Semantic Web Interest Group IRC Chat Logs for 2003-07-29

00:36:15 <bitsko> are there any examples of sites that are using Dublin Core extensively within their own web published collections, in or not in RDF?

00:38:47 <deltab> I've seen some managed-content sites using it

00:40:20 <bitsko> I'm mostly interested in whether or not they include the DC in the published resources

00:40:31 <deltab> some news site, I think

00:42:41 <deltab> cnn has description, author, section, subsection, date

00:44:15 <deltab> cnet has description and keywords

00:45:32 <deltab> msn has really messed-up html

01:45:21 <jimH_>http://obo.sourceforge.net/#

01:45:22 <dc_rdfig> A: http://obo.sourceforge.net/# from jimH_

01:45:46 <jimH_> A:| Open Biological Ontologies (OBO) at sourceforge

01:45:47 <dc_rdfig> Titled item A.

01:46:04 <jimH_> A: library for lots of ontologies

01:46:04 <dc_rdfig> Added comment A1.

01:46:37 <jimH_> A: most in DAG-edit (gene ontology) format - some now in DAML and RDFS and OWL

01:46:37 <dc_rdfig> Added comment A2.

01:47:59 <jimH_> A: goals: "ontologies must be in a common, shared syntax" GO and OWL only two listed

01:47:59 <dc_rdfig> Added comment A3.

05:15:27 <mdupont> hi burtonator

05:17:48 <burtonator_> yo

09:41:37 <darobin_> darobin_ is now known as darobin

10:16:10 <arnarl> hi

12:10:18 * danbri sends rdfcore minutes; sorry didn't get to these yesterday

12:33:04 <ericP> is jan_g around?

12:33:49 <ericP> (though i don't know if the 'g' suffix indicated a location or something like that)

12:38:32 <dajobe> he's in a meeting

12:40:44 <ericP> tx, i'll try later

12:40:58 <dajobe> I've forwarded the message

12:41:57 <ericP> been trying to figure out if a p1 ?t1. ?t1 p2 ?t2. is a constriant or a cross product

12:42:09 <ericP> or, more likely, a constrained cross product

12:46:50 * danbri assumed 'g' meant 'Grant'

12:50:33 <arnarl> arnarl is now known as arnarl|away

13:03:44 <Dorward> OK, I have an RDF Calendar, and I have the databook xslt file, now I need to find myself an XLST processor. Any recomendations?

13:05:45 <danbri> i'd expect http://xml.apache.org/xalan-j/ to be good

13:06:01 <Dorward> s/XLST/XSLT/

13:06:37 <Dorward> The only one I've tried already :)

13:06:41 <Dorward> But its spittiing out SystemId Unknown; Line -1; Column -1; XSLT Error (javax.xml.transform.TransformerConfigurationException): javax.xml.transform.TransformerException: javax.xml.transform.TransformerException: xsl:variable is not allowed in this position in the stylesheet!

13:06:58 <Dorward> And my experience with xslt is next to nil, so I'm not sure what to do about it.

13:07:11 <Dorward> I was going to try a different app to see if it was a flaw in the xslt or the processor.

13:08:03 * danbri not too familiar w/ xslt either, sorry

13:08:12 * danbri wonders if there's an xslt validator anywhere

13:08:13 <Dorward> Ah, not to worry.

13:08:22 <Dorward> That's a good point...

13:33:25 <danbri>http://www.w3.org/2001/11/IsaViz/gss/gssmanual.html

13:33:26 <dc_rdfig> B: http://www.w3.org/2001/11/IsaViz/gss/gssmanual.html from danbri

13:33:35 <danbri> B:|Graph Stylesheets in IsaViz

13:33:35 <dc_rdfig> Titled item B.

13:37:40 <wkearney_away> wkearney_away is now known as wkearney99

13:49:09 <arnarl|away> arnarl|away is now known as arnarl

14:49:30 <dajobe>http://lists.w3.org/Archives/Public/www-rdf-interest/2003Jul/0199.html

14:49:31 <dc_rdfig> C: http://lists.w3.org/Archives/Public/www-rdf-interest/2003Jul/0199.html from dajobe

14:49:45 <dajobe> C:|Raptor 0.9.11 - now passes all latest RDF Core tests

14:49:45 <dc_rdfig> Titled item C.

