Semantic Web Interest Group IRC Chat Logs for 2003-03-29

01:46:43 <taaz> is there a good way to embed arbitrary xml in rdf? for instance, embedded xhtml markup?

01:47:08 <taaz> using CDATA sort of works for what i'm doing but it seems kind of a lame way to do it

04:58:25 <DanC> taaz, have you tried <myProp rdf:parseType="Literal">any <markup/> here</myProp> ?

06:59:18 <taaz_> taaz_ is now known as taaz

10:29:13 <tav|offline> tav|offline is now known as tav

10:51:18 <golbeck> golbeck is now known as golbeck_zzz

11:37:57 <arnarl-home> hi

11:40:16 <danbri> morning

12:54:18 <Wack> hi

12:54:35 <Wack> well, morning, it's 13:54 here :]

13:08:30 * sandro and danbri, discussing whether speciing and ontology alone (and "RDF/XML") would be enough of a data format for RSS....

13:09:40 <sandro> I think OWL-Lite/DL's Data-Valued Properties might really do what you want. I think a Cardinality=1 Data-Valued Property is the same as a required field. But I'm not sure, and I've heard suggestions otherwise.

13:10:20 <sandro> I guess leaving it out might not show up as an error in any standard OWL validators; but it probably could/should.

13:10:36 <danbri> logger_1, pointer?

13:10:36 <danbri> See http://ilrt.org/discovery/chatlogs/rdfig/2003-03-29#T13-10-36

13:10:42 <danbri>http://ilrt.org/discovery/chatlogs/rdfig/2003-03-29#T13-10-36

13:10:43 <dc_rdfig> A: http://ilrt.org/discovery/chatlogs/rdfig/2003-03-29#T13-10-36 from danbri

13:10:52 <danbri> A:|sandro and danbri, discussing whether speciing and ontology alone (and "RDF/XML") would be enough of a data format for RSS....

13:10:52 <dc_rdfig> Titled item A.

13:11:00 <danbri> A:Placeholder to come back to, I have to go offline now!

13:11:01 <dc_rdfig> Added comment A1.

13:11:05 <sandro> Ciao!

13:11:09 <danbri> cu

14:08:31 <timbl> Sandro, re yesterday's comment about a cardinality=1 data value being a required field ( http://ilrt.org/discovery/chatlogs/rdfig/2003-03-29#T13-09-40 ), remember it applies to the thing not the form.

14:09:32 <timbl> So, for example, it might apply to the registration numebr of a car.

14:09:47 <timbl> Every car has one and only one registration number.

14:10:24 <timbl> When you build a form about the car, then if there is a registration number field on it, clearly no one should be able to fill in two different numbers.

14:10:28 <JibberJim> at any point in time...

14:10:55 <timbl> But that doesn't mean that a person has to fillin a number on ant particular form.

14:11:23 <timbl> It doesn't mean that any particupalr form has to have that slot.

14:11:40 <timbl> It doesn't mean that any particulare software object has to have a slot for it.

14:12:22 * timbl yes JibberJim, in the particular universe of discourse under any hypothetical conditions, etc

14:13:30 <timbl> The question of what slots/form fields are displayed or asked for is one which isaViz puts in a style sheet. A style sheet can be a whole bidirectional link between semweb and UI. It can give you all you need to build an application.

14:14:05 <timbl> Maybe a "style sheet" is an idea to use for mapping OO systems onto teh semweb, too.

14:15:39 * timbl is done :)

14:21:31 <timbl> logger_1, pointer?

14:21:31 <timbl> See http://ilrt.org/discovery/chatlogs/rdfig/2003-03-29#T14-21-31

14:21:35 <timbl> A:

14:21:36 <dc_rdfig>http://ilrt.org/discovery/chatlogs/rdfig/2003-03-29#T13-10-36

14:21:37 <dc_rdfig> sandro and danbri, discussing whether speciing and ontology alone (and "RDF/XML") would be enough of a data format for RSS....

14:21:38 <dc_rdfig> (1:danbri) Placeholder to come back to, I have to go offline now!

14:22:38 <timbl> A: [tim continues| http://ilrt.org/discovery/chatlogs/rdfig/2003-03-29#T14-21-31]

14:22:39 <dc_rdfig> Added comment A2.

14:23:32 <timbl> JibberJim, you have done a lot of cool SVG-RDF stuff.

