Semantic Web Interest Group IRC Chat Logs for 2001-07-23

00:24:28 <GlobalMessage> [Global] We're about to split two major hubs in an attempt to reduce the amount of lag experienced on the network.

00:29:08 <GlobalMessage> [Global] All done. Thanks for your patience.

04:32:11 <AaronSw>http://www.markbaker.ca/2001/07/DeRPC/

04:32:11 <dc_rdfig> A: http://www.markbaker.ca/2001/07/DeRPC/ from AaronSw

04:32:46 <AaronSw> A:|DeRPC: Mark Baker volunteers to redefine RPC systems in the application semantics of HTTP

04:32:47 <dc_rdfig> titled item A

04:33:22 <AaronSw> A:It's always seemed to me that if we're going to be working on the Semantic Web, it's probably a good idea to be sure that we retain the semantics in things like HTTP, and HTML.

04:33:22 <dc_rdfig> commented item A

04:33:49 <AaronSw> A:The prevalence of throw a SOAP/XML-RPC message in an HTTP POST doesn't seem to be helping matters.

04:33:49 <dc_rdfig> commented item A

05:09:50 <ajbusy> ajbusy is now known as ajmitch

05:39:19 <ajmitch> ajmitch is now known as ajbusy

05:56:59 <ajbusy> ajbusy is now known as ajmitch

06:44:34 <ajmitch> ajmitch is now known as ajbusy

12:57:57 <hollawyer> hello

12:58:33 <hollawyer> need some help pls

13:00:25 <libby> hi

13:00:31 <libby> what about ?

13:03:17 <hollawyer> ssl on apache

13:03:50 <hollawyer> i am using openSa, but i guess it doesn't matter.

13:04:02 <hollawyer> i just can't set up ssl pages

13:04:23 <hollawyer> do i have to make some seperate .conf files?

13:04:25 <libby> sorry - I don't know how to help you there

13:04:36 <hollawyer> well ok thx

13:04:55 <libby> this might not be the best channel for those types of questions

13:05:15 <hollawyer> i just searched the net to find channels

13:05:15 <libby> (i don't know where you'd go though)

13:05:23 <hollawyer> whats this channel about?

13:05:44 <libby> the 'semantic web' and specifically RDF

13:06:10 <hollawyer> oh, hehe, i thought it has something to do with apache ;-)

13:06:23 <libby> nah.... ;-)

13:06:36 <hollawyer> ok than never mind, bye

13:06:49 <libby> see ya

13:48:15 <ajbusy> ajbusy is now known as ajzzzz

13:50:28 <danbri> anyone seen brian today?

13:54:26 <DanCon> I've seen email from him

13:55:07 * DanCon enjoys a spam-free morning, thanks to gerald's whitelist-based spam filtering http://impressive.net/people/gerald/2000/12/spam-filtering.html

14:00:18 <dajobe> DanCon: did you get the updated ntriples2kif I sent you or was that lost?

14:02:40 <dajobe> AaronSw: you asked for that file to be updated

14:03:01 <AaronSw> what file?

14:03:19 <dajobe> ntriples2kif.pl

14:09:46 <DanCon> dajobe, got it.

14:21:58 * danbri grumbles about http://www.ukoln.ac.uk/metadata/resources/dc/dc-xml-guidelines/

14:22:34 <danbri> what could've been a really useful doc just seems sheer bloodyminded.

14:22:35 <danbri> [[

14:22:35 <danbri> Note: It is anticipated that descriptions will be encoded within a container XML element of some kind. This document makes

14:22:36 <danbri> no recommendations for the name of any container element, nor for the namespace that the element should be taken from.

14:22:36 <danbri> Candidate container element names include <dc>, <dublinCore>, <description> and <metadata>.

14:22:36 <danbri> ]]

14:23:01 <danbri> apart from that, files written to these guidelines could pass for rdf.

14:23:10 <AaronSw> oh geez, no wonder people were so upse about it

14:23:15 * dajobe more than grumbles about it

14:24:00 <danbri> "Guidance for implementing Dublin Core in Unicode"?

14:25:40 <AaronSw> [[[This document does not consider the encoding of Dublin Core in RDF/XML. Furthermore, this document does not take a position on the relative merits of encoding metadata in 'plain' XML rather than RDF/XML. Mechanisms for encoding Dublin Core metadata in RDF/XML are available elsewhere.]]]

14:25:50 <AaronSw> I always thought the Dublin Core plan was to tell everyone to use RDF...

14:26:40 <AaronSw> Hmm, doesn't Andy understand that you can embed RDF in XML?

14:28:45 <danbri> He's responding to the need (apparently) widely felt to do a "plain" version of DC in XML, without all this RDF cruft.

14:28:52 <danbri> I thought Dave's doc showed that...

14:28:57 <dajobe> which ammounts to about 30 chars

14:29:03 <danbri> yeah

14:29:11 <dajobe> his xml schema "cruft" is bigger

14:29:49 <danbri> I talked to Makx about the status of dave's doc btw, and its going through DC AC at same time as the DC in RDF spec. So it vanished for a while but is back on (some kind of) track.

14:30:28 <dajobe> would be nice if they did this properly rather than via "my mate" Dan, who happens to be on the right DCAC mailing lists

14:31:40 <danbri> Is em around?

14:31:41 <danbri> em?

14:37:55 <danbri>http://209.198.125.242/itdsw/default.asp

14:37:55 <dc_rdfig> B: http://209.198.125.242/itdsw/default.asp from danbri

14:38:17 <danbri> B:|"Welcome to Intellidimension's Semantic Web Base"

14:38:17 <dc_rdfig> titled item B

14:38:36 <danbri> B:"This site serves as a public demonstration

14:38:36 <dc_rdfig> commented item B

14:38:36 <danbri> of RDF Gateway and as a testing ground for Semantic Web ideas in general

14:38:41 <dajobe> hey Jim

14:38:41 <libby> I couln't fing the de,mo link at all!

14:38:45 <danbri> B: of RDF Gateway and as a testing ground for Semantic Web ideas in general

14:38:45 <dc_rdfig> commented item B

14:39:05 * JHendler howdy all

14:39:15 <JHendler> JHendler is now known as JimH

14:40:34 <libby> hi jimh

14:41:03 <AaronSw> It's a little sad when one of the commonly used classes is http://www.w3.org/TR/WD-rdf-schema#Class

14:41:48 <AaronSw> We should tell those folks to shape up their act...

14:42:05 <danbri> there were only 5 uses of it

14:42:18 <AaronSw> that's even more sad... ;-)

14:42:51 <danbri> hmm, i thought your usual answer is "oh we'll just write some inference rules to map across"

14:43:03 <AaronSw> oh we'll just write some inference rules to map across

14:43:17 <AaronSw> but, really, it indicates that some folks are really out-of-touch

14:43:29 <AaronSw> which is important to keep in mind...

14:43:35 <dajobe> but is that us or them? :-)

14:44:09 <AaronSw> hmmm... if it's us....hey, i know dajobe, you can staff the RDF-mobile!