14:53:26 <rO|_> hi

14:54:11 <sandro> Wooo hoo dajobe!

14:54:18 <rO|_> what are 'triples'?

14:54:54 <deltab> groupings of three items

14:55:13 <dajobe> try the RDF primer basic concepts: http://www.w3.org/TR/rdf-primer/#basicconcepts

14:55:16 <dajobe> three-part sentences

15:00:58 <rO|_> thx

15:20:31 <arnarl> arnarl is now known as arnarl|shopping

15:41:24 <arnarl|shopping> arnarl|shopping is now known as arnarl

16:16:26 <mhgrove> #mindswap

16:58:11 <tav> tav is now known as tav|offline

17:24:42 <danja> Dorward - still having probs with XSLT?

17:25:06 <Dorward> danja: Indeed. No progress since I sent that email.

17:25:12 <Dorward> err

17:25:18 <Dorward> Nope, I sent the email to DanC :)

17:25:32 <danja> I saw your note in the logs

17:25:39 <danja> which XSLT file?

17:25:46 <Dorward> I tried a different processor and got a more useful error, so I'm sure that its with the file.

17:26:03 <Dorward>http://dev.w3.org/cvsweb/~checkout~/2001/palmagent/datebook2html.xsl?rev=1.14

17:26:03 <dc_rdfig> D: http://dev.w3.org/cvsweb/~checkout~/2001/palmagent/datebook2html.xsl?rev=1.14 from Dorward

17:26:08 <danja> heh

17:26:40 <Dorward> I don't think I wanted that to happen :)

17:30:11 <DanC> somebody sent me a fix for that; I wonder if it's in the www-rdf-calendar archive

17:31:03 <tav|offline> tav|offline is now known as tav

17:31:11 <danja> which is a test file?

17:31:18 <danja> datebook sample

17:33:43 <Dorward> Can't find it in the archive.

17:35:18 <danja> ok, got one, got same error re. xsl:variable

17:35:28 <danja> using Xalan

17:36:08 <Dorward> Error at xsl:variable on line 153 of file:/home/david/tmp/cal/datebook2html-1.14.xsl:

17:36:09 <Dorward> Variable is already declared (on line 113)

17:36:11 <Dorward> From Saxan

17:36:15 <Dorward> Saxon even

17:37:03 * danja must add line numbering to his XSLT editor hack...

17:50:01 <Dorward> DanC: Any joy in finding that fix?

17:50:20 <DanC> no; didn't really look.

17:50:45 <DanC> I know it's easy, once one is in XSLT-hacking mode. I'm not likely to get there soon, alas.

17:50:47 <mortenf> sablotron barfs on the use of variables in template match attributes

17:51:04 <Dorward> DanC: Ah, thanks anyway.

17:51:14 <mortenf> what's the error you get dorward?

17:51:39 <DanC> hmm... gaim doesn't do readline style editing.

17:52:00 <DanC> and I thought I told it to log everything, but that doesn't seem to have stuck.

17:52:10 <DanC> so xchat stays, for at least another day

17:52:16 <Dorward> $ java net.sf.saxon.Transform cal.rdf datebook2html-1.14.xsl

17:52:16 <Dorward> Error at xsl:variable on line 153 of file:/home/david/tmp/cal/datebook2html-1.14.xsl:

17:52:16 <Dorward> Variable is already declared (on line 113)

17:52:16 <Dorward> Transformation failed: Failed to compile stylesheet. 1 error detected.

17:52:56 <mortenf> hmm

17:53:16 <mortenf> that's not even a variable declaration?

17:54:09 <Dorward> mortenf: I think its a conflict between <xsl:param name="kb"/> and <xsl:variable name="kb" select="$kb"/> - two different XSLT processers threw errors on it anyway.

