Semantic Web Interest Group IRC Chat Logs for 2003-01-22

00:05:13 * DanC does a little prep for RDF cal tomorrow...

00:07:06 <danbri>http://news.com.com/2100-1001-981256.html

00:07:07 <dc_rdfig> A: http://news.com.com/2100-1001-981256.html from danbri

00:07:30 <danbri> A:|IBM aims to get smart about AI

00:07:30 <dc_rdfig> Titled item A.

00:07:41 * danbri pings guha, to see if he's heard of this

00:08:14 <danbri> A:"The Unstructured Information Management Architecture (UIMA) is an XML-based data retrieval architecture under development at IBM. UIMA will greatly expand and enhance the retrieval techniques underlying databases, said Alfred Spector, vice president of services and software at IBM's Research division."

00:08:14 <dc_rdfig> Added comment A1.

00:08:53 <DanC> A:hmm...

00:08:53 <dc_rdfig> Added comment A2.

00:09:35 <DanC> $ make TESTCASE=test/20030122mtg testcase

00:09:36 <DanC> perl ical2rdf.pl --xnames http://www.w3.org/2002/12/cal/prod_evo# test/20030122mtg.ics >test/20030122mtg.rdf

00:09:36 <DanC> unknown iprop uri; dunno what to do with value 'Ping' at ical2rdf.pl line 316, <> line 21.

00:10:20 <danbri> A:See [http://www2002.org/spector.pdf|www2002 paper (pdf)] (or [http://216.239.33.100/search?q=cache:w8-dwzbubP0C:www2002.org/spector.pdf+UIMA+RDF+IBM&hl=en&ie=UTF-8|google html].

00:10:20 <dc_rdfig> Added comment A3.

00:13:03 <danbri> A3:See [http://www2002.org/spector.pdf|www2002 paper (pdf)] (or [http://216.239.33.100/search?q=cache:w8-dwzbubP0C:www2002.org/spector.pdf+UIMA+RDF+IBM&hl=en&ie=UTF-8|google html]).

00:13:04 <dc_rdfig> Replaced comment A3.

00:14:19 * danbri heads off

00:14:59 <none> none is now known as mrlc

00:30:40 * TimBL_ catches up

00:32:16 <mdupont> how hard is it to write in cwm a query like

00:32:29 <mdupont> find all the uris of the nodes that have this subject

00:59:55 <mdupont> any cwm experts awake?

01:10:02 <mdupont> TimBL_: you wrote most of cwm?

01:11:14 <TimBL_> hi

01:11:18 <TimBL_> wassup

01:11:26 <mdupont> just a quick question about cwm

01:11:33 <mdupont> i have a set of ntriples

01:11:35 <TimBL_> k

01:11:49 <mdupont> and want to just pull out all the objects for a given subject

01:12:45 <mdupont> and then pull out all the objects again from all those objects returned from the first query, for all the statements where they are subject, but filtered by a predicate

01:12:56 <TimBL_> { :joe ?p ?o } => { ?o a :OBJECT_I_WAS_LOOKING_FOR }.

01:12:59 <mdupont> so, like step by step, traversing a data set

01:13:13 <mdupont> let me see

01:13:20 <TimBL_> recursively or just a finite number of steps?

01:13:23 <mdupont> joe is the subject of the first

01:13:27 <mdupont> just for one step

01:13:30 <mdupont> for now

01:14:14 <mdupont> Object_was_looking_for, is that a variable?

01:14:25 <mdupont> "a" is the predicate filter

01:14:33 <TimBL_> Ok, so in the first step yo mean any object of *any* predicate where joe is the subject?

01:14:37 <mdupont> is that a form of unification?

01:14:38 <mdupont> yes

01:14:44 <mdupont> joe is the subject

01:15:10 <mdupont> and "a" is the predicate of all the statements where o is the subject

01:15:26 <mdupont> and o is the object of joes statements

01:15:30 <mdupont> cool

01:15:30 <TimBL_> "a" is just short for "rdf:type"

01:15:35 <mdupont> that is it

01:15:43 <TimBL_> Sounds like you need something like

01:15:48 <mdupont> well, i want to filter by predicate

01:16:25 <TimBL_> { :joe ?p ?o. ?o :phoneNumber ?n } => { ?n a :RESULT }}.