14:44:15 <danbri> libby, http://www.visionary2000.com/pilot/ -- did you say you've updated the .pdb to hybrid convertor?

14:44:57 <libby> yeah

14:45:09 * danbri is looking for a timezone-sensitive version of palm datebook

14:45:22 <dajobe> I did a gentle XML/RDF comparison in my semantic web talk a while back; maybe we should be doing a strong sell - "dumb XML", "smart RDF". I don't want to annoy XML-people since XML is still great.

14:46:51 <libby>http://swordfish.rdfweb.org/calendar/import.html

14:46:51 <dc_rdfig> C: http://swordfish.rdfweb.org/calendar/import.html from libby

14:46:53 <AaronSw> i think maybe we should wait until rdf 2.0 before the strong sell

14:47:14 <libby> B:|Generate RDF from your Palm Datebook file - 'hybrid' RDF schema

14:47:14 <dc_rdfig> titled item B

14:47:33 <DanCon> B:is this tool invertable/lossless?

14:47:33 <dc_rdfig> commented item B

14:47:48 <libby> B: changed to use the hybrid RDF calendar schema

14:47:48 <dc_rdfig> commented item B

14:48:03 <libby> B:no, just a bunch of print statements

14:48:04 <dc_rdfig> commented item B

14:48:45 <libby> the timezone thing is a pai - you have to enter a tz id from a list, assuming you put all your data in in one timezone

14:52:23 * dajobe reports ntriples and daml:collections support now in redland

14:52:24 <libby> B:doesn't do anything with repeating events

14:52:24 <dc_rdfig> commented item B

15:00:36 <JimH> dajobe - way to go on ntriples and daml:collections...

15:01:31 <dajobe> just merging in daml:collections donation by Aaron Michal - working on DAML somewhere

15:02:07 <dajobe> oh, at intellidimension.com - how topical

15:03:22 <JimH>http://www.daml.org/PalmDAML/

15:03:22 <dc_rdfig> D: http://www.daml.org/PalmDAML/ from JimH

15:03:39 <JimH> D: Scrape DAML (RDF) KBs onto your Palm

15:03:39 <dc_rdfig> commented item D

15:03:55 <JimH> D: might be a good link up to the RDF Palm Calendars!

15:03:55 <dc_rdfig> commented item D

15:04:18 <libby> ooh maybe

15:05:00 <libby> I always thuoght that the palm daml stuff would be good for http://rdfweb.org

15:05:34 <libby> interesting both are applications that the palm already does - calendar and addressbook

15:05:43 <JimH> libby - yes - also, would be really great for wireless/PDA apps because you can browse metadata (esp. semantics) without having to donload web pages -

15:05:59 <libby> ooh yes

15:06:22 <libby> for slow wireless modems especially

15:07:48 <JimH> PalmDAML used to browse big web repositories (I use it for thesaurus stuff) - keeps the URI up top, so browse in semantics space - then double click and you're able to download - very nice for bandwidth

15:08:36 <JimH> D:|Palm DAML

15:08:36 <dc_rdfig> titled item D

15:09:30 <libby> is there an online version?

15:10:25 <JimH> not sure what you mean by "online version"

15:11:26 <libby> a thing which would generate a pdb for you to download.

15:11:56 <libby> I keep thinking about managemnet types who often have palms but won't run java

15:13:34 <JimH> we've not yet done that -- mainly because there's just not enough organized DAML/RDF stuff around that made it a big win -- but we got to go there...

15:17:24 <JimH>http://www.daml.org/tools/

15:17:24 <dc_rdfig> E: http://www.daml.org/tools/ from JimH

15:17:36 <JimH> E:|DAML tools (also of use to RDF)

15:17:36 <dc_rdfig> titled item E

15:18:33 <JimH> E: The Daml tools area expanded a lot lately thanks to the recet DARPA DAML meeting - tools of use to RDF/S types abound (and a few of the RDF tools have now been used by DAML folks)

15:18:33 <dc_rdfig> commented item E

15:19:26 <danbri> Is any RDF data not DAML data? Or vice-versa?

15:21:22 <JimH> DAML makes some assumptions about underlying "meaning" -- i.e. stronger semantic model than RDF for some of its constructs. So it depends on whether you mean "syntactically" or semantically -- certainly a lot of the tools work on both

15:21:55 <JimH> example - we use hyperDAML for DAML, but should work just fine for RDF(S) -- very nice for managing namespaces

15:22:19 <JimH>http://www.daml.org/2001/04/hyperdaml/

15:22:19 <dc_rdfig> F: http://www.daml.org/2001/04/hyperdaml/ from JimH

15:22:28 <danbri> Let's take "semantically"; are there cases where a document is DAML but not RDF (or v-versa)?

15:22:45 <JimH> F:|HyperDAML - like hyper mail for RDF files

15:22:46 <dc_rdfig> titled item F

15:22:56 <JimH> F: check it out -- suspect it would be of use to RDF hackers

15:22:57 <dc_rdfig> commented item F

15:23:21 <JimH> hmm, a DAML document is always RDF, not the other way round -- sort of like RDF and XML

15:23:59 <JimH> i.e. An RDF doc not using the DAML keywords won't get any benefit from the DAML semantics

15:24:05 <danbri> Can you give me an example of an RDF document that isn't a DAML document?

15:24:14 <danbri> what's a keyword?

15:25:25 <JimH> sorry -- casual usage. Just as RDFS adds special meaning to the term "Class", DAML adds meaning to some special terms like "restriction" and etc. So if those terms are used in the defintions the DAML semantics knows what to do with them

15:25:28 <danbri> If I use a property, say util:personalMailbox, and it is rdfs:subPropertyOf foaf:mbox, which is a property I happen to have defined as a daml:UnambiguousProperty, the DAML semantics are useful in instance data usingthe 1st property.

15:25:50 <danbri> yes; the DAML stuff can be some way away from the instance data though.

15:26:14 <danbri> Also, just because the data happens not to draw on the DAML value-adding stuff, do you want to say it "isn't DAML data"?

15:26:29 <JimH> yes - agreed, that's what I was grappling with. In a web of RDF/RDFS/DAML it's hard to know exactly how to define terms

15:27:03 <JimH> I was not trying to say "it isn't DAML data" -- I was driving at "it doesn't get any benefit from DAML " -- although as you point out it may be linked.

15:27:27 <danbri> So, all RDF data is DAML data; all DAML data is RDF?

15:27:57 <JimH> the problem is I agree with each part, but think there's an asymmetry not revealed in the English --

15:28:10 <JimH> all RDF data is DAML - is true, but in the sense of "compatibility"

15:28:20 <DanCon> I have some specific points re DAML/RDF:

15:28:40 <DanCon> in fact, an RDF document that doesn't use any DAML vocabulary may very well benefit from DAML semantics...

15:29:26 <DanCon> the sorts of documents that *don't* benefit from DAML semantics are documents that don't stay with the "every property is either a datatype property or an object property" constraint.

