH> Tom's paper: http://www.w3.org/2001/sw/Europe/events/200407-cph/baker-pos.html

09:32:37 <bengtfCPH> to track changes of ontologies the KAON project came up with a evolve ontology that might be similar to what is discussed in the first section

09:34:08 <mortenf> uri?

09:34:43 <bengtfCPH> kaon.semanticweb.org

09:35:50 <chaalsCPH> Liddy: I think Tom is asking if people would do the same thing in the same way?

09:36:08 <chaalsCPH> Morten: I don't see anything controversial - is this a compatible approach with CCF / SKOS

09:38:36 <chaalsCPH>http://esw.w3.org/topic/SchemaChangeWithOwl

09:38:37 <dc_rdfig> E: http://esw.w3.org/topic/SchemaChangeWithOwl from chaalsCPH

09:39:04 <chaalsCPH> E:|Using OWL to manage vocabulary changes, without changing more than you need to...

09:39:04 <dc_rdfig> Titled item E.

09:40:18 <chaalsCPH> E: An idea that Alistair Miles and I discussed recently, about how to do versioning in a way that you could track the changes

09:40:18 <dc_rdfig> Added comment E1.

09:40:27 * danja appreciates chaals' cunning

09:50:58 <chaalsCPH> Bengt: There is a problem with how you use latest version identifiers. If you use version 1 of something, and point to the latest version, and version 2 comes out and changes some meaning, you have a problem

09:51:18 <chaalsCPH> chaals: When I use Dublin Core, I always use a particular version of each term. Is that normal?

09:51:52 <chaalsCPH> Hosein: We don't use versioned terms - normally term versioning is just maintenance

09:51:53 <bengtfCPH> mortenf: this is the schema for the kaon evolution log http://km.aifb.uni-karlsruhe.de/kaon2/ontos/evolution.kaon

09:52:15 <chaalsCPH> Liddy: We would use a versioned term normally, to make sure that we don't get caught if there is an incompatible change.

09:53:51 <chaalsCPH> E:The idea is that you can have machine-processable information that tells you what bits of your model (in a particular version) can be treated as if they belong in a different version of something (for example SKOS) without changing more than the minimal number of terms (so avoiding have millions of owl:sameAs statements as part of versioning a whole ontology).

09:53:51 <dc_rdfig> Added comment E2.

09:54:30 <chaalsCPH> Hosein: Oh, no we do point to a particular version of a term.

09:55:48 <chaalsCPH> Liddy: So do we think it's fair to say to Tom that there is nothing alarming about what he is doing, and that we suggest he consider the versioning idea?

09:56:15 <chaalsCPH> Morten: there is a small vocabulary used in FOAF to identify the status of a term - whether it is in testing, unstable, or stable...

09:57:36 <mortenf>http://www.w3.org/2003/06/sw-vocab-status/ns

09:57:36 <dc_rdfig> F: http://www.w3.org/2003/06/sw-vocab-status/ns from mortenf

09:57:51 <mortenf> F:|SemWeb Vocab Status ontology

09:57:51 <dc_rdfig> Titled item F.

10:01:02 <mortenf> F:used by FOAF

10:01:02 <dc_rdfig> Added comment F1.

10:07:08 <q8uv> q8uv is now known as d8uv

10:43:02 <q8uv> q8uv is now known as d8uv

10:49:25 <Emmy-Sidar> Emmy-Sidar is now known as Emmy

10:49:52 <Emmy> Good morning :)

11:40:13 * shellac is away: I'm busy

12:19:04 <chaalsCPH> we're back

12:25:00 <domCPH> [we're going to go through the list of issues we've identified during the workshop and see whether we can think of a solution, in terms of vocabularies, tools or projects]

12:25:31 <domCPH> 1st issue: addresses

12:25:54 <domCPH> Hossein: UN has a specification to describe addresses

12:26:28 <domCPH> ... it identifies the order of tokens in the various countries

12:26:34 <domCPH> ... probably a good start

12:26:49 <domCPH> Richard: OASIS has also something about addresses

12:27:01 <domCPH> ... but it was very complicated, and seems to have holes

12:27:42 <domCPH> Liddy: what about vcard, InetID (used with LDAP)?