14:24:31 <timbl> Do you think a common language for style sheets would work?

14:27:09 <JibberJim> I don't know, I'm very much a scripter in all I do, and I'm not particularly happy with stylesheets at all, I don't have the right mindset.

14:43:04 <sandro> thanks, TimBL. I was kind of forgetting that. It's a use/mention kind of thing, I guess, to blur that distinction.

15:26:00 <timbl> Oh, and sandro - you asked for number literals in cwm - you've got 'em

16:29:07 <sandro> with math support, Tim? :-)

16:30:53 <sandro> Could the RSS (ontology as syntax) problem be addressed by a query instead? "An RSS Document is an RDF/XML document which has one or more matches to this query ...." ?

16:31:19 <sandro> ... and the SEMANTICS of the terms in the query would still come from the ontology.

16:44:34 * bitsko wonders the context of this dicsussion

16:49:30 <bitsko>http://dev2dev.bea.com/articles/JSchneider_XML.jsp

16:49:31 <dc_rdfig> B: http://dev2dev.bea.com/articles/JSchneider_XML.jsp from bitsko

16:49:52 <bitsko> B:|Native XML Scripting

16:49:53 <dc_rdfig> Titled item B.

16:50:41 <bitsko> B:Describes adding XML literals and operators to ECMAScript.

16:50:41 <dc_rdfig> Added comment B1.

16:51:55 <bitsko> B:An XML literal is exactly as one would expect: foo = <p>some <i>random</i> well-formed markup</p>;

16:51:55 <dc_rdfig> Added comment B2.

16:54:49 <bitsko> B:«A very common task in XML processing is navigation. Native XML scripting addresses this problem by adopting the familiar dot operator ".", commonly used for accessing the fields, methods and properties of an object, for accessing the children of an XML value. In addition, it adds a new dot-dot operator ".." for accessing descendents (i.e., children, grandchildren, etc.) of an XML value. These operators are conceptually similar to the "/" and "

16:54:50 <bitsko> /" operators found in XPath, but they use syntax designed to be more intuitive for developers familiar with navigating objects (and they don?t conflict with the use of "/" and "//" for numeric division and comments).»

16:54:51 <dc_rdfig> Added comment B3.

16:56:30 <bitsko> B3:«A very common task in XML processing is navigation. Native XML scripting addresses this problem by adopting the familiar dot operator ".", commonly used for accessing the fields, methods and properties of an object, for accessing the children of an XML value. In addition, it adds a new dot-dot operator ".." for accessing descendents (i.e., children, grandchildren, etc.) of an XML value. These operators are conceptually similar to the "/" and

16:56:30 <dc_rdfig> Replaced comment B3.

16:56:42 <bitsko> B:"/" operators found in XPath, but they use syntax designed to be more intuitive for developers familiar with navigating objects (and they don?t conflict with the use of "/" and "//" for numeric division and comments).»

16:56:43 <dc_rdfig> Added comment B4.

17:12:06 <sandro> DanC, I wonder if you could make a page to elucidate on "the pattern language we're building".

17:28:08 <silvernerd> hi

17:28:28 <silvernerd> I've developed a way of relative addressing for the GNU.RDF project

17:29:15 <silvernerd> called link paths that are of the form 'resource.property.property....'

17:29:16 <bitsko> relative addressing?

17:29:28 <bitsko> ah, navigation

17:29:41 <silvernerd> yes

17:29:55 <silvernerd> not hardlinking with uri's

17:30:11 <silvernerd> IMHO that's a potential source of problems once things change in the web

17:31:15 <silvernerd> the trick with link paths is that they can navigate through collections

17:31:27 <silvernerd> by specifying coditions

17:31:45 <silvernerd> because link paths get translated into GNU.RDF.QL (the query language)

17:35:19 <bitsko> do you have more examples?

17:39:40 <silvernerd> there's a whole discussion thread about the query language starting at: http://dotgnu.org/pipermail/developers/2003-March/010251.html

17:40:27 <silvernerd> the link path idea came up at: http://dotgnu.org/pipermail/developers/2003-March/010278.html

17:40:52 <silvernerd> it was worked out further at: http://dotgnu.org/pipermail/developers/2003-March/010295.html

17:48:43 <bitsko> what does '@@' do?