15:29:29 <JimH> How would you feel about "All RDF is RDFS, All RDFS is RDF"?

15:29:55 <DanCon> "All RDF is RDFS, All RDFS is RDF" absolutely. (though it's somewhat meaningless, technically)

15:30:35 <danbri> Feels like a category mistake; RDFS is a strategy for describing RDF vocabularies and some utility constructs for doing so.

15:30:42 <JimH> exactly - it's the need for that parenthesized remark that is the problem...

15:30:49 <danbri> good question...

15:31:24 <JimH> so DAML is a strategy for describing RDF (and RDFS) vocabs and some utility constructs for doing so...

15:31:46 <JimH> I sort of like that as an explanation of what DAML is

15:31:57 <DanCon> I guess I see it this way: the class of WFFs is the same for RDF, RDFS, and DAML. i.e. there's only one syntax to parse.

15:32:23 <DanCon> i.e. everybody can tell the subjects from the predicates from the objects.

15:32:32 <sandro> Paraphrasing Drew McDermott: DAML is a type system for RDF.

15:32:51 <DanCon> Then there's the question of what inferences are licensed by the use of what vocabularies. RDFS and DAML+OIL are points in this space of vocabularies.

15:33:23 <sandro> Writing in DAML can thus mean either: writing type declarations, or writing RDF which includes and conforms to type declarations.

15:34:06 <DanCon> yeah, the type system analogy fits pretty well.

15:36:34 <JimH> I agree type analogy works okay, but of course the real fun begins with mapping between ontologies -- which would sort of be mapping one type system to another, and that's going to stress the analogy -- but that's just my "mishugas"

15:38:12 <DanCon> one type system to another: no, that doesn't follow from the analogy. different DAML ontologies would be different sets of declarations written in the DAML type system.

15:39:11 <JimH> dancon - good point

16:01:02 <jang> to recap:

16:01:14 <jang> 1. we can subclass rdfs:Literal, eg: into xsd:Date

16:01:18 <jang> and use that as a range.

16:01:32 <jang> 2. the "what is 2001-09" question

16:01:34 * danbri and jang surface offchannel rdfs discussion

16:02:03 <jang> - the clue comes from the idea of slapping a rdf:type arc on an anonymous node with an rdf:value arc

16:02:16 <jang> in that a literal isn't always the same resource

16:02:21 <jang> (I think)

16:02:37 <jang> because a literal may get an implicit type if you do the subclassing and range thing

16:03:02 <jang> so "2001-09" might be an arithmetic expression at one point, and a date at another

16:03:10 <danbri> I'm not sure what you mean by "In that a literal isn't always the same resource"

16:03:16 <jang> in other words, instances of literals are the resources (I think)

16:03:27 <danbri> you mean different occurences of literals are shadowed by different 'wrapper' resources?

16:03:37 <jang> seems like that makes the most sense

16:04:02 <danbri> So how to couch this in terms of practical test cases?

16:04:08 <jang> otherwise it looks like the date of my birth is an arithmetic expression

16:04:26 <jang> ...which doesn't make much sense

16:04:37 <jang> this is because we're dealing with the unicode representation of different ideas

16:04:45 <jang> which, as (I think Ron) points out, is the problem area

16:05:06 <jang> test cases are more like use cases: eg, the subclassing of rdfs:Literal thing

16:05:18 <jang> (which is something I've done for the EASEL semantic registry)

16:05:46 <jang> it's less something you can throu at a parser and more an example with a "this is how this works" kind of thing

16:05:48 <danbri> Describing scenarios I can do; I wanted something I could do 'make test' over and get a bunch of pass/fails

16:05:59 <jang> it's not a pass/fail thing

16:06:11 <jang> it's a "what are the implications of literal is a resource"

16:06:24 <jang> which will impact on rdfs

16:07:17 <jang> and the big question is whether if we have, say, the range of bornOn as xsd:Date....

16:07:35 <jang> and the range of hasExpr as ArithmeticExpression

16:07:47 <jang> then whether ...

16:08:00 <jang> <jan> <bornOn> "1970-12-08" .

16:08:16 <danbri> One question: how many things of rdf:type xsd:Date do we consider there to be? multiple for 1970-12-08 or just the one?

16:08:19 <jang> <jans_hard_sum> <hasExpr> "1970-12-08"

16:08:25 <jang> refer to the same literal

16:08:53 <jang> it's a good question though

16:09:12 <jang> we'd like the xsd:Date occurrences of "1970-12-08" to compare equal

16:09:21 <jang> but does it make sense to have a date equal an expression?

16:09:45 <jang> or do we say "1970-12-08" is BOTH a date and an arithmetic expression,

16:10:03 <danbri> Did you see... [rummages]

16:10:03 <jang> making the tacit assumption that we've ellided a "is the unicode representation of" in there somewhere?

16:10:43 <danbri> ...my SW CG msg on the XML work in this area? defining URIs for different flavours of 'equals', w.r.t. unicode etc.

16:11:21 <danbri> see http://lists.w3.org/Archives/Member/w3c-semweb-cg/ -> http://lists.w3.org/Archives/Member/w3c-semweb-cg/2001Jul/0023.html

16:11:29 <danbri> (member only URL, sorry)

16:14:19 <libby> jang, have you done any querying using datatypes?

16:14:44 <jang> not yet really...

16:15:04 <jang> I can derive the datatype of incoming streamed RDF by referring to the schema element definition

16:15:05 <danbri> you mean based on which xml schema datatypes are from ordered value spaces?

16:15:12 <jang> so my literals all have types associated with them

16:15:21 <jang> but that's easel-only stuff, not generic RDF

16:15:36 <jang> 'though I've got some methods for "find me the literals of this type"

16:15:56 <libby> that's cool

16:16:18 <jang> it's ok, proved less useful than "find the the values at the end of this path" stuff

16:16:33 <libby> the squishql stuff andy seaborne and I've been doing is limited by java casts at the moment

16:16:59 <libby> it's nasty because it can be sql-implementation dependant

16:17:54 <jang> not a lot you can do about that, I think...

16:20:05 <libby> I guess so

16:30:23 <jang> home I think

17:04:09 <JimH> JimH is now known as Jah-nashua

17:24:11 <dajobe>http://www.redland.opensource.ac.uk/demo/

17:24:11 <dc_rdfig> G: http://www.redland.opensource.ac.uk/demo/ from dajobe

17:24:16 <dajobe> G:|Redland RDF demo

17:24:16 <dc_rdfig> titled item G

17:25:18 <dajobe> G:now with N-Triples e.g. [http://www.redland.opensource.ac.uk/demo?db=ntriples;parser=ntriples;command=print|example]

17:25:18 <dc_rdfig> commented item G

17:28:09 <dajobe> G:and daml:collection support e.g. [http://www.redland.opensource.ac.uk/demo?db=daml;parser=raptor;command=print|here]

17:28:09 <dc_rdfig> commented item G

18:10:40 <Jah-nashua> dajobe - checked out "G:" very nice, and thanks for including the daml stuff

18:11:22 <dajobe> no problem

18:14:16 <dajobe> dajobe has changed the topic to: RDF and Semantic Web Chat (home: http://www.w3.org/RDF/Interest/#irc blog: http://rdfig.xmlhack.com/)

18:54:22 <Jah-nashua> Jah-nashua is now known as JimH-NH

19:12:37 <leigh> evening all

19:13:07 <danbri> hello

19:13:36 <libby> hey leigh

19:13:45 <leigh> hi libby

19:14:05 <libby> how are you doing?