12:28:07 <domCPH> Hossein: the Danish Governement made a core elements repository

12:28:15 <domCPH> ... with many schemas, one of the about addresses

12:28:22 <domCPH> ... see isb.oio.dk

12:28:35 <domCPH> Chaals: do they covert I18N addresses?

12:28:40 <domCPH> Hossein: only danish

12:28:45 <domCPH> ... but can serve as an inspiration

12:29:02 <domCPH> Chaals: all the tools I know use vcard

12:29:12 <domCPH> ... which may explain why they're broken

12:29:25 <domCPH> ... there is an RDF schema for vCard in a W3C Note

12:29:32 <domCPH> ... I don't know how good it is

12:29:51 <domCPH> ... What about phone numbers?

12:30:12 <domCPH> Dom: ITU has specified that, hasn't it?

12:30:21 <domCPH> Richard: it depends on what you do with phone #

12:30:38 <domCPH> ... if you want to distinguish area code from other parts

12:31:05 <domCPH> Chaals: plus many have countries have this rule where you drop the leading zero when dialing from outside the country

12:32:06 <domCPH> Liddy: other formats to describe people (may include addresses): viaf, intercard

12:32:33 <danbri_dna> the RDF schema for vcard is very basic, just labels for properties.

12:33:15 <domCPH> ... Inetorgperson (an IETF standard)

12:33:22 <domCPH> Chaals: what about measurements, numbers?

12:33:46 <domCPH> Liddy: It's not only differences in names

12:34:06 <domCPH> ... in some cultures, you measure things in a completely different way than in others

12:34:24 <domCPH> ... e.g. some islands in Pacific

12:34:37 <CloCkWeRX> do tell

12:34:38 <danbri_dna> BLURB:<foaf:Person foaf:nick="timbl" foaf:title="Sir"/>

12:34:39 <dc_rdfig> G: <foaf:Person foaf:nick="timbl" foaf:title="Sir"/> from danbri_dna

12:34:46 * CloCkWeRX is curious how else you'd measure something

12:34:55 <danbri_dna> G:see also: [http://www.w3.org/2004/07/timbl_knighted|press release]

12:34:56 <dc_rdfig> Added comment G1.

12:35:29 <danbri_dna> G:Congrats :)

12:35:29 <dc_rdfig> Added comment G2.

12:36:22 <domCPH> Chaals: depending on countries, you count differently; e.g. a Frenchman doesn't give the same answer of the # of days till thursday as I would

12:36:33 <domCPH> ... but I'm not sure there is anything to address about that

12:37:08 <domCPH> Hossein: for names, ebXML has something

12:37:18 <domCPH> ... @@@based on edifact (?)

12:37:30 <domCPH> Morten: I made a survey of FOAF last year

12:37:52 <domCPH> Chaals: FOAF looked at this name question at least 3 times

12:38:17 <domCPH> ... question of structure of a name

12:41:22 <domCPH> Chaals: currency?

12:41:35 <domCPH> ... there are unit, subunits

12:41:43 <domCPH> hossein: also unofficial units

12:42:37 <domCPH> [this is not only for currency, but all the things based on units (?)]

12:42:42 <r12a> Oasis Customer Information Management TC covers names and addresses: http://www.oasis-open.org/committees/ciq/ciq.html

12:42:54 <r12a> (last time i looked it wasn't perfect)

12:43:29 <domCPH> Chaals: also, currencies varies over time

12:43:44 <domCPH> ... the meaning of an australian dollar is different now from what it was 100 years ago

12:44:39 <domCPH> Richard: ISO has 3-letters code for currencies

12:46:52 <domCPH> [discussion about pronouncing numbers, currencies, etc, and relations to SSML]

12:47:13 <domCPH> Richard: phone numbers have also different rules for pronunciation

12:47:28 <domCPH> ... but there are many other rules about currencies, numbers, etc.

12:48:10 <domCPH> ... I'm not sure what our goal is here

12:48:20 <domCPH> ... how representation relates to the semantic aspect?