17:49:07 <silvernerd> nothing

17:49:15 <silvernerd> it's a syntax thing

17:49:18 <silvernerd> like ; in C

17:50:56 <bitsko> by coincidence, this morning I chumped an article that describes an XML navigation pattern similar to the link paths you describe, see today's rdfig.xmlhack.com

17:51:50 <bitsko> I had been tempted to add, "now we just need something like this for RDF"

17:58:28 <silvernerd> would link path be a good RDF replacement for this?

17:58:40 <bitsko> appears so

18:05:16 <bitsko> I had that working in Perl at one point, $resource->{Property}{Property}, but it's not as useful without conditions.

18:06:42 <silvernerd> only problem is that the conditions are in GNU.RDF.QL

18:07:06 <silvernerd> but that means it's like SQL

18:10:13 <bitsko> I think there's two seperate patterns. one is where you query a large base of data to retrieve selected resources (SQL-like) and one where, once you have a resource, you navigate thru properties of the resource (XPath-like). I've found my needs tend towards the latter

18:11:12 <bitsko> I would also think (and maybe this is what you meant) that the latter is used within the former

18:11:53 <silvernerd> yes

18:12:10 <silvernerd> link paths are technically a shorthand for GNU.RDF.QL queries

18:12:31 <silvernerd> but they are also used inside those queries

18:14:00 <silvernerd> I've got to go

18:14:03 <silvernerd> bye

18:14:11 <bitsko> later

18:39:09 <danbri> <sandro> Could the RSS (ontology as syntax) problem be addressed by a query instead? "An RSS Document is an RDF/XML document which has one or more matches to this query ...." ?

18:39:28 <danbri> I think so, eventually, if xml schemas could be reformulated in terms of an ontology for xml document types

18:39:48 <danbri> libby and i played around with this a bit, the schemerama thing, sort of like Schematron but over top of rdf query

18:40:53 <danbri>http://ilrt.org/discovery/2001/02/schemarama/

18:40:54 <dc_rdfig> C: http://ilrt.org/discovery/2001/02/schemarama/ from danbri

18:41:00 <danbri> C:|Schemarama experiment(s)

18:41:00 <dc_rdfig> Titled item C.

18:41:12 <danbri> C:Using RDF query for RDF/XML document typing

18:41:12 <dc_rdfig> Added comment C1.

18:41:22 <danbri> C:A bit rough but imho an area worth revisiting now we have OWL

18:41:22 <dc_rdfig> Added comment C2.

18:42:00 <danbri> C:with [http://swordfish.rdfweb.org/discovery/2001/01/schemarama/|online demo] (rss example...)

18:42:00 <dc_rdfig> Added comment C3.

19:00:56 <danbri>http://downlode.org/perl/wikibot/

19:00:56 <dc_rdfig> D: http://downlode.org/perl/wikibot/ from danbri

19:01:07 <danbri> D:|Wikibot is an IRC bot that puts text on a wiki

19:01:08 <dc_rdfig> Titled item D.

19:01:51 <danbri> D:Related to accretor, see [http://0xff.org/accretion.pl?IRC_Scratchpad|irc/wiki scratchpad page]

19:01:51 <dc_rdfig> Added comment D1.

19:02:08 <danbri> D:Hex/Grault knows more. I'm just rummaging around this stuff now.

19:02:08 <dc_rdfig> Added comment D2.

19:03:12 <danbri> D:Wonder how it deals with parallel updates problem...

19:03:12 <dc_rdfig> Added comment D3.

19:06:31 <danbri> D:<hex> accretor is an instance of wikibot.

19:06:31 <dc_rdfig> Added comment D4.

19:06:51 <grault> grault is now known as earle

19:07:31 * earle waves

19:07:52 <earle> D:hex is earle on this network

19:07:52 <dc_rdfig> Added comment D5.

19:07:53 * danbri waves

19:08:34 <earle> D: as to parallel updates, um, I have no idea. It's not very sophisticated :)

19:08:34 <dc_rdfig> Added comment D6.

19:10:04 <earle> what's the syntax for a hyperlink to IRC?

19:10:15 <danbri> irc: something

19:10:19 * danbri rummages

19:11:17 <earle> oh, it's on the scratchpad :)

19:11:29 <earle> . irc://irc.freenode.net:6667/rdfig

19:11:47 <danbri> . http://esw.w3.org/topic/AboutThisService -> I thought linked to more info, but can't find that now

19:12:22 <danbri> I keep meaning to read up on how # is dealt with, since # is really part of the channel name, but has special meaning in URIs

19:12:23 <earle> D:accretor is running in <a href="irc://london.rhizomatic.net:6667/core">#core on rhizomatic</a>.