19:14:20 <leigh> not bad thanks

19:14:32 <leigh> so, whats this rdf-calendaring lark all about then? :)

19:14:49 * leigh is writing an article on the rdf-calendaring effort...

19:14:59 <libby> heh, erm

19:15:21 <leigh> I've done a quick surf through the archives, and collected some pointers

19:15:34 <leigh> would it be safe to say that currently the state of play is:

19:15:48 <leigh> -working on calendaring schema (based round iCalendar)

19:16:01 <libby> sure

19:16:02 <leigh> -some prototypes produced (e.g. dancs palm pilot doobrie)

19:16:06 <danbri> collected in an xml-deviantish sort of a way?

19:16:14 <leigh> yep, this weeks deviant

19:16:21 <danbri> cool :)

19:16:21 <libby> at the moment what we have is very closely tied to iCalendar

19:16:38 <libby> schema-wise

19:16:52 <danbri> some things aren't, eg danc's stuff tends to reflect out Palm's schema rather than map into iCal

19:17:03 <libby> DanC has been working on a schema in N3 for the Palm datebook format

19:17:06 <libby> right

19:17:14 <leigh> btw. is it ok to quote from here?

19:17:19 <danbri> it's public, yes

19:17:24 <libby> I guess - all logged anyway

19:17:29 <leigh> cool.

19:17:49 <danbri> archived an everything; you can cite discussions by url too, since dave's logger generates html anchors

19:17:58 <leigh> cheers, thanks

19:17:58 <danbri> and theres the http://rdfig.xmlhack.com/ blogs too :)

19:18:08 <dajobe> eek!

19:18:14 <libby> we have some prototypes for getting data from Palms, such as danC's, and my rather hacked about perl script

19:18:28 <danbri> eek? I said too much? thought you said the urls wouldn't break?

19:18:40 <libby> danC has been investigating rules for recurrance and suchlike in N3

19:18:42 <dajobe> :-) nah but they run on my box, which I have to move this week

19:18:45 <DanCon> that reminds me... I have been meaning to write rules to relate my palm schema to the vcard stuff... more relevant now that my palmpilot broke and my cell-phone groks vcalendar!

19:19:12 <danbri> "vcalandar" meaning icalendar?

19:19:17 * leigh adopts pseudo-journalist mode

19:19:26 <leigh> so how would you (all) describe the effort?

19:19:33 <DanCon> libby, have you hacked anything that takes RDF in the hybrid schema and spits out actual icalendar stuff that, say, korganizer would read?

19:19:53 <libby> nope

19:20:06 <libby> I guess wouldn't be too hard

19:20:16 <DanCon> describe: as I've said, the bane of my existence is doing things I know the computer could do for me. I'm having a great time getting the computer to figure out things about my schedule

19:20:51 <DanCon> e.g. I wrote a tool to convert semi-structured plain-text itineraries in the format that our travel agent spits them out into RDF...

19:21:17 <libby> leigh: I'd charaterise the effort as trying to get an RDF model for event data - specifically things like meetings and conferences - as quickly as we can

19:21:29 <DanCon> then I took two such itineraries, as RDF, and wrote some rules expressing constraints that my wife and I had agreed to, and I was able to get the machine to decide that one of the itineraries didn't meet our constraints.

19:21:54 <libby> at Danc's suggestion we've been trying to make data avilable inn the schema and write demos to test it on real data

19:22:19 <leigh> are you aiming for anything in particular. For example is this just an RDF

19:22:20 <DanCon> results: http://www.w3.org/2001/08swws67/itin1check.n3

19:22:21 <leigh> test bed?

19:22:32 <leigh> DanCon: thanks

19:23:00 <DanCon> aiming: I'm aiming at automaing my day-to-day decisions and queries. e.g. "I've got yet another invitation for tuesday at 2pm. Do I have any pre-existing obligations?"

19:23:04 <libby> I'd like to be able to export may data - all in different formats - to RDF, and then query it in one place

19:23:25 <libby> so not just a RDF testbed, but trying to get somewhere near solving practical problems

19:23:43 <DanCon> the reality is: my schedule info is never collected in one place. it's a multi-party (peer to peer) setup.

19:23:51 <libby> although of course useful adta for testing query stuff for example

19:24:58 <leigh> yep, thats what I thought.

19:25:06 <leigh> I find this posting interesting though:

19:25:11 <leigh>http://lists.w3.org/Archives/Public/www-rdf-calendar/2001Jul/0009.html

19:25:11 <dc_rdfig> H: http://lists.w3.org/Archives/Public/www-rdf-calendar/2001Jul/0009.html from leigh

19:25:33 <leigh> para beginning 'One main problem...'

19:26:00 <leigh> seems like calendaring might offer an interesting learning curve

19:27:01 <libby> sure, and at a very basic level, what different people call events is very different

19:27:30 <libby> it's difficult to see what lots of different sorts of wht we might agree are events have in common

19:27:59 <danbri> H:|An www-rdf-calendar discussion on the strengths/weaknesses of the proposed RSS 1.0 Event module

19:27:59 <dc_rdfig> titled item H

19:28:06 <libby> there's a lot of philosophical literature on what consititutes an event, that I've skimmed the surface of

19:28:37 * danbri just bought Davidson's Actions and Events; interesting but not something I'd want www-rdf-calendar folk to have to read...

19:28:54 <libby> we had an interesting discussion about flights for example, a few days ago, and another about the relationships between locations and datetimes

19:28:56 <leigh> re: discovering commonality among events; thats why I thought its an interesting RDF topic

19:29:03 <DanCon> re 0009: he makes a good point: calendaring seems simple from the perspective of any one party in the system. But it's very hard to model the union of everybodys perspectives.

19:29:46 <DanCon> I'm explicitly *not* trying to model the union. I modelled my palmpilot, and I modelled the data we get from our travel agent. Then I wrote rules to map them: http://www.w3.org/2001/07dc-bos/itin2datebook.n3

19:29:59 <leigh> DanCon: semweb in miniature?

19:30:23 <DanCon> miniature?

19:30:41 <leigh> well, you're exploring issues that will be repeated elsewhere

19:30:48 <leigh> 'union of perspectives'

19:31:25 <DanCon> I'm not following

19:31:28 <leigh> libby: thanks, I'll look up that discussion

19:31:30 <libby> .some discussion here:http://ilrt.org/discovery/chatlogs/rdfig/2001-07-18.txt

19:31:32 <danbri> Leigh, here's why the calendaring work appeals to me: its a very practically grounded (palm pilots; meetings etc) area that shows some of the potential strengths and challenges of the Semantic Web. Specifically, the need for calendar/schedule data to be intermingled with other related information, eg. RSS for syndication, DC for document metadata, white pages info etc.