01:16:39 <mdupont> yes

01:16:46 <mdupont> all of joes phone numbers

01:16:50 <mdupont> cool!

01:16:52 <mdupont> thanks

01:17:02 <TimBL_> So fro all obejcts (with any predicate) of joe, then for that object, any phone number it has,

01:17:07 <mdupont> yes

01:17:12 <mdupont> that is right

01:17:25 <TimBL_> Well, that is all the phone number of anything related to joe by any relationship, which is a littl euncorntrolled.

01:17:30 <mdupont> sure

01:17:38 <mdupont> but that is fine for me now

01:17:46 <mdupont> the data is very organized

01:17:50 <mdupont> and strict

01:18:07 <mdupont> i want to query the results of the gcc compiler dumps

01:18:20 <mdupont> right now i am writing greps on the ntriples

01:18:25 <mdupont> as a prototype

01:18:35 <mdupont> but there has to be a better way, i thougth

01:18:41 <mdupont> and then i remembered cwm

01:18:49 <mdupont> but it was alway very scary to me

01:18:52 <mdupont> :)

01:18:58 <TimBL_> I just generate some redicate in upper case so its easy to spot in the resulting output.

01:18:58 <TimBL_> Ok

01:18:58 <TimBL_> Ok

01:18:58 <TimBL_> I'll trust that what you say you want to do you want to do!

01:19:16 <TimBL_> If you have a lot of data, cwm may be slow.

01:19:28 <mdupont> well, i need a prototyping tool

01:19:33 <mdupont> redland is good

01:19:38 <mdupont> and handles our needs

01:19:47 <mdupont> but it more complex to program

01:20:06 <mdupont> it would be great to have a command like redland tool

01:20:20 <TimBL_> like? line?

01:20:27 <mdupont> to just exec queries against a bdb dataset

01:20:34 <mdupont> yes s/like/line/

01:20:42 <TimBL_> mmm

01:20:49 <mdupont> and save results into some form of named set

01:21:00 <TimBL_> I think command line tools are great.

01:21:08 <mdupont> i have a 6mb ntriples file

01:21:26 <mdupont> that reprsents the rdf from the guile scheme eval function!

01:21:35 <mdupont> it is massive!

01:22:07 <TimBL_> When cwm is slow, there is Euler too.

01:22:38 <mdupont> ok. i think it is time to learn how to use these tools

01:22:56 <mdupont> how is Euler faster?

01:24:00 * mdupont rtfm

01:24:10 <mdupont> thanks tim

01:24:16 <mdupont> i will do my homework

01:24:28 <TimBL_> welcome.

01:24:29 <mdupont> you gave me a good push in the right direction

01:24:47 <mdupont> i need to get some sleep now

01:24:56 <mdupont> it is 2:26 am

01:25:35 <none2> bye all

01:26:24 <mdupont> bye

01:32:38 <TimBL_> TimBL_ is now known as timbl-away

03:55:53 <DanC> wow... sometimes it pays to surf advogato randomly...

03:55:56 <DanC> "Web based proof exchange systems

03:56:01 <DanC> " -- http://www.advogato.org/person/chalst/

07:19:50 <mdupont> morning

10:45:43 <danbri> BLURB:Ideas for SWAD-Europe project promotion giveaways?

10:45:44 <dc_rdfig> B: Ideas for SWAD-Europe project promotion giveaways? from danbri

10:46:43 <danbri> B:We have some $$$ in the SWAD-Europe project for buying some project promotional stuff, like those things you get at conferences and throw away or hoard (t-shirts, pens, mousemats, temorary tattoos, etc...).

10:46:43 <dc_rdfig> Added comment B1.

10:47:40 <danbri> B:I'd like us to promote the SW, rather than the project itself, and come up with something we could give away to developers etc., that they might actually want or be amused by, and that we could use as bounty/prizees informally...