12:48:37 <domCPH> Chaals: we're looking to things that are useful in a semantic web world

12:49:43 <domCPH> ... trying to assess the difficulties on solving the various issues

12:49:49 <domCPH> ... we've identified

12:50:50 <domCPH> mortenf: but whatever the way you present the number

12:50:56 <domCPH> ... the semantics are the same

12:51:35 <domCPH> chaals: but when you want to present the concept in a way people can understand

12:51:43 <domCPH> ... you need to adapt it

12:53:25 <domCPH> ... what solutions are available to solve this type of issues?

12:53:33 <domCPH> mortenf: datatypes are relevant

12:54:01 <stork_> stork_ is now known as balbinus

12:54:38 <domCPH> chaals: times/dates?

12:54:59 <domCPH> ... iCal, RDF/iCal, XML Schema datatypes, dc:date, ISO date

12:55:12 <domCPH> ... eventSherpa

12:55:20 <mortenf> survey on names in foaf, from last year with lots of pointers: http://rdfweb.org/topic/NamesInFoaf

12:56:08 <domCPH> ... history, museum and astronomy people have always troubles with dates

12:57:05 <domCPH> richard: IBM has a book on the different calendaring systems

12:58:44 <domCPH> Chaals: about places?

12:58:56 <domCPH> ... country names, airport names and codes

12:58:59 <domCPH> ... GPS

12:59:11 <bengtfCPH>http://www.calendarzone.com/

12:59:11 <dc_rdfig> H: http://www.calendarzone.com/ from bengtfCPH

12:59:52 <domCPH> ... ISO country names and codes

13:00:10 <domCPH> Richard: I'm working on a UN document listing all the countries names in their original language and script

13:00:32 <Liddy> types of calendars - see http://www.hermetic.ch/cal_stud/lunarcal/types.htm

13:00:34 <domCPH> Iso for country codes: 3166

13:01:04 <domCPH> what about former names? former countries?

13:01:27 <domCPH> ISO has some of them, but doesn't go back e.g. 1800 years ago

13:01:41 <r12a> richard's document (still being developed and checked!) is http://people.w3.org/rishida/names/UNcountries

13:02:45 <domCPH> what about names of countries/places in the different languages?

13:03:08 <domCPH> chaals: other solutions include geo lat/long

13:04:22 <domCPH> liddy: there are also local versions of descriptions of the place they live in

13:04:46 <domCPH> ... e.g. in austrlia, there are legal documents that give names to the different regions of australia

13:06:12 <domCPH> other resources: TGN, OpenGIS, GML, Alexandria Digital Library, Geowanking(?)

13:06:15 <evamen> TGN: Getty Thesaurus for Geographical Names: http://www.getty.edu/research/conducting_research/vocabularies/tgn/

13:06:16 <dc_rdfig> Label TGN not found.

13:06:30 <evamen>http://www.getty.edu/research/conducting_research/vocabularies/tgn/

13:06:30 <dc_rdfig> I: http://www.getty.edu/research/conducting_research/vocabularies/tgn/ from evamen

13:07:03 <crschmidt> I:| Getty Thesaurus for Geographical Names

13:07:04 <dc_rdfig> Titled item I.

13:07:46 <evamen> Alexandria Digital Library (ADL) Gazetteer: http://www.alexandria.ucsb.edu/gazetteer/

13:09:14 <Pike> Pike is now known as Pike_

13:09:37 <domCPH> chaals: what about separators, colors, "text"?

13:09:48 <domCPH> ... this relates to the discussion about numbers

13:11:21 <domCPH> richard: many people have tried to come up with this cultural registry

13:11:42 <domCPH> ... with all these symbolisms

13:11:47 <domCPH> ... but that hasn't happened

13:11:57 <domCPH> chaals: so that's in the "very hard" category

13:12:31 <domCPH> ... for accessibility, we have EARL, AccMD, AccLIP, DC accessibility group

13:12:43 <domCPH> ... a few things at W3C ;)

13:14:00 <domCPH> ... for languages, we have SKOS, CCF, ISO 8639

13:15:06 <r12a>http://www.loc.gov/standards/iso639-2/langcodes.html

13:15:06 <dc_rdfig> J: http://www.loc.gov/standards/iso639-2/langcodes.html from r12a

13:15:15 <domCPH> ... Inkel LHE

13:16:19 <evamen> Inkel's schema: http://f14web.com.ar/inkel/rdf/schemas/lang/

13:16:28 <domCPH> richard: with regard to symbols, you have iso-690 (not all known)

13:17:10 <domCPH> chaals: what about grammar specs?