17:55:44 <mortenf> has anyone ever run this - it seems quite full of errors?

17:56:17 <Dorward> According to the archives, DanC has been using it to export his calendar for friends and family.

17:56:26 <mortenf> that line is not permitted there, and it seems it should be xsl:with-param instead

17:57:21 <DanC> try renaming the param

17:57:34 <DanC> er... sorry; try renaming the local variable

17:57:44 <DanC> either way.

17:58:15 <mortenf> ok, changed that variable line to xsl:with-param instead, removed pre and div at top, and manually copied the content of the thisTrans variable (a static URI) to where it was used; now compiles.

17:58:35 <mortenf> anyone have some instance data?

17:59:03 <Dorward> mortenf: Ah, excellent. Could you stick it somewhere I could grab it from please? I'll test it in my Evolution data

17:59:58 <mortenf> hang on...

18:00:13 <_joshua> DanC: You might try using a proxy IRC server for the logging

18:01:06 <mortenf> http://www.wasab.dk/morten/07/datebook2html-1.14-mof.xsl

18:02:02 <Dorward> mortenf: 404

18:02:30 <mortenf> sorry, morten/2003/07/

18:02:49 <Dorward> ah, that's my favourite trick :)

18:05:49 <Dorward> hmm... had to convert some "--" to "-" inside comments.

18:06:16 <mortenf> yeah, forgot that...

18:06:44 <mortenf> fixed as well...

18:07:27 <Dorward> OK, this is odd - it doesn't seem to be performing any transformation.

18:09:34 <mortenf> do you have some sample input?

18:10:02 <Dorward> I'm just trying to grab some simplified input data.

18:16:21 * DanC wanedered off to a telcon...

18:17:29 * Dorward mutters about his Evolution data containing way too much personal stuff that is ancient...

18:24:27 <Dorward> OK, different set of data. http://david.us-lot.org/tmp/rdf/

18:24:31 <Dorward> The XSL seems to be outputting RDF

18:24:33 * Dorward notes the upload hasn't started transfering yet.

18:24:37 <Dorward> OK, they are up.

18:25:46 <Dorward> OK, that was interesting. http://david.us-lot.org/tmp/rdf/ contains the files I've been playing with. The XSL seems to be outputting RDF instead of HTML.

18:28:08 <mortenf> i don't know what this XSLT is supposed to do, but it seems it reads HTML, not iCal?

18:28:31 <mortenf> erh, s/iCal/rdf/

18:30:55 <Dorward> That would explain why it isn't working... it doesn't explain why the name is databook2html!

18:31:11 <mortenf> it's an identity transformation, except that it catches html:object elements with classid and data attributes, the last of which seems to be poiting to an rdf/XML doc

18:31:17 <mortenf> indeed...

18:31:52 <Dorward> Well, this is driving me nuts. I'll get something to eat then think about it again.

19:09:31 <collord> anyone know of any ontologies of emotions?

19:39:21 * ericP bursts into tears in a pitiful confession that he know now such ontology

19:52:20 * DanC waves to ghuo

19:52:46 <DanC> frap; it seems that the 1 RdfCalendar test file that we had round-tripping has fallen over

19:52:56 <DanC> connolly@dirk:~/w3ccvs/WWW/2002/12/cal$ make test

19:52:56 <DanC> PYTHONPATH=../../../2000/10/swap python ../../../2000/10/swap/pim/toIcal.py test/mtg.rdf >test/mtg,.ics

19:52:57 <DanC> ...

19:53:06 <DanC> File "../../../2000/10/swap/sax2rdf.py", line 583, in feed

19:53:07 <DanC> self._p.feed(data)

19:53:07 <DanC> File "/usr/lib/python2.2/site-packages/_xmlplus/sax/expatreader.py", line 211, in feed

19:53:07 <DanC> self._err_handler.fatalError(exc)

19:53:07 <DanC> File "/usr/lib/python2.2/site-packages/_xmlplus/sax/handler.py", line 38, in fatalError

19:53:07 <DanC> raise exception

19:53:09 <DanC> xml.sax._exceptions.SAXParseException: <unknown>:1:0: syntax error

19:53:11 <DanC> make: *** [test/mtg-compare.n3] Error 1

19:53:55 <DanC> ah; my workspace was screwey... cvs updating...

