Experimental IRC log swig-2007-09-24

Available formats: content-negotiated html turtle (see SIOC for the vocabulary)

Back to channel and daily index: content-negotiated html turtle

These logs are provided as an experiment in indexing discussions using IRCHub.py, Irc2RDF.hs, and SIOC.

02:08:34<dajobe>sigh, where is the patch to rdflib to make it work properly with the grddl test runner
02:12:11<dajobe>http://lists.w3.org/Archives/Public/public-grddl-wg/2007May/0077.html
02:12:36<dajobe>A:|patch to rdflib to make it work with grddl testft.py (May 2007)
02:12:44<dajobe>A:patches rdflib 2.4.0 (Apr 2007)
03:13:26<chimezie_>A:I see the diff has been applied to trunk: http://svn.rdflib.net/trunk/rdflib/syntax/parsers/RDFXMLHandler.py
03:16:44<chimezie_>A:"r1136 | chimezie | 2007-05-31 17:22:58 -0400 (Thu, 31 May 2007) | 1 line - fixes to RDF/XML parser to better handle xml:base (in particular the source URI of the RDF sink is handed down and used in the absence of an xml:base)."
03:17:26<dajobe>if there was a working bug tracker, a viewcv view of the repository, a working SVN repo or a changelog, I might have been able to find it earlier
03:19:21<chimezie>ACTION wonders if there is any literature on reducing SPARQL to datalog queries for relational databases
04:40:26<dajobe>I give up, can anyone explain to me why the GRDDL loopx tests should return anything?
04:40:38<dajobe>starting at http://www.w3.org/2001/sw/grddl-wg/td/grddl-tests#loopx
04:42:00<dajobe>"For this particular test, an XML document is its own namespace document" => no it isn't
05:16:44<kasei>crontab -e
05:16:56<kasei>oops. wrong window.
05:48:09<dajobe>ACTION emails his cry for help to grddl-comments and calls it quits
09:45:21<xi_>ugh
09:45:46<xi_>the 'microformats' API that just got merged into firefox is a 146.24 KB patch
09:45:58<xi_>my jQuery RDFa discoverer/parser is abuot 10 lines of js
09:46:16<xi_> i guess they like to specal case their discovery for each format.. well then they dont have any other options
09:48:15<xi_>full of lines like:
09:48:23<xi_>if (((((propnode.localName.toLowerCase() == "abbr") || (propnode.localName.toLowerCase() == "html:abbr")) && !propnode.namespaceURI) || ((propnode.localName.toLowerCase() == "abbr") && (propnode.namespaceURI == "http://www.w3.org/1999/xhtml"))) && (propnode.getAttribute("title"))) {
09:48:28<xi_>ACTION rolls over and pukes
09:48:40<m94mni>eugh, really?
09:48:59<xi_>https://bugzilla.mozilla.org/attachment.cgi?id=279849
09:49:07<m94mni>well, all the easier to argue for RDFa / GRDDL
09:49:13<xi_>B: firefox 'Microformat API' rolled into trunk
09:49:30<xi_>yeah. whats so hard about realizing 'write the parser once - data infinitely extensible'
09:50:45<xi_> var dateArray = string.match(/(\d\d\d\d)(?:-?(\d\d)(?:-?(\d\d)(?:[T ](\d\d)(?::?(\d\d)(?::?(\d\d)(?:\.(\d+))?)?)?(?:([-+Z])(?:(\d\d)(?::?(\d\d))?)?)?)?)?)?/);
09:51:00<xi_>nice of them to tuck that regex instead of make JS's Date class _finaly_ support iso8601. what year is this?
09:52:45<Arnia>xi_: 'worse is better'... apparently
09:52:53<xi_> if (this.org && this.org[0]["organization-name"] && (this.fn != this.org[0]["organization-name"])) {
09:52:57<Arnia>Even when it is obviously a case of 'worse is worse'
09:53:17<xi_>sweet, unqualified unprefixed string representing organization-name, hardcoded into firefox source
09:53:22<Arnia>xi_: Is your RDFa jQuery API available anywhere?
09:53:25<m94mni>well "works is better", I suppose
09:53:33<xi_>Arnia: i need to update it to support datatyped literals and such
09:53:35<xi_>Arnia: then i'll mentino it again
09:53:47<Arnia>ah cool :)
09:53:59<xi_>Arnia: my big problem is the huge chasm between teh XHTML 'theoretical' world and the browser 'real world' where innerHTML doesnt even match up with what you sent in
09:54:39<Arnia>Ouch
09:55:26<m94mni>xi_: to be fair though, half that patch is just testcases
10:03:51<xi_>yah.
10:04:12<xi_>id still like a firefox w/o all the crap. like all this sqlitey bookmark junk
10:04:28<xi_>opera keyword indexes every visited page now, google style
10:04:34<xi_>and returns results inline in the address bar
10:04:59<xi_>too bad it kills non GET/PUT/POST methods and it seems no way to bypass
10:05:12<xi_>what fun is REST if you have to tunnel an additional method selector inside POST..