13:17:13 <domCPH> [break]

13:21:56 <r12a>http://www.iso.ch/iso/en/stdsdevelopment/tc/tclist/TechnicalCommitteeStandardsListPage.TechnicalCommitteeStandardsList?COMMID=4768&printable=true

13:21:57 <dc_rdfig> K: http://www.iso.ch/iso/en/stdsdevelopment/tc/tclist/TechnicalCommitteeStandardsListPage.TechnicalCommitteeStandardsList?COMMID=4768&printable=true from r12a

13:22:12 <r12a> ISO - International Organization for Standardization

13:22:12 <r12a> ... ISO/IEC 11581-1:2000, Information technology -- User system interfaces and symbols -- Icon symbols and functions -- Part 1: Icons -- General. ...

13:26:15 <ndw> Does anyone know what that 9 page spec actually *says*, nevermind what it might say that's worth 61$CH

13:27:54 <r12a> i'd like to know what it contains

13:28:31 <r12a> i believe it is a set of 'internationally standard' icons

13:28:36 <ndw> I wonder if the 9 pages includes the, what, three or four pages of boilerplate that's always in specs and the bibliography at the back :-)

13:29:22 <ndw> If I had an absurdly large amount of money, I'd just order the whole set of ISO specs.

13:29:34 <r12a> though i know that some are not instantly recognizable...

13:39:48 <bparsia_> bparsia_ is now known as bijan

13:44:52 <r12a> libby, read about pinyin here [see especially the section entitled Inputting Ideographic Text]: http://people.w3.org/rishida/scripts/tutorial/script-intro-1.pdf

13:47:27 <r12a> actually, specifically slide 37

13:47:47 <r12a> oops, s/libby/liddy/

13:48:48 <domCPH> [discussion about terms and concepts references]

13:49:09 <domCPH> Wordnet, though treasure, TEI, some other reference that Richard has to dig up

13:49:20 <domCPH> s/though/thought/

13:50:01 <domCPH> chaals: swad-eu is focused on short term goals, without breaking long-term ones

13:50:15 <domCPH> ... formalizing languages grammar seems definitely long term

13:50:47 <r12a> TBX http://www.lisa.org/tbx/ "TBX is an open XML-based standard format for terminological data. "

13:51:29 <domCPH> chaals: re. interfaces for metadata in a multilingual context

13:51:39 <domCPH> ... HTTP conneg is relevant

13:52:05 <r12a> also http://www.lisa.org/term/ - LISA's teminology SIG

13:52:42 <domCPH> ... FOAF read/write languages too

13:52:45 <domCPH> ... DC has stuff too

13:53:19 <chaalsCPH> a [tool with content-negotiation for entering data|http://www.sidar.org/hera]

13:57:38 <flaw> chaalsCPH: thanks for pointing me towards xhtml-mod to show that DTDs can do namespaces; Although, I'm curious as to why w3 chose to support schema as opposed to improving DTDs/doctypes/relate.. Do you or anyone else have links to discussions regarding this topic?

13:58:15 <domCPH> for catalogs (e.g. museum, products) I18N: UNSPSC

13:58:37 <evamen> United Nations Standard Products and Services Code: http://www.unspsc.org

13:58:38 <ndw> flaw, I think it mostly had to do with the fact that they're (DTDs) in an alternate syntax

13:58:59 <domCPH> liddy: what about different level of languages?

13:59:16 <domCPH> ... (level of complexity in a given language)

13:59:36 <domCPH> chaals: the FOGG index is a method to calculate the complexity of a text

13:59:50 <domCPH> ... it gives a reasonable order of magnitude

14:00:03 <domCPH> ... but it's arbitrary in many regards

14:00:17 <flaw> ndw: ok, so it was meant to be an "easier" form of DTD?