19:54:55 * ghuo waves to DanC

19:55:13 <ghuo> okay, let me try this ``phone'' contraption

19:59:56 <DanC> odd...

19:59:57 <DanC> Include can only work on formulae 40) short=0, 0_work :: `test/mtg,.rdf` log:includes G1(G1)?.

20:00:06 <ghuo> hmmm

20:00:14 <ghuo> i don't see that error

20:00:51 <Dorward> ghuo: Phone? That's one of those small plastic boxes that server to route TCP/IP over GPRS+bluetooth isn't it? :)

20:01:01 <DanC> <prodid>-//Apple Computer\, Inc//iCal 1.0//EN</prodid>

20:02:35 <DanC> hm... sequence...

20:03:49 <DanC> wierd... too many slashes in http://www.w3.org/2002/12/cal/prod/w3.org/200_fe11812d4533096f#

20:04:35 <DanC> took out prodid, sequence, wrTimezone

20:05:56 <danbri>http://www.mapbureau.com/rdfmapper/

20:05:56 <dc_rdfig> E: http://www.mapbureau.com/rdfmapper/ from danbri

20:06:06 <danbri> E:|RDFMapper: an RDF-Based Web Mapping Service

20:06:06 <dc_rdfig> Titled item E.

20:06:50 <danbri> E:Recently [http://lists.w3.org/Archives/Public/www-rdf-interest/2003Jul/0204.html|announced] to www-rdf-interest.

20:06:50 <dc_rdfig> Added comment E1.

20:08:49 <DanC> xmlns:x=''

20:09:18 <tim142> inclusdes G1? was G1 a variable?

20:09:55 <DanC> timbl, G1 comes from http://www.w3.org/2002/12/cal/test/graphCompare.n3

20:10:43 * DanC wonders why this test ever appeared to pass

20:12:26 * DanC wonders if this phone thing is helping

20:13:20 <ghuo> that's my hacked version

20:15:12 <DanC> discussion of retest.py vs Makefile...

20:16:15 <DanC> ... discussion of costs of moving things...

20:19:07 <DanC> gee... how to make sense of prodId in the context of round-trip testing?

20:27:01 <DanC> explaining our skeptical approach to ical...

20:27:35 <DanC> prioritizing tests based on production systems over stuff that happens to be in 2002/12/cal/test/

20:27:50 * ghuo nods

20:30:42 * tim142 wondres whether DanC and ghuo are still on the bridge

20:31:20 <DanC> no, not on the bridge, but on the phone separately

20:31:35 <tim142> Make should be top level, call manifest-based test harness if necessary.

20:32:51 <ghuo> what does `manifest-based' mean?

20:33:03 <ghuo> i.e. regression.n3 for the cwm tests?

20:37:27 <tim142> re GPL: Ooops. I hadn't noticed. I didn't realize that was owned by RDFIG. Pity. That code is running. I guess I could eiethr move it out or accesss control it.

20:38:43 <tim142> yes, like retest.py

20:38:53 <tim142> manifest - rdf list of tests.

20:39:25 * tim142 wondres about combining make and manifest.

20:40:47 <tim142> There is a make2n3.py which will read a makefile and turn it into n3

20:47:46 <mortenf> ndw, would you mind sharing how you set up your conneg; i want to provide conneg with qs but without having to define each and every set of files seperately (apparently this isn't possible with MultiViews)?

20:48:46 <ndw> mortenf, i define each and every set of files seperately, alas. They're just built by the XSLT.

20:48:56 <mortenf> hmm, darn :)

20:49:52 <ndw> yeah, i thought it was pretty goofy but it seems to be the way

20:50:42 <mortenf> at least to try to combat IE...