19:12:23 <dc_rdfig> Added comment D7.

19:12:26 <danbri> guess it should be escaped

19:12:29 * earle nods

19:13:03 <earle> oops

19:13:06 <danbri> D:Ok for #rdfig folk to join #core and play about a bit? I'd be interested in possibility of running it here...

19:13:06 <dc_rdfig> Added comment D8.

19:13:38 <earle> D7:accretor is running in [#core on rhizomatic|irc://london.rhizomatic.net:6667/core]

19:13:39 <dc_rdfig> Replaced comment D7.

19:13:44 <earle> D:absolutely

19:13:44 <dc_rdfig> Added comment D9.

19:26:34 <danbri>http://mydomain.com/about/aljazeera

19:26:34 <dc_rdfig> E: http://mydomain.com/about/aljazeera from danbri

19:26:47 <danbri> E:|ALJAZEERA.NET DOMAIN HACKED AND MOVED TO MYDOMAIN.COM

19:26:47 <dc_rdfig> Titled item E.

19:26:54 <danbri> E:" Arab satellite TV network?s nameservers changed to MyDomain URL forwarding service by hacker; MyDomain reacts by suspending service."

19:26:54 <dc_rdfig> Added comment E1.

19:27:10 <danbri> E:"VANCOUVER, Wash. ? MyDomain.com, host to more than 600,000 domain names using its free URL forwarding and e-mail services, discovered this morning the domain aljazeera.net had been moved to MyDomain servers. MyDomain has learned from NavLink, the company that hosts the aljazeera.net Web site from its data centers in France, that Al Jazeera?s domain name account at Network Solutions was compromised. The hacker changed the Web site?s na

19:27:10 <danbri> to point to MyDomain nameservers and erected a page displaying the American flag."

19:27:10 <dc_rdfig> Added comment E2.

19:27:20 <danbri> E:Where's the Web Of Trust when you need one?

19:27:20 <dc_rdfig> Added comment E3.

19:27:57 <danbri> E2:"VANCOUVER, Wash. ? MyDomain.com, host to more than 600,000 domain names using its free URL forwarding and e-mail services, discovered this morning the domain aljazeera.net had been moved to MyDomain servers. MyDomain has learned from NavLink, the company that hosts the aljazeera.net Web site from its data centers in France, that Al Jazeera?s domain name account at Network Solutions was compromised.(...)"

19:27:57 <dc_rdfig> Replaced comment E2.

19:28:17 <danbri> E:"(...) The hacker changed the Web site?s nameservers to point to MyDomain nameservers and erected a page displaying the American flag."

19:28:17 <dc_rdfig> Added comment E4.

19:28:45 <danbri> E:A Semantic Web use case? proving you own an internet domain...

19:28:45 <dc_rdfig> Added comment E5.

19:29:19 <niq> E:Proof of the reality of cyberterrorism (as if we needed it)

19:29:19 <dc_rdfig> Added comment E6.

19:29:22 <danbri> E:I read this hack was pulled using faked letter headed paper. It's like PGP never existed... :(

19:29:22 <dc_rdfig> Added comment E7.

19:29:53 <niq> theregister had a story about this recently

19:31:51 <danbri> earle, http://esw.w3.org/topic/ChannelsAreResourcesToo re irc URIs

19:34:09 <earle> danke

19:37:46 <golbeck_zzz> golbeck_zzz is now known as golbeck

20:53:03 * DanC waves

20:53:26 * DanC heads out...

21:30:23 <earle> earle is now known as grault

21:58:36 <jeremiah> hello

22:23:32 <c> dc_rdfig:view

22:23:32 <dc_rdfig> A: sandro and danbri, discussing whether speciing and ontology alone (and "RDF/XML") would be enough of a data format for RSS.... (http://ilrt.org/discovery/chatlogs/rdfig/2003-03-29#T13-10-36)

22:23:33 <dc_rdfig> B: Native XML Scripting (http://dev2dev.bea.com/articles/JSchneider_XML.jsp)

22:23:34 <dc_rdfig> C: Schemarama experiment(s) (http://ilrt.org/discovery/2001/02/schemarama/)

22:23:36 <dc_rdfig> D: Wikibot is an IRC bot that puts text on a wiki (http://downlode.org/perl/wikibot/)

22:23:36 <dc_rdfig> E: ALJAZEERA.NET DOMAIN HACKED AND MOVED TO MYDOMAIN.COM (http://mydomain.com/about/aljazeera)

22:23:38 <c> c is now known as danbri_lap

22:24:29 <danbri_lap> D:I forked off a POE-based version, [http://rdfweb.org/2003/03/oaf/oaf.pl|oaf.pl], as I want one for #foaf and Net::IRC doesn't work from the rdfweb.org server.