19:31:55 <leigh> what he said!

19:31:58 <leigh> :)

19:32:19 <danbri> The sorts of things I want to do with calendaring data usually require me to draw on other sources of info at same time; eg. who else is attending a meeting, what the required reading was.

19:32:53 <libby> jan grant has some interesting duiscussion here http://ioctl.org/jan/cal/critique

19:33:15 <danbri> iCalendar, vCard etc sort of live in little islands; RDF's grand claims tend to be that it builds up some commonality between these islands. www-rdf-calendar is a group trying to find out if we can live up to the rhetoric...

19:33:16 <DanCon> intermingled: exactly. note that the result of mapping from travel itinerary to palmpilot doesn't lose the travel itinerary info. it's all just merged together.

19:33:55 <libby> a really important problem is getteing that data out there

19:34:01 <leigh> libby: thanks

19:34:08 <libby> that's why the palm conversion tools could be so useful

19:34:22 <DanCon> did anybody see the little 30 line ical->xml hack perl script I chumped yesterday to look at the data from my cellphone? I'm interested to see an actual spec for ical so I can see how far my hack should work.

19:34:23 <danbri> It's also a good testbed area for the specs: people have long asked how we should deal with XML Schema datatypes in RDF; some of the discussion on xmlschema-dev and the rdf cal list gets (at last) stuck into the detail.

19:34:42 <danbri> I saw it; and went so far as downloading the rpm you referenced before realising i wasn't quite sure what it did.

19:34:48 <DanCon> rpm???

19:35:04 <libby> that's alos why it would be good to get RSS 1.0 event module people on board

19:35:08 <danbri> some tool you used, on sourceforge

19:35:32 <libby> I've a program somewhere that a friend of Greg fitzpatrick wrote which does icalendar -> xml

19:35:34 <DanCon> oh... the talk-to-the-cellphone tool.

19:35:40 <leigh> libby: so you can piggy back off that infrastructure?

19:35:51 <libby> sure

19:35:54 <libby> yes

19:36:26 <libby> and also - as danC suggests, create RDF files from existing icalendar files - iCalendar is used in many apps (including palms)

19:36:32 <danbri> i have a bunch of phone numbers in my Motorola/Timeport 250e, and much the same in my Palm... Seems a shame.

19:36:45 <libby> it's so annoying

19:37:02 <libby> but the key then is - as you've saif before dambri - to get the data off these devisces.

19:37:07 <leigh> so whats the best way someone can get involved initially? data conversion tools?

19:37:23 <libby> that's where proprietory tools - howevere good they are - often fall down

19:37:33 <DanCon> my new nokia is very happy talking to various open-source tools on my vaio laptop. such a pleasant surprise after all that proprietary crap with the qualcomm/sprintpcs setup!

19:37:53 <libby> sorry about the spelling

19:38:02 <libby> It's a good way, yep.

19:38:29 <DanCon> danbri, never mind the obex tools... the 30 line perl gem is this: http://dev.w3.org/cvsweb/2001/palmagent/vcal2xml.pl?rev=1.1&content-type=text/x-cvsweb-markup

19:38:30 <libby> I think DanC's right about testing the schema - which is not by any means complete - by creating real data

19:38:32 <leigh> I saw a mention that you were working towards a suggested schema by end of this month,

19:38:36 <leigh> that still on the cards?

19:39:00 <libby> it's difficult - they guy who did a lot of the work, micheal arick, is away at the moment

19:39:05 <DanCon> the use-case that I (we?) are working toward is an event in San Jose at the end of the month, yes.

19:39:23 <libby> we have a strawman schema, and we're working on that

19:39:39 <libby> 'suggested' is knind of vague I guess

19:39:51 <libby> I think I underestimeated how much testing needed to be done

19:40:04 <DanCon> my contribution to the game: http://www.w3.org/2001/08swws67/

19:40:20 <leigh> yes, I saw danbri's clarification that there's no 'official' effort here

19:40:59 <libby> the didfficulty with things based on the icalendar work is that it is very comprehensive, veryt complex

19:41:07 <libby> much too much so for most people

19:41:23 <leigh> hmmm, is the RDF any more accessible?

19:41:27 <libby> ...and the complexity is such that they are still testing it after 5 years

19:41:53 * danbri returns; not sure why I keep getting disconnected

19:41:55 <libby> not in the current version

19:42:21 <libby> but I think there is room for us to suggest using simplified parts, for example by using different namespaces

19:42:29 <libby> I hope we can do this

19:42:30 <DanCon> er... danbri, the "Generate RDF from your Palm Datebook file..." item that you chumped... following that link puts me somewhere that doesnt' seem very relevant.

19:43:09 <libby> at the moment I atr least am trying to get an overall view of what calendaring encompassses, and the enormous effort put into iCal;endar is a very good place to start

19:43:17 * DanCon is looking for info about the hybrid schema...

19:43:45 <DanCon> the value of iCal, to me, is in the running code: in my cellphone, in korganizer, etc.

19:43:57 <libby> content: http://www.ilrt.bris.ac.uk/discovery/2001/06/content/ (not all is hybrid yet)

19:44:17 <libby> schema (annotated) http://www.ilrt.bris.ac.uk/discovery/2001/06/schemas/ical-full/hybrid.rdf

19:45:04 <libby> datbook ->hybrid http://swordfish.rdfweb.org/calendar/import.html

19:45:17 <libby> (working on s writeuup of the hybrid but not finished)

19:45:26 <leigh> ta

19:45:29 <DanCon> thx... that should keep me hacking for a bit...

19:45:40 <libby> 'hybrid because micheal arick and I simultaneously wrote a schema

19:46:14 <danbri> I'm interested in how this relates to the xml schema date/time stuff; which I don't know too well but remember being controversial and/or confusing

19:46:17 <libby> and he merged it together

19:46:52 <leigh> libby: I saw a reference to his critique somewhere I think

19:46:55 * leigh goes to check

19:47:28 <libby> sure - that was pre-merge

19:48:00 <libby> xml datatypes allows utc- or

19:48:10 <AaronSw> hello all

19:48:13 <leigh> where else would it be useful to scrape data from? Palm seems to be covered...

19:48:13 <libby> format for

19:48:22 <AaronSw> Hi Leigh, good to see you here!

19:48:23 <libby> date times

19:48:44 <DanCon> where else: oag.com

19:48:48 <leigh> hi aaron. I'm doing some research on calendaring for this weeks deviant.

19:49:09 <DanCon> where else: see other suggestions in my "todo list" message to www-rdf-calendar

19:49:16 <libby> whereas icalendar inly aloows free dloating, UTC or timezone properties

19:50:05 <leigh> DanCon: thx.

