Experimental IRC log swig-2007-07-22

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.

12:17:33<ericP>phenny, tell me something good
12:17:35<phenny>Hey, I'm not as stupid as Monty you know!
12:17:37<phenny>ericP: 21 Jul 22:39Z <chimezie> tell ericP that if he is still interested in auto-generation of SPARQL facet XPath tests, I have a script (http://pygments.org/demo/400/?style=native) which I used to generate all the "core" permutations from the grammar (http://rafb.net/p/JZdi3Q54.html). It could be used to generate corresponding XPath patterns as well
16:20:39<tommorris>research.talis.com seems to be timing out on me, and it makes triplr run slowly when parsing an eRDF page.
16:21:21<tommorris>just in case any talis people are about - it would be quite useful. ;)
16:32:33<tommorris>Is there any formalised idea of what should be the subject of triples that have been GRDDLed?
16:33:06<tommorris>Or rather, where one should get the data from of what document is currently being transformed?
17:32:47<ericP>chimezie, around?
17:33:48<chimezie>yes, I'm here (sorta)
17:34:27<ericP>howdy
17:34:31<chimezie>tommorris: the subject defaults (by several paths) to the retrieval URI
17:34:42<chimezie>so this is the 'assumed' subject unless otherwise specified (in content)
17:34:46<chimezie>hey there
17:34:49<ericP>am thinking about how to lable exhaustive tests vs. those that test for existance of features
17:34:53<tommorris>okay,
17:35:21<tommorris>I've just published a GRDDL transform at http://rdf.opiumfield.com/grddl/nsfw.xsl
17:35:30<chimezie>the latter being those that only test syntax?
17:36:17<ericP>latter to make sure someone does UNION, but not necessarily UNION withing an OPTION in a GraphGraphPattern
17:36:51<tommorris>a GRDDL processor will presumably take back the RDF/XML I return, load it in to a memory model and then turn any triples with a subject of <> to the URI of the originating document.
17:37:34<chimezie>"regression" vs "unit" test?
17:37:36<ericP>ACTION would also like to talk about HCLS questionnaire, but afraid chimezie is already juggling enough conversations
17:38:18<chimezie>tommorris: yes, unless you have: @xml:base, /html/head/base, or the server returned a 30(2/3)
17:38:36<ericP>chimezie, dunno. my intuition was that unit test was more thorough than our what i meant by "feature" tests
17:38:38<chimezie>now is a good time, my rugrats are sleeping :)
17:38:45<ericP>sweet!
17:39:20<ericP>i was sorta talked out of s/integrate/coordinate/ on some, but here are the categories i have now:
17:39:32<tommorris>I hope the URL param in my XSL won't fsck anything up. Or would I be better served by removing it and letting the processor sort it out?
17:39:47<ericP>Integrate health care term sets (e.g. HL7, LOINC)
17:39:47<ericP>Integrate health care ontologies (e.g. SNOMED CT, openEHR, GALEN)
17:39:50<ericP>Coordinate healthcare rulesets with ongoing effort towards rule standards(e.g. LOA Medical Theories)
17:39:53<ericP>.
17:40:23<ericP>got a better idea than term set?
17:41:14<ericP>(if i understand from susie and you, HL7 and LOINC give coded terms for things, but no connection between them)
17:41:28<chimezie>ericP: yes!
17:41:49<chimezie>term sets actually is much better than controlled vocabulary, I think
17:42:42<chimezie>LOINC, CPT, and such are mostly flattened sets of coded terms, there is no 'formal' connection
17:42:49<chimezie>they mostly drive pick lists
17:43:03<chimezie>that phrasing above works
17:43:32<chimezie>ACTION points tommorris to jeremy carrols excellent work on XSLT modules for GRDDL processing: http://www.w3.org/2001/sw/grddl-wg/library
17:43:58<tommorris>thank you.
17:44:06<chimezie>np
17:44:16<ericP>chimezie, final check of http://www.w3.org/2007/06/HCLSForm#healthCare ?
17:44:39<chimezie>okay
17:45:29<chimezie>looks good =)
17:47:08<chimezie>in particular http://www.w3.org/2003/g/xml-attributes was meant to be imported to handle situations such as BaseURI processing
18:27:40<ericP>what's an RDF type for a project (like a software project)?
18:28:04<tommorris>ericP: DOAP - Description of a Project
18:28:13<ericP>aaahya, tx
18:28:21<tommorris>see http://usefulinc.com/doap/
18:29:52<ericP>cool, schema is poly-lingual!
19:55:28<recluxus>hi ihave to write a xml schema defining email data. i have "body may contain arbitrary markup from any XML namespace" , what's that suppose to mean ?
20:05:11<ericP>maybe it's considered self-describing (agent will recognize certain xml namespace/localname root pairs), or maybe it relies on auxilliary data, like a mime-type, to determine how the body should be interpreted
20:12:56<tommorris>or it could be a CDATA section - like how we put tag soup HTML into Atom/RSS etc.
20:17:32<c>http://bigdata.cvs.sourceforge.net/bigdata/bigdata-rdf/src/java/com/bigdata/rdf/
21:54:16<tommorris>http://tommorris.org/profiles/nsfw
21:54:34<tommorris>B|:Not Safe For Work (Tom Morris)
21:54:58<tommorris>B:|Not Safe For Work (Tom Morris)
21:55:06<tommorris>never remember the syntax for titles. :(
21:55:11<tommorris>B: A GRDDL profile that lets you mark up links as not safe for work.
21:55:34<tommorris>B: GRDDL allows you to resolve 'cross-cultural' issues too - my 'not safe' may be different from yours.
22:33:07<tommorris>danja_: research.talis.com is not responding.

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