11:46:53<chimezie>xi_: re: whats so hard about realizing 'write the parser once - data infiitely extensible' - what's hard is that it requires overcoming microformat zealotry =)
11:51:33<tommorris>chimezie: practical examples on the way.
11:52:32<tommorris>Hopefully, sometime today I'll release some ten line Python scripts that show how simple RDF parsing can be. I'm bloody fed up with the "RDF tax" myth - that RDF is *so* complex that us *mere mortals* can't understand it.
11:54:20<m94mni>that's the spirit
11:57:04<tommorris>Sorry, just been arguing with a borderline troll about the Semweb
11:57:17<tommorris>He seems to think that the SemWeb was an idea conjured up a good two years after RDF.
11:57:31<m94mni>rather than in 1989 :-)
11:57:51<tommorris>Well, I pulled up timbl's WWW94 slides and posted a link. See what happens.
11:58:09<tommorris>He is also part of the RDF == XML camp
11:58:43<tommorris>Anyone know what the best practice is to represent an integer literal age is in a foaf profile?
11:59:51<TedThibodeauJr>convert and store it as dob with day=month=0 ?
12:00:09<tommorris>Hmm. XSLT. ;)
12:00:25<tommorris>foaf:age seems to have been pulled from the latest version of FOAF
12:00:28<TedThibodeauJr>(without guaranteed date-of-triple, integer age will guaranteed be incorrect/changed in a year...)
12:01:01<tommorris>Well, it's being pulled from an API, so the database should change it in line with their DOB
12:01:15<tommorris>Meaning it might be incorrect for 60 minutes while my cache updates.
12:01:34<TedThibodeauJr>oh, I see ...
12:02:07<tommorris>I'll just use foaf:age, despite it being deprecated in the latest foaf.
12:02:45<TedThibodeauJr>can't find why it was deprected?
12:04:17<tommorris>Probably for the quite reasonable reason you just provided.
12:05:40<m94mni>and no replacement giving date of birth?
12:07:29<tommorris>Yep, foaf:birthday
12:07:54<tommorris>But I can't infer yyyy-mm of birth from age, alas.
12:08:30<tommorris>http://lists.foaf-project.org/pipermail/foaf-dev/2005-March/008054.html
12:08:38<tommorris>C|: Birthday Issue
12:08:45<tommorris>C:|Birthday Issue
12:09:30<tommorris>C: Mailing list post about the birthday issue in FOAF.
12:09:55<TedThibodeauJr>does foaf:birthday only have hasValue? or can it be set isGreaterThan/isLessThan ? (depending on data source, you could infer that...)
12:10:05<TedThibodeauJr>*thinks* *clicks*
12:10:37<TedThibodeauJr>also, if you're pulling age from an API, does that API also have a dob call?
12:17:59<m94mni>tommorris: http://vocab.org/bio/0.1/#sec-examples
12:18:21<tommorris>TebThibodeauJ: No, alas not.
12:19:20<tommorris>m94mni: Thanks. I'm just using foaf:age though
12:19:55<m94mni>it's certainly not good practice to use a deprecated property :-)
12:20:12<TedThibodeauJr>clearly the API is broken. ;-)
12:22:21<TedThibodeauJr>(especially because given a year, you could learn the birthDate exactly)
12:24:17<TedThibodeauJr>but now I must play in traffic ...
12:30:38<facefaceface>hello!
12:43:46<tommorris>Just for context, I'm working on mapping last.fm to RDF using XSLT.
12:43:55<Arnia>Dynamism? Who needs dynamism in a perfect shiny model-theoretic world?
12:44:23<m94mni>tommorris: nice
12:45:01<m94mni>tommorris: why don't you ask for advise on the FOAF list as well?
12:45:38<tommorris>I will when I come across problems that are more complex.
12:45:59<m94mni>they'd be interested in knowing about your project, I'm sure
12:46:00<tommorris>Song plays should be an interesting one.
12:46:12<m94mni>http://lists.foaf-project.org/mailman/listinfo/foaf-dev
12:46:21<tommorris>A bit o' luck and I'll have something up before this evening
12:46:26<m94mni>D:| FOAF mailing list
13:31:53<tommorris>See: http://rdf.opiumfield.com/lastfm/friends/tommorris and http://rdf.opiumfield.com/lastfm/profile/tommorris
13:33:15<m94mni>tommorris: nice
13:35:07<tommorris>When I get around to marking up the music, it'll follow the Linked Open Data principle and link through to MusicBrainz!
13:36:18<Arnia>tommorris: going to use MeNow?
13:36:39<tommorris>Hmm. Maybe.
13:36:49<tommorris>Just looking at the examples
13:37:10<tommorris>If I'd known about MeKnow a few months ago I'd have plotted Twitter to it.
13:38:56<tommorris>I'll update the Twitter script to support MeNow
13:42:52<m94mni>tommorris: have you seen http://openfriendformat.com/
13:43:15<tommorris>Yes. Seems like a fancy way of saying "use FOAF"
13:43:32<m94mni>http://openfriendformat.com/foaf/