14:00:54 <domCPH> matt: it's an English-only method, with limitations

14:01:09 <domCPH> ... built by the US Navy to evaluate its manuals

14:01:15 <ndw> Uhm. It was meant to be a form that was expressed in XML. I'm not sure easier played into it. It sure didn't play into the result :-)

14:01:24 <domCPH> ... based on interpolations from experiences

14:01:35 <flaw> hehe

14:02:44 <domCPH> hossein: needs to distinguish classification vs identification [not sure about the context for this]

14:03:28 <domCPH> chaals: re complexity, SKOS may be relevant; there are also dictionaries and thesaurus

14:06:31 <domCPH> classifying objects: (ack they are context dependent)

14:06:48 <domCPH> ... no specific refs

14:06:59 <bengtfCPH_> bengtfCPH_ is now known as bengtCPH

14:07:36 <domCPH> describing people, adapting content to user, accessibility: CC/PP (mostly used for describing devices, but can also be used for user preferences)

14:08:49 <domCPH> ... AccLIP, V2 Universal Remote Console (VURC), FOAF

14:10:44 <domCPH> ... AccLIP is instantaneous and can describe several profiles of one person, FOAF is more permanent

14:11:27 <flaw> Seems odd, I asked in here a while ago about if there were any discussion/movements in developing a replacement for CSS to something XML based(maybe like RDF+XPath+some formatting goodness), and received a response along the lines: "I see no reason to throw away CSS syntax".. I guess DTDs aren't getting thrown away in favor of schemas, but I _assumed_(; that that was the direction the w3 was wanting to go.. thus my confusion/curiosity...

14:12:52 <ndw> I've been tempted several times to write my own XML-based interpretation of CSS and convert back to the text syntax with XSLT. But I've never found the time.

14:13:02 <flaw> personally, I'm starting to like DTDs, but I find them a bit inelegant..

14:13:43 <ndw> Someone's giving a talk at Extreme about extending them to do namespaces and datatypes.

14:13:50 <flaw> nice

14:14:10 <flaw> I'd be interested in hearing it

14:14:13 <ndw> It'd be pretty easy. I've also been tempted to write NS support in DTD validation on more than one occassion. But RELAX NG came along before I got around to it. It'd just be a matter of dealing with the prefixes.

14:14:16 <balbinus> balbinus is now known as stork

14:14:38 <ndw> Come to Montreal for Extreme! :-)

14:16:07 <flaw> hehe, I'm at least a couple thousand miles away and without funding for such a trip :)

14:16:38 <danbri_dna> I'd love to, but can't. Maybe next year(tm)

14:16:40 <domCPH> describing events, human rights violations: UN? standards, A.B.C. (lugosi et al)

14:16:55 <domCPH> describin law: tricky

14:17:10 <mortenf> H:|The Calendar Zone

14:17:11 <dc_rdfig> Titled item H.

14:17:25 <ndw> Well, I hear you. The proceedings goes online after the conference. The NS stuff might be in last years proceedings. They're doing datatypes this year, I think.

14:17:43 <domCPH> history: broad, tricky

14:17:46 <ndw> IIRC, they wanted to preserve DTD syntax so they used a PI where I would have used a <!NAMESPACE decl, but...

14:18:42 <mortenf> J:|ISO 639.2

14:18:42 <dc_rdfig> Titled item J.

14:18:53 <dajobe> I was thinking of going to Extreme, but it's a bit, er, markupy :)

14:19:22 <ndw> You mean there's something else?

14:20:38 <domCPH> multilingual documents: dc.language, another ref from richard about multilingual dossiers

14:20:45 <flaw> hrm, as far as the XML based CSS, I was thinking perhaps something a bit more general would be useful. So that formatting for other MLs could be specified and loaded by a user-agent, custom markup presentation?..

14:21:36 <ndw> Yes, more could certainly be done. I've just not had reason to be that ambitious.

14:23:50 <mortenf> K:ISO/IEC 11581-1:2000

14:23:51 <dc_rdfig> Added comment K1.

14:24:02 <mortenf> K:|ISO/IEC 11581-1:2000

14:24:03 <dc_rdfig> Titled item K.

14:24:18 <mortenf> K1:International Organization for Standardization, Information technology -- User system interfaces and symbols -- Icon symbols and functions -- Part 1: Icons -- General

14:24:18 <dc_rdfig> Replaced comment K1.