22:24:29 <dc_rdfig> Added comment D10.

22:25:53 <danbri_lap> D:OAF doesn't work yet, but the things-to-do-next should be obvious. It does run, and uses POE to connect to two IRC channels, demo'd by copying text between them. Note that this makes a multi-headed chatbot look rather feasible.

22:25:53 <dc_rdfig> Added comment D11.

22:26:10 <danbri_lap> D:...although let's not get ahead of ourselves. It don't work at all yet.

22:26:10 <dc_rdfig> Added comment D12.

22:26:27 <danbri_lap> D:But multi-headed chatbot would be sweet. Did I mention that yet? ;)

22:26:27 <dc_rdfig> Added comment D13.

23:35:16 <pi314159> pi314159 is now known as golbeck

23:38:54 <danbri_lap> D:Ok, it works now. Multi-headed (after a fashion) too. Running as 'oafbot' in #foaf and #oaf, scrawling into [http://rdfweb.org/rweb/wiki/wiki?ScratchPad FoafWiki:ScratchPad].

23:39:12 <dc_rdfig> Added comment D14.

23:39:26 <danbri_lap> D:

23:39:26 <dc_rdfig>http://downlode.org/perl/wikibot/

23:39:27 <dc_rdfig> Wikibot is an IRC bot that puts text on a wiki

23:39:28 <dc_rdfig> (1:danbri) Related to accretor, see [http://0xff.org/accretion.pl?IRC_Scratchpad|irc/wiki scratchpad page]

23:39:29 <dc_rdfig> (2:danbri) Hex/Grault knows more. I'm just rummaging around this stuff now.

23:39:30 <dc_rdfig> (3:danbri) Wonder how it deals with parallel updates problem...

23:39:31 <dc_rdfig> (4:danbri) <hex> accretor is an instance of wikibot.

23:39:32 <dc_rdfig> (5:earle) hex is earle on this network

23:39:33 <dc_rdfig> (6:earle) as to parallel updates, um, I have no idea. It's not very sophisticated :)

23:39:34 <dc_rdfig> (7:earle) accretor is running in [#core on rhizomatic|irc://london.rhizomatic.net:6667/core]

23:39:35 <dc_rdfig> (8:danbri) Ok for #rdfig folk to join #core and play about a bit? I'd be interested in possibility of running it here...

23:39:38 <dc_rdfig> (9:earle) absolutely

23:39:40 <dc_rdfig> (10:danbri_lap) I forked off a POE-based version, [http://rdfweb.org/2003/03/oaf/oaf.pl|oaf.pl], as I want one for #foaf and Net::IRC doesn't work from the rdfweb.org server.

23:39:43 <dc_rdfig> (11:danbri_lap) OAF doesn't work yet, but the things-to-do-next should be obvious. It does run, and uses POE to connect to two IRC channels, demo'd by copying text between them. Note that this makes a multi-headed chatbot look rather feasible.

23:39:47 <dc_rdfig> (12:danbri_lap) ...although let's not get ahead of ourselves. It don't work at all yet.

23:39:49 <dc_rdfig> (13:danbri_lap) But multi-headed chatbot would be sweet. Did I mention that yet? ;)

23:39:51 <dc_rdfig> (14:danbri_lap) Ok, it works now. Multi-headed (after a fashion) too. Running as 'oafbot' in #foaf and #oaf, scrawling into [http://rdfweb.org/rweb/wiki/wiki?ScratchPad FoafWiki:ScratchPad].

23:40:09 <danbri_lap> D14:Ok, it works now. Multi-headed (after a fashion) too. Running as 'oafbot' in #foaf and #oaf, scrawling into [http://rdfweb.org/rweb/wiki/wiki?ScratchPad|FoafWiki:ScratchPad].

23:40:09 <dc_rdfig> Replaced comment D14.