13:43:54<m94mni>E:|OpenFriend FOAF generator
13:44:53<m94mni>(taken from discussions on FOAF list)
13:45:06<m94mni>example: http://openfriendformat.com/foaf/gen.php?account=chrisvr&avatars=true&network=all
13:46:45<m94mni>(it crashed on tommorris :-/)
13:47:34<tommorris>Probably the MySpace parser.
13:48:03<tommorris>Oh, oaky, it's doing it for last.fm
13:48:16<tommorris>Hmm. MySpace is identified as an rss:channel.
13:48:58<m94mni>it tries to do quite a few of them at a time, see the dropdown at http://openfriendformat.com/foaf/
13:51:06<m94mni>tommorris: link to discussion http://lists.foaf-project.org/pipermail/foaf-dev/2007-August/008656.html
13:52:49<tommorris>Oh, okay, that explains it (despite top-posting). ;)
13:54:25<m94mni>indeed
14:50:39<rawles>does anyone know about approaches to ontology merging? are there accepted good solutions, or is it something of a 'black art'
14:52:06<dajobe>the second one
14:52:38<rawles>so if i wanted to try to develop some strategy for it, where should i start looking? or is it just as good to try to define my own strategy?
14:53:11<dajobe>best review the literature first, ISWC would be a good start
14:53:26<rawles>i've been reading about SMART
14:53:31<rawles>but that's quite old now
14:58:21<gromgull>ontology merging is hopelessly domain dependent... take your own domain and hack something togehter
14:58:32<gromgull>assuming you just want a solution now... if you want to do some science I can't help you
14:58:34<gromgull>:)
14:59:13<rawles>yeah thats it, i just want a solution now... i'll do the domain dependant thing!
15:00:57<Arnia>Meh, ontology merging isn't something you can give a general solution to in a boolean semantics
15:01:04<Arnia>Not enough space
15:01:49<rawles>makes sense now you say it
15:02:02<rawles>this ontology stuff is all a bit cool and murky
15:19:48<jimbojw>rawles: i'm new to ontologies, but it would seem to me that merging them would be akin to merging say two XML files - you only know when to overwrite vs append values based on context
18:08:18<danja>isn't there any swig at this years TP? http://www.w3.org/2007/11/TPAC/
18:08:20<phenny>danja: 22 Sep 12:44Z <sbp> tell danja heh, I don't think I'd seen the "and getting it wrong!" bit before
19:24:22<DanC>danja, I remember some discussion of whether SWIG would meet this year's TP. I don't recall the conclusion.
19:24:24<phenny>DanC: 15 Sep 17:54Z <sbp> tell DanC http://www.w3.org/2001/sw/grddl-wg/ needs updating--shows the specs at Proposed Recommendation (unless I've got a stale mirror)
19:24:58<DanC>sbp, hhalpin just updated /sw/grddl-wg/ , I thought
19:25:23<sbp>yep. belay that order, &c.
19:26:23<DanC>hmmm... "and getting it wrong!" sounds familiar; is that quoting me?
19:28:38<sbp>well remembered. from a recent dannyayers.com post
19:28:58<sbp>(whose URI I would paste, but I seem to be having network woes)
19:30:41<sbp>ah, at last. http://dannyayers.com/2007/09/22/connollys-bane
19:50:39<tommorris>Note: I share DanC's pain at computers not making my life as easy as they should.
19:51:04<sbp>it's a universal pain, I'm sure
19:51:35<tommorris>That said, I find that now we have reasonably nice scripting languages (Python, Ruby etc.) a lot of that pain goes out the window.
20:51:42<DanC>@prefix foaf?
20:51:46<DanC>.g foaf namespace
20:51:49<phenny>DanC: http://xmlns.com/foaf/0.1/
20:57:58<MFen>.g rdfs namespace
20:58:01<phenny>MFen: http://www.dajobe.org/2004/01/turtle/tests/rdfs-namespace.out
20:58:16<MFen>@prefix rdfs?
20:58:27<MFen>hmm
20:58:30<MFen>well, THAT wasn't right
20:58:36<MFen>.g rdf namespace
20:58:39<phenny>MFen: http://www.w3.org/TR/REC-rdf-syntax/
20:58:49<MFen>that isn't right either. what the heck
20:59:27<chimezie>.nsBound rdf
20:59:28<Emeka>$rdf = http://www.w3.org/1999/02/22-rdf-syntax-ns#
21:04:37<tommorris>http://tommorris.org/blog/2007/09/24#When:21:03:04
21:04:51<tommorris>F:|Simple RDF Querying with Python (Tom Morris)
21:05:41<tommorris>F: How to query RDF using Python's RDFLib. Specifically oriented towards people writing small-scale mashups who get excited by the idea of creating stuff in less than ten lines.
21:48:14<DanC>very nice... http://www.semergence.com/2007/09/23/qotd-5/
22:03:03<drobilla>anyone use fact++? particularly anyone who knows an easy way to make it do what it does not involving 900 other technologies...

Back to channel and daily index: content-negotiated html turtle