19:50:06 <danbri> H:Also interesting, [http://lists.w3.org/Archives/Public/www-rdf-calendar/2001Jul/0012.html|relationship to xml-schema datatypes], perhaps providing input into the [http://www.w3.org/2001/sw/RDFCore|RDFCore] work on schema and typing.

19:50:06 <dc_rdfig> commented item H

19:50:26 <libby> I'm trying to get conferemces and workshops to give me their data, and then when more settled, to outputt their data as rdf

19:51:11 <libby> edd helped me out with that for xmleurope, and stefan, for SWWS

19:51:29 <leigh> other suggestions: school timetables? address databases (for locations of events)?

19:51:59 <libby> ooh, greg fitzpatrick and skical - events like concerts, cinema, footie matches

19:52:07 <libby> railtrack

19:52:18 <DanC_> hmm... libby, didn't you already provide the SWWS event data in RDF? part of it, at least?

19:52:18 <leigh> that'd be a god-send

19:52:24 <libby> i think the locations stuff is extreemly interesting

19:52:31 <danbri> a uk reference! foreigners se http://www.railtrack.co.uk/ -- uk train times

19:52:54 <libby> sure, yep, on http://www.ilrt.bris.ac.uk/discovery/2001/06/content/ near the top

19:53:12 <leigh> (uk train times in an ideal world)

19:53:19 <danbri> Event discussions can be hard to scope, since they relate to attempts to describe anything that can happen!

19:53:27 <libby> icalendar has a geo property, which is supposed to have longitude and latuitude

19:54:16 <leigh> cool

19:54:18 <libby> I've been going a little work on trying to match people gegraphically as part of William loughborough's barnraising project, but not much to show yet

19:54:38 <libby> danbri's right - everything's connected

19:54:52 <libby> and it all starts to shade off into other things

19:55:13 <libby> yep, railtrack - pure fantasy

19:55:51 <leigh> there's a paper title there: 'rdf-calendaring, the interconnectedness of all things : a holistic approach' :)

19:55:56 <danbri> everything's connected: RDF's promise and curse...

19:56:05 <libby> skical were the first to start to put this sort of data in rdf

19:56:49 <libby> .http://lists.w3.org/Archives/Public/www-rdf-interest/2000Aug/0109.html

19:57:07 <libby> leigh ;-)

19:57:15 <danbri> RDF shows (I believe) that you can put all this data into a common model, and write rules, queries etc to exploit that. But we've a lot to figure out regarding practicalities of getting all those compatible ontologies agreed.

19:58:12 <leigh> thats why I (poorly) described it as 'the semweb in miniature' - that seems to be on of the key probs to solve

19:58:16 * leigh waves hands a lot

19:58:28 <danbri> leigh, libby and me are funded in part through harmony project, hence http://metadata.net/harmony/Publications.htm has pointers to some event-basd modelling we're involved in

19:58:40 <libby> ooh yes

19:58:49 * DanC_ starts to make http://www.w3.org/2001/08swws67/ by adding a HEADER.html ...

19:58:59 <DanC_> make it readable, that is...

20:00:01 <libby> the model is the tricky bit for events

20:00:38 <danbri> I'm not totally sold on the current ABC model, http://metadata.net/harmony/ABCV2.htm but the original draft i'm still fond of: http://www.ilrt.bris.ac.uk/discovery/harmony/docs/abc/abc_draft.html

20:00:52 * danbri should make time to writeup his worries about V2 instead of grouching in irc

20:01:01 <libby> like, place-times need locations, at least in the broad sense of timezone

20:01:24 <libby> not all events have a location; often the particpants are at different location

20:01:35 <libby> s, like teleeconferencing

20:01:52 <libby> some events (like flights) don't have one location

20:02:09 <dajobe> (like this event)

20:02:11 <leigh> hmmm. but the conference room can be assigned a URI

20:02:15 <leigh> virtual or no

20:02:20 * sandro wonders who dingbat_hp (on slashdot) is. (an RDF supporter who works at HP Labs in Bristol; I'm guessing they're here....)

20:02:31 <libby> sure

20:03:20 <libby> but it would be good to make data from many sources compatible, i.e. not rely on identifiers for, say, events that are centrally providede

20:03:41 <leigh> centrally provided?

20:04:50 <libby> well, syncing is easy if both devices know about the event identifier, but if not, it's hard

20:05:10 <libby> (although some things will need identifiers)

20:05:35 <leigh> I'm confused. How could you do away with identifiers?

20:06:14 <libby> sorry, you couldn't. but as regards idetifiers for the events themselves, you might not know the official id for a conference talk for example

20:06:31 <danbri> Individuating events is a really hairy problem; identifying people/organisations etc is much easier. Specific types of events (eg. meetings) you can work around, since they're not so nebulous. But merging descriptions of arbitrary events is too hard.

20:06:59 <libby> I've been wondering if queries can help idetify types of events, see http://www.ilrt.bris.ac.uk/discovery/2001/05/rdf_cal/slide5.html

20:07:25 <DanC_> re individuating events: it's hard/impossible in general, but straightforward in many cases that matter. I made a rule just yestarday that allows me to recognize when 2 W3C telcons are the same:

20:07:48 <DanC_> # implicit in W3C proces...

20:07:48 <DanC_> # Remote meetings happen at most once per day.

20:07:48 <DanC_> { :thisMeeting a org:RemoteMeeting;

20:07:48 <DanC_> is m:meeting of :grp;

20:07:48 <DanC_> m:date :str.

20:07:49 <DanC_> :thatMeeting a org:RemoteMeeting;

20:07:51 <DanC_> is m:meeting of :grp;

20:07:53 <DanC_> m:date :str.

20:07:55 <DanC_> }

20:07:57 <DanC_> log:implies { :thisMeeting = :thatMeeting }.

20:08:50 <DanC_> I have similar rules for knowing that two groups are the same cuz they have the same contact:mailbox.

20:10:05 <leigh> aha, hence the comment about social practice (u derived rule from W3C process)

20:10:32 <DanC_> yup.

20:10:36 <leigh> cool

20:12:28 <libby> goeff chappell has done some interestig stuff with rules and query language with danc's calendar data

20:13:33 <leigh> got any pointers?

20:13:55 <libby> hang on

20:14:29 <danbri> danc re individuating events, yep: we can use rules to write down (and sign up to) our shared fictions about how we'll go about individuating events

20:14:43 <danbri> looking for the 'right answer' is philosophical rathole territory

20:15:19 <libby> . http://lists.w3.org/Archives/Public/www-rdf-calendar/2001Jun/0011.html

20:15:32 <leigh> ta libby

20:15:54 <libby> interesting approach

20:16:07 <libby> clever database, simple ql

20:16:33 <leigh> yeah I want to tinker with some ql stuff - thats why I grabbed squish

20:16:52 <leigh> I was going to write some junit test cases for it to get to grips with the code :)

20:17:01 <libby> ooh

20:17:04 <libby> do!

20:17:06 <danbri> leigh, Geoff was the guy who did the 'logic plugin' for Mozilla a couple years ago, http://www.mozilla.org/rdf/doc/inference.html (plugged a Prolog system in behind their RDF API)

20:17:21 <danbri> mmm, test cases...