14:24:56 <flaw> hehe, come on, think about *just loading* a docbook in mozilla without any XSL transformations.. ;)

14:27:05 <chaalsCPH> r12a: there are cases where you want to describe a single document with multiple primary language

14:27:22 <chaalsCPH> the frontpage for the European Union 9and many of its agencies) is an example

14:28:15 <chaalsCPH> Bengt: People might be writing in one language, describing things in another, with a User Interface that is presented in a third. (Use case described)

14:29:01 <ndw> It's hard to do DocBook with just CSS; there's a few places where reordering is really useful.

14:31:40 <domCPH> richard: about languages, xerox has made a lot of works in in-line help

14:34:36 <r12a> reasoning about the different between declaring the language for processing a document and meta-data about the intended audience (ie. html lang=... vs. content-language): http://www.w3.org/International/questions/qa-http-and-lang

14:37:21 <evamen> AccCLIP: http://www.imsglobal.org/developers/ims/imsforum/messageview.cfm?catid=16&threadid=26

14:38:50 <Liddy> ummm for AccLIP - see http://www.imsproject.org/accessibility - there is also the matching AccMD . Together these are for user profiles and resource profiles. please send us comments etc...

14:41:57 <chaalsCPH> hosein: some generic ontologies are useful.

14:42:15 <chaalsCPH> I forgot Cyc, which allegedly encodes pretty much the entire universe

14:45:12 <evamen>http://www.dublincore.org/groups/languages/

14:45:12 <dc_rdfig> L: http://www.dublincore.org/groups/languages/ from evamen

14:49:10 <mortenf> L:|DCMI Localization and Internationalization Working Group

14:49:10 <dc_rdfig> Titled item L.

14:52:48 <stork> stork is now known as balbinus

14:53:44 <lyle> lyle is now known as lyle|away

16:06:34 <Talliesin> Are URIs in rdf:about and rdf:resource meant have chars above U+0080 escaped, or unescaped?

16:13:23 <OliB> take a look at http://www.isi.edu/in-notes/rfc2396.txt ;-)

16:14:59 <dajobe> can somebody change the topic if/when the CPH event is over? thanks

16:15:04 <OliB> > 2.4 Escape Sequences

16:15:32 <OliB> > Data must be escaped if it does not have a representation using an unreserved character; this includes data that does not correspond to a printable character of the US-ASCII coded character set [...]

16:33:03 <`flaw> `flaw is now known as flaw

17:13:11 <libby> any cwm-knowledgeable people about?

17:13:45 <libby> I'm getting an obsolete maths built-in error from an old n3 file

17:30:33 <mortenf> mortenf has changed the topic to: Semantic Web Chat - (b)logs http://rdfig.xmlhack.com/ and http://www.ilrt.bris.ac.uk/discovery/chatlogs/rdfig/

17:40:09 <danbri_dna> libby, i think karl has some knowhow re the math module. may be misremembering.

17:42:16 <libby> thanks

17:42:40 <libby> I haven't really looked into it myself yet, just wondered if there was any instant knowledge on the channel

17:55:01 * sh1mmer waves

18:01:19 <libby> heya sh1m

18:01:30 <libby> you were right about carry on films of course

18:11:49 <sh1mmer> libby im busy playing with my foaf file

18:12:02 <sh1mmer> libby eaon had some fun things in his so i stole them

18:12:03 <sh1mmer> :D

18:15:51 <libby> :)

19:08:17 <bblfish> bblfish is now known as bblfish_gone_for

19:08:32 <bblfish_gone_for> bblfish_gone_for is now known as bblfish_dinner

19:18:46 <eaon> nice people of #rdfig

19:19:03 <eaon> am i an idiot when i do <foaf:homepage rdf:resource="http://example.com" dc:Title="My homepage" /> ? is that valid?

19:19:21 <eaon> shouldn't i use rdf:Description to apply the title?

19:20:07 <danja> looks ok, except for dc:title => dc:title

19:20:21 <danja> best test is the RDF Validator

19:20:31 <danja> .google rdf validator

19:20:31 <mattmcc> You get away with not needing an rdf:Description because in RDF/XML it comes out the same.

19:20:56 <kota> literal properties can be abbreviated. so it's fine

19:21:02 <eaon> thanks

19:21:22 <eaon> RDF/XML feels weird :P

19:21:41 <kota> everyone knows that :P

19:22:13 <mattmcc> It tends to take people awhile to get used to different XML files resulting in the same graph.. :)

19:23:38 <eaon> well i know that, but i didn't know i can do <foaf:homepage rdf:resource="http://example.com/"><dc:title>My homepage</dc:title></foaf:homepage>

19:23:58 <crschmidt> feels weird to who?