20:54:50 <DanC> I'd much prefer you moved phpicalendar out of 2002/12/cal/ , timbl

20:55:21 <tim142> Ok

20:56:07 <tim142> Do you know how to delete a subtree from CVS?

20:56:42 <jimH_> No, how do you delete a subtree from CVS?? (Oh, sorry, thought that sounded like the first line of a joke)

20:56:52 <DanC> you cd into the repository and rm -rf

20:57:03 <DanC> or you mail sysreq

20:57:50 <DanC> there's a way to delete files in a reversible way, but (a) I don't know if it works for directories, and (b) do we want this to be reversible?

20:58:10 <ndw> it works for directories iff you remember to use -P when you do updates in the future

20:58:15 <DanC> I'd rather phpicalendar were in dev.w3.org

20:58:23 <ndw> (in other words, it doesn't really work, but -P tells cvs to ignore empty directories)

20:59:11 <DanC> did you make changes to phpicalendar, tim?

20:59:30 * DanC doesn't see the point of using php if not for database access

21:00:14 * tim142 wonders what DanC woul dsue instead

21:00:24 <DanC> python

21:01:01 <DanC> actually, if the goal is to browse .ics files, I'd use ical2rdf.pl to convert to .rdf and then XSLT to format as .html

21:01:09 <tim142> phpicalendar is a sourceforge application which works and happens to use php, and which I use for making my calendar available on the web. Do you know of a python equivalent?

21:01:32 <DanC> well, for that special meaning of "works" I suppose

21:02:12 <DanC> bert wrote a C program that produces really nice looking html/css output; it segfaulted when I tried it, though.

21:02:32 <ndw> an even more special meaning of "works", DanC :-)?

21:02:38 <DanC> :)

21:03:25 <tim142> phpicalendar has a timezone bug.

21:03:29 <DanC> i.e. if phpiclandar really works out of the box, very well. but you checked in into cvs, which suggests you're maintaining it.

21:04:00 <DanC> what it does is not that complicated. I'd rather rewrite it in python and/or XSLT than maintain it in php

21:05:05 <DanC> phpicalendar also has problem with multi-day untimed events

21:05:16 <tim142> I thought that was a space intto which I could check out a veriosn to run for real life. I made it team accessible only. I use it in real life. i didn't relize that that URI pre4fix was spoken for.

21:05:47 <DanC> yeah, you use it in real life, and somehow support requests go to me.

21:06:04 <DanC> who told Ivan that I had anything to do with phpicalendar? guilt by association, I guess.

21:06:11 <jimH_>http://wonderweb.man.ac.uk/owl/javadoc/

21:06:11 <dc_rdfig> F: http://wonderweb.man.ac.uk/owl/javadoc/ from jimH_

21:06:30 <jimH_> F:| OWL API from Manchester (Javadoc)

21:06:30 <dc_rdfig> Titled item F.

21:07:07 <DanC> greek spam. wow.

21:08:13 <DanC> (a) wow that the greek letters actually show up on my linux box, and (b) wow that somebody thinks it's cost-effective to spam me in greek

21:08:46 <ndw> Someone thinks it's cost-effective to spam me for a time machine parts, Greek seems positively mundane by comparison.

21:09:08 <DanC> in order for the greek letters to show up on my screen, (1) the spammer had to follow some pretty obscure RFCs for putting non-ascii text in email, rather than just doing whatever outlook likes...

21:09:14 <ghuo> yeah, the time machine one was pretty excellent

21:09:21 * jimH_ saw a great cartoon recently -- entitled "Where Spam comes from" and shows Osama Bin Laden sitting at a laptop saying "FIrst I will get them to question their manhood..."

21:09:28 <DanC> (2) evolution had to grok those RFCs, and (3) I had to have the relevant fonts installed and correctly configured.

21:10:27 * ndw is fighting a slightly borked machine today, having decided perhaps foolishly that unstable was, uhm, stable enough

21:11:12 <jimH_>http://www.cs.umbc.edu/~zding1/pub/HICSS37Final.pdf

21:11:12 <dc_rdfig> G: http://www.cs.umbc.edu/~zding1/pub/HICSS37Final.pdf from jimH_

21:11:26 <tim142> phpicalendar is gone

21:11:34 <jimH_> G:| (PDF link) A paper on a probablistic Extension to OWL

21:11:34 <dc_rdfig> Titled item G.