20:17:45 <leigh> is 0.45 the most recent stuff?

20:18:25 <leigh> no promises mind. its just that I've caught refactoring in a big way and test cases are a must. trying to keep some time free for it

20:18:25 <DanC_> hmm... hybrid schema... does it have a dc:source link to the ical spec?

20:18:32 <libby> I think that simple ql like squish can stilll be usefiul - and should be - for calendar queries, although rules based stuff is much more powerful and therefore useful

20:18:39 <libby> no, sorry

20:19:03 <DanC_> danbri, on how to use cwm ala javascript-prolog: I can sketch that for you if you like.

20:20:11 <JimH-NH>http://ittalks.org

20:20:11 <dc_rdfig> I: http://ittalks.org from JimH-NH

20:20:23 <danbri> DId you see the followon discussion w/ Aaron yesterday; I think he cracked it for me

20:20:27 <JimH-NH> I:|DAML used for talk announcements

20:20:27 <dc_rdfig> titled item I

20:20:36 <DanC_> no, didn't see that danbri. Aaron++

20:20:44 <JimH-NH> I:A good source for dated events tagged in DAML

20:20:44 <dc_rdfig> commented item I

20:21:09 <JimH-NH> I: generated from a DB, but view source on the talks to see the DAML

20:21:09 <dc_rdfig> commented item I

20:21:37 <JimH-NH> I: tim Finin (site developer) says he would be willing to give access to DB to those doing calendaring

20:21:37 <dc_rdfig> commented item I

20:21:42 <danbri> see http://rdfig.xmlhack.com/2001/07/22/2001-07-22.html 'doing database style queries in cwm'

20:23:29 <libby> thanks!

20:23:39 <DanC_> libby, good detail work on swws2001-07-31.rdf . it seems you addressed all the details about people/places/things vs. their names.

20:24:14 <libby> thanks

20:26:20 <leigh> libby/danbri I saw comments that said you were looking for ways to make RDF querying accessible through JSP/JDBC apps. Any progress?

20:27:17 <libby> I've some demos, that's it so far

20:27:40 * DanC_ scans ical spec http://www.ietf.org/rfc/rfc2445.txt

20:28:07 <em> libby are you going to present any of you calendaring work at swws?

20:28:10 <libby> .http://swordfish.rdfweb.org/discovery/2001/07/swws/

20:28:23 <libby> er, if you like :)

20:28:40 <em> yes please! :)

20:28:55 <libby> ok then !

20:29:22 <em> i;d like to talk a bit about this in my talk, but reference some additional presentation/demo where people can get more info

20:29:23 <leigh> libby: neat

20:29:44 <libby> thanks

20:29:47 <libby> sure

20:30:21 <em> will you be planning on doing a demo of sort? have you contacted Martin Lacher?

20:30:40 <libby> er, no...

20:30:58 <em> [[[

20:31:00 <em> You are a company willing to sponsor this event or give a demo?

20:31:00 <em> Sponsoring companies are are given the opportunity to present their

20:31:00 <em> software in the demo session. Please contact Martin Lacher (lacher@db.stanford.edu).

20:31:07 <em> ]]] -- http://www.semanticweb.org/SWWS/

20:31:09 <libby> I didn't realize I had to - hadn't thought it through

20:31:26 <libby> h'm not really a sponsoring compnay ;)

20:31:44 <DanC_> ugh.. ical unfolding (section 4.1 Content Lines)

20:31:57 <em> yeah... but your highlighted on the page under Semantic Web Resources affiliated with SWWS

20:32:06 <em> so I think you qualify :)

20:32:10 <JimH-NH> wasn't clear if demos were only for the sponsors - think Stefan is willing to have some others shown - suggest you contact him ( stefan@db.stanford.edu)

20:32:36 <JimH-NH> btw, if not - DARPA is a sponsoring agency and would be pleased to have Libby demo in our slot

20:32:52 <em> basically, i was wondering if you'd be willing to talk more about this work... i think its quite excellent and i imagine people will want to know more

20:33:26 <em> thanks JimH-NH for the extended invitation! :)

20:34:21 <libby> aw thanks

20:34:22 * em notes the following "A bottle of champagne (from Stefan) for

20:34:22 <em> the first one who imports the data (automatically) into the calendar

20:34:22 <em> of MyYahoo or Outlook."

20:34:51 * em notes as well.. the topicmap demo seemed to be no-worky

20:34:54 <JimH-NH> wow - all Tim ever offers is a beer!

20:35:05 <em> :)

20:35:08 <libby> heh

20:36:18 <libby> I'll sort some demo stuff out. thanks :)

20:36:31 * AaronSw wonders when he can cash in his brownie points

20:36:41 <AaronSw> ;-)

20:36:47 <danbri> when you're 21!

20:37:13 * AaronSw moves to have the next F2F in a more lenient country

20:37:41 <leigh> 'bout time there were some UK events :)

20:38:09 <DanC_> yikes... some of this ical stuff looks pretty wierd, huh? e.g.

20:38:12 <DanC_> ATTENDEE;DELEGATED-FROM="MAILTO:jsmith@host.com":MAILTO:

20:38:12 <DanC_> jdoe@host.com

20:38:36 <danbri> uk events: hence http://www.yahoogroups.com/group/semanticweb-southwest/ :)

20:39:33 <leigh> cool. of course the trick would be to use RDF to schedule 'em :)

20:39:46 <libby> heh soon...

20:42:23 <libby> danc, yeah;

20:42:47 <libby> there's not a clear model in part, to me at least

20:43:34 <danbri> can the clear bits be put in a separate namespace, so folk can use the entire thing to accurately represent ical cruft, or a nice simple vocab for those without ical commitments?

20:43:44 <danbri> that sounded a bit rude; i'm sure they had their reasons...

20:44:36 * DanC_ grabs vcard/vcal sdk from http://www.imc.org/pdi/pdiproddev.html

20:44:58 <libby> i was wondering if they had a uml model somewhere

20:45:39 <libby> I could put the bits clear to me in a separate place maybe

20:45:42 * danbri heads off in search of a us bank account

20:45:58 <danbri> clear: that'd be interesting to see

21:32:37 <leigh> I'm afraid I've got to do (must sleep tonite :). Many thanks for all your comments, much appreciated. Look out for Deviant on Thurs.

21:32:41 <leigh> bye

21:55:01 <ChanServ> This channel is logged and blogged: http://logicerror.com/rdfIRCWelcome

22:15:28 <fitzix> hello

22:15:39 <sbp> Hi

22:16:00 <fitzix> sbp: Hey there, how are ya?

22:16:27 <AaronSw> hi there fitzix!

22:16:39 <fitzix> AaronSw: Hey man, did you go to the protest? :)

22:16:57 <AaronSw> I did indeed.

22:17:00 <sbp> Not bad, thanks

22:17:12 <fitzix> sbp: cool :)

22:17:15 <fitzix> AaronSw: How was it?