19:24:03 <crschmidt> It feels just fine to me :)

19:39:35 <danja> hang on,

19:46:01 <danja> not sure about that last version

19:48:37 <danja> I /think/ the all-attributes version is ok, but I'm not sure about the nested <title> version

19:49:24 <danja> grr, isaviz playing up

19:50:23 <mattmcc> eaon: That you can't do.

19:51:08 <jsled> well .... you can do it, but it'll mean something different?

19:51:24 <mattmcc> No, that breaks striping.

19:51:58 <danja> thanks matt - thought it looked funny

19:52:47 <eaon> so for that i'd need rdf:Description

19:52:59 <danja> think so, yep

19:53:01 <mattmcc> A class of some sort.

19:53:12 <danja> or use the attributes version

19:53:25 <mattmcc> The range of foaf:homepage is a foaf:Document, so I'd use that.

19:54:52 <eaon> so what does dc:title refer to then when i'm not using a class?

19:58:05 <mattmcc> The resource.

20:00:43 <eaon> so this is <foaf:homepage rdf:resource="http://example.com/" rdf:nodeID="Foo" /> <dc:title rdf:nodeID="Foo">My homepage</dc:title> ?

20:01:54 <mattmcc> Ah, nodeId and resource don't mix.

20:02:09 <eaon> yeah right but i you know what i mean

20:03:20 <mattmcc> The look you're probably going for is <foaf:homepage rdf:resource="http://example.com/" /> and <rdf:Description rdf:about="http://example.com/"><dc:title>My homepage</dc:title></rdf:Description>

20:03:33 <danja> {Person} -homepage-> http://example.org

20:03:42 <mattmcc> Which is roughly what I do in my foaf.rdf. http://mmcc.cx/foaf.rdf

20:10:11 <eaon> so <foaf:homepage rdf:resource="http://example.com" dc:Title="My homepage" /> is a(n imho) heavy exception?

20:16:57 <mortenf> C:

20:16:57 <dc_rdfig>http://www.w3.org/2001/sw/Europe/events/200407-cph/

20:16:59 <dc_rdfig> Workshop on Metadata for a multilingual world - Day 2

20:16:59 <dc_rdfig> (1:mortenf_) [scribles|http://rdfig.xmlhack.com/2004/07/15/2004-07-15.html], [chatlogs|http://www.ilrt.bris.ac.uk/discovery/chatlogs/rdfig/2004-07-15.html] and [Morten's pictures|http://www.wasab.dk/morten/2004/07/photos/cph/1/] from day 1

20:17:18 <mortenf> C:[pictures from day 2|http://www.wasab.dk/morten/2004/07/photos/cph/2/]

20:17:19 <dc_rdfig> Added comment C2.

20:38:06 <dajobe> stefano is just so wrong

20:38:42 <eaon> with?

20:39:00 <dajobe> his http://www.betaversion.org/~stefano/linotype/news/57/

20:39:35 <dajobe> replying to it would be of zero use

20:40:17 <eaon> uhm, what exactly is wrong with it?

20:40:28 <dajobe> facts

20:41:54 <eaon> i haven't read it properly, can you point me a specific section? :)

20:42:03 <dajobe> well, what does OWL stand for?

20:42:09 <dajobe> what is an XML person?

20:42:28 <jsled> dajobe: who doesn't believe they're an XML person, these days?

20:42:36 <dajobe> and then yawn ivory towers yadda yadda

20:43:03 <dajobe> I've done (codeded, created, read about) enough XML to justify that

20:43:23 * Arnia thinks stefano should read more AA Milne

20:44:58 <dajobe> lovely prom that

20:45:08 <dajobe> elgar++

20:45:45 <Arnia> The organ sounds great now

23:15:37 <drumm> not entirely RDF, but the namespace url at http://web.resource.org/rss/1.0/modules/audio/ is a 404

23:16:02 <drumm> what now?

23:17:34 <jsled> no it's not...

23:17:42 <jsled> sorry, the namespace URL.