21:11:56 <jimH_> G: adds probability properties in various places, converts to Bayesian network

21:11:56 <dc_rdfig> Added comment G1.

21:12:36 <jimH_>http://km.aifb.uni-karlsruhe.de/owl/

21:12:37 <dc_rdfig> H: http://km.aifb.uni-karlsruhe.de/owl/ from jimH_

21:13:30 <jimH_> H:| Demo of a reasoner that uses horn logics to do OWL reasoning

21:13:30 <dc_rdfig> Titled item H.

21:29:00 <bitsko> ndw, or anyone: re. http://norman.walsh.name/2003/07/29/circle

21:29:35 <bitsko> how does "having a fragment identifier (#)" make a difference?

21:29:53 <mortenf> bitsko - don't go there! :)

21:30:47 <DanC> "weekly.pl is a Perl script generating an XHTML weekly view of an RDF calendar document." -- http://dev.w3.org/cvsweb/2003/weekly-view/

21:30:50 <bitsko> hopefully, my question is on a much smaller point than httpRange-14 as a whole :/

21:34:58 <ndw> if you use a fragid, then what you point to is defined by the spec that defines the media type of the representation you get back.

21:35:40 <ndw> ndw is now known as ndw^afk

21:39:03 <bitsko> ah. a very fine hair, I must say.

21:41:52 <DanC> re [[ Things that start ?http:? used to be called URLs. ]] they were 1st called UDI (universal document identifiers) then URIs, then URLs.

21:42:05 <DanC> cf www-talk archives from ~1991

21:42:21 <DanC> and IETF iiir archives

22:11:03 * bitsko better understands why mortenf said not to go there

22:11:11 <mortenf> :)

22:11:39 <danbri> <ndw> if you use a fragid, then what you point to is defined by the spec that defines the media type of the representation you get back.

22:11:51 <danbri> especially don't go asking what happens if the thing is content negotiable...

22:12:25 <mortenf> which i've just been doing...

22:15:50 <mortenf> hmm, wordnet:Camera is restricted to analog cameras - "with ... film"

22:20:19 <bitsko> I understand that FOAF decided to go with unambiguous properties instead of trying to "identify the person" with a URI. Are there any examples, proposed or otherwise, for [http:] URIs that are intended to identify a thing rather than a document about a thing?

22:21:25 <bitsko> uhm, outside of web-based documents themselves, if that helps

22:21:56 <mortenf> sorry, i won't join that discussion...

22:34:38 <dajobe> mdupont: I just added a very early csharp thing to redland

22:34:50 <dajobe> since I discovered the swig I have here has -csharp on it

22:40:17 <mortenf> on a related note, i tested the php interface the other day: http://www.wasab.dk/morten/2003/07/redland-test.phps

22:41:10 <dajobe> does it work?

22:41:10 <mortenf> as you can probably tell from what i've commented out, the language, datatype and xmlwf stuff doesn't seem to work (0.9.12)

22:41:20 <dajobe> I wouldn't be surprised

22:41:22 <mortenf> there's a .php as well

22:41:33 <dajobe> I take the Linus approach, it rots unless someone uses it

22:41:38 <mortenf> heh

22:42:01 <mortenf> the part that's active does work though

22:42:33 <mortenf> ... and implements a possible feature request: serialization grouped by subject

22:43:46 <dajobe> I'm likely to push most of that into C at some point

22:44:03 <dajobe> or into the parser, since it has lots of XML code anyway, get the escapes right

22:44:36 <mortenf> that would be nice, possibly as a flag or so, if there's a performance penaly (it's only really needed for XML processing and byte count...)

22:45:24 * dajobe stares dubiously at print(htmlspecialchars($xml));

22:45:34 <mortenf> :) - testing...

22:46:02 <mortenf> it escapes the big5