22:17:19 <AaronSw> protest: http://ufo.chicago.il.us/free-sklyarov/

22:17:24 <AaronSw> i thought it was a lot of fun, actuall

22:17:29 <AaronSw> got some good pictures of folks

22:17:30 <sbp> Did you turn over any cars?

22:17:44 <AaronSw> No, it was a non-violent protest

22:18:05 <fitzix> cool :)

22:18:11 <sbp> Aw, come on: it's not a protest if you don't turn cars over

22:18:22 * fitzix thinks all non-violent protests need a little violence to get them going...

22:18:28 <fitzix> j/k

22:18:29 <fitzix> :)

22:18:32 <sbp> :-)

22:19:03 <sbp> I was joking too (just about)

22:19:28 <fitzix> sbp: Sometimes, civil disorder can be violent without being unethical - sometimes

22:19:36 * AaronSw kept wondering where all the Adobe-sponsored protesters were... ;-)

22:19:38 <sbp> Very very rarely

22:19:42 <fitzix> AaronSw: How many people were there?

22:20:01 <fitzix> sbp: Yeah, but there are times - but, like you said - rare times

22:20:04 <dajobe> logger, off

22:29:04 <fitzix> later everyone :)

22:29:14 <DanCon> AaronSw, you put some swws calendar/travel info together, no? wanna remind me where? I'm collecting what I know about the event

22:29:16 <sbp> I'm not sure that RDF will ever become the de facto language for expressing relationships between resources, it's not like the World is set up for another HTML

22:29:40 <DanCon> oh ye of little faith, sbp. ;--0

22:29:40 <AaronSw> DanCon, at http://logicerror.com/SWWS-plans

22:29:44 <AaronSw> not much there yet...

22:30:08 <sbp> Maybe :-)

22:30:19 <DanCon> thx, AaronSw

22:30:33 <dajobe> logger, on

22:30:49 <sbp> Hey DanBri

22:30:51 * danbri catches up

22:33:23 <ajzzzz> ajzzzz is now known as ajmitch

22:39:04 * danbri gets rdf query over SOAP working; gives libby a preview by phone; wanders off for a while

22:39:51 <AaronSw> rdf query over SOAP? ick! talk about destroying semantics....

22:39:56 <AaronSw> why not the other way around?

22:40:54 <danbri> SOAP is supposed to be a language for defining Web-based protocols, right? So I'm trying it out, see how it works for me.

22:41:19 <AaronSw> i know, i've just been brain-dulled by all the soap semantics arguments

22:41:43 <danbri> I'll use RDF to document SOAP method semantics, when that time comes.

22:42:03 <AaronSw> err, SOAP is a langauge for sending web-based RPC messages...

22:42:13 <AaronSw> WSDL is "a language for defining Web-based protocols"

22:43:04 <danbri> ok, a system for implementing web based protocols...

22:43:17 <AaronSw> so is HTTP ;-)

22:43:21 <danbri> but anyway, i mean all this mainstream Web Services stuff

22:43:39 <danbri> rearrange these words to form a popular phrase or slogan

22:43:41 <danbri> converted

22:43:43 <danbri> to

22:43:46 <danbri> preaching

22:43:48 <danbri> ;-)

22:43:55 <GabeW> SOAP is a way for encoding an invocation of a behavior (a method) - but it can carry any xml-wrappped message...

22:44:47 * danbri trying to be a good XML citizen

22:45:18 * AaronSw wonders what that means... should we all be RDF patriots?

22:45:30 <sbp> We need an RDF flag

22:45:52 <danbri> I mean, RDF heads have a tendency to appear snooty w.r.t. mainstream XML stuff. And that's appealling but wrong.

22:46:33 <danbri> If XML mainstream and a W3C WG think SOAP is what we should use, I feel I owe them an attempt to use it before proposing anything in the 'Semantic Web Services' area

22:47:18 <AaronSw> That's the problem with SOAP... it's addictive... until you wash your mouth out with it, i suppose

22:47:34 <ajmitch> ajmitch is now known as ajbusy

22:47:34 <dajobe> I don't see why we shouldn't use it, or XML-RPC, as well as HTTP, BXXP/BEEP, ...

22:47:50 <danbri> So I'm happy enough using Perl SOAP::Lite; if SOAP's bloated etc as accused, the module seems to have protected me from the worst.

22:47:53 <AaronSw> because it loses the semantics that underly HTTP

22:48:15 <AaronSw> and thus the semantic web

22:48:23 <GabeW> SOAP can be layered on many transports (HTTP, TCP, BEEP) - you guys shouldn't care too much

22:48:28 <AaronSw> From the chump:

22:48:35 <AaronSw> A:

22:48:35 <dc_rdfig>http://www.markbaker.ca/2001/07/DeRPC/

22:48:36 <dc_rdfig> DeRPC: Mark Baker volunteers to redefine RPC systems in the application semantics of HTTP

22:48:37 <dc_rdfig> (AaronSw) It's always seemed to me that if we're going to be working on the Semantic Web, it's probably a good idea to be sure that we retain the semantics in things like HTTP, and HTML.

22:48:38 <dc_rdfig> (AaronSw) The prevalence of throw a SOAP/XML-RPC message in an HTTP POST doesn't seem to be helping matters.

22:49:14 <AaronSw> A:s/throw/throwing

22:49:15 <dc_rdfig> commented item A

22:50:11 <shellac> I better go - 45 mins. Eek

22:50:17 <AaronSw> C:|Generate RDF from your Palm Datebook file

22:50:18 <dc_rdfig> titled item C

22:51:41 <danbri> A:I totally agree about the value of HTTP's existing methods; see [http://www.xent.com/jan00/0867.html|my previous fork rant] on this.

22:51:41 <dc_rdfig> commented item A

22:53:04 * sbp wonders if http://c2.com/cgi/wiki?PalmWiki is of any interest to anyone (esp. DanC)

22:53:59 <sbp> It'd be cool to export the relationships between the pages as RDF. Aaron, can Blogspace do that?

22:54:59 <AaronSw> of course

22:55:05 <sbp> Example?

22:55:20 <AaronSw> well if there's a rel attribute on a link then it uses the REL as the property

22:55:25 <AaronSw> otherwise it uses blogspace:link

22:55:30 <AaronSw> the pages are the nodes

22:55:54 <sbp> No I mean, is there a publically available example of this being exported from LogicError?

22:56:16 <AaronSw> umm, not at present

22:56:19 <sbp> If you ran it through SiRPAC and got the SVG out, you'd have a map of your site

22:56:26 <AaronSw> that'd be cool

22:56:38 <AaronSw> i'm still fiddling with the semantics

22:56:52 <sbp> Cool

22:56:57 <AaronSw> i suppose i could do one of just the plain old links

22:58:21 <sbp> That would probably be just as interesting. You'd still get the overview

22:59:37 <AaronSw> here, i'll get one for you...

23:30:50 <AaronSw> ok, here it is: http://logicerror.com/links