10:47:41 <dc_rdfig> Added comment B2.

10:48:01 <danbri> libby, I'm scribbing some notes here on swad-e giveaway ideas, in case folk have a wishlist

10:48:28 <libby> heh, good idea

10:48:32 <josek> danbri, maybe you could use some of that $$$ to produce an annotated version of some spec?

10:48:36 <danbri> B:Eg. t-shirt to the first developer(s) to get their RDF such-no-so system to pass a certain number of tests...

10:48:36 <dc_rdfig> Added comment B3.

10:48:44 <mdupont> irc://irc.freenode.net/introspector

10:48:45 <dc_rdfig> C: irc://irc.freenode.net/introspector from mdupont

10:49:07 <mdupont> C:|Introspector Chatroot

10:49:07 <dc_rdfig> Titled item C.

10:49:07 <danbri> josek, things are carved up in workpackages, and there is a certain amount allocated for outreach/promo stuff; most of it goes on real work, I hasten to add!

10:49:25 <danbri> do you have a particular spec in mind? hopefully the new RDF Primer fullfills some of that need?

10:50:20 <josek> well, I read the primer myself at the end of last year and I added some comments to it. I must confess I could understand it because I had some prior knowledge of RDf. It helped me to understand more things

10:50:33 <danbri> B:Since #rdfig folk are one target audience, I'd be interested to hear ideas from developers here. T-shirts are maybe old hat, but maybe you can never have too many? I keep pushing the temporary tatoo idea but all I get is funny looks!

10:50:33 <dc_rdfig> Added comment B4.

10:50:53 <josek> hmm, no particular spec right now. I was only thinking about the annotated XML spec :)

10:50:57 <danbri> B:Basically it'd be nice to be able to give some small rewards for interesting things folk do...

10:50:58 <dc_rdfig> Added comment B5.

10:51:29 * danbri liked that work... maybe a commentary on the RDF Semantics might make sense.

10:51:56 <danbri> Actually I'm trying to write something on schemas currently, especially given the current thread on rdfweb-dev regarding 'validation'; that issue needs explaining much more clearly.

10:53:24 <danbri> shell of a schema doc is at http://www.w3.org/2001/sw/Europe/reports/xml_schema_tools_techniques_report/ including an image i originally drew for the RDFS spec, but got removed from there.

10:54:02 <mdupont> so i talked to ursala bartens from the rdfig mailling list

10:54:16 <mdupont> she had a question about rdfs that maybe you can help with?

10:54:28 * danbri ducks

10:54:29 <mdupont> danbri: or dc_rdfig ?

10:54:31 <danbri> sure, ask away...

10:54:38 <mdupont> it is the question about validating rdf files

10:54:56 <mdupont> is it possible to exclude a user from adding in new properties

10:55:02 <danbri> I'd rather work on the doc and get back to you next week, than try a half-answer in irc again...

10:55:09 <mdupont> i think that it is a feature of rdf that you can add in new propertieds

10:55:28 <mdupont> me? doc? ok

10:55:37 <danbri> exclude: only if you use a non-RDF form of validation (eg. XML Schema or DTD), or else you could try writing an OWL ontology for document types, but that is research territory.

10:55:45 <danbri> does that help at all?

10:55:54 <mdupont> that makes sense

10:56:00 <mdupont> that is what i thought

10:56:27 <mdupont> is there a command line rdf validator that checks a doc against a rdfs?

10:56:56 <mdupont> anyway, i tried to help the lady,

10:57:06 <danbri> Not really, the notion of 'validation' doesn't really apply in RDFS, only wrt RDF's XML encoding.

10:57:09 <mdupont> only after she wrote to me in german did i understand he question

10:57:14 <mdupont> ok great

10:57:15 <danbri> ...where it is more akin to 'well formedness'

10:57:17 <mdupont> thanks dude

10:57:35 <mdupont> catch ya later

10:57:40 <danbri> ooh, once we get our stuff written up, any chance you could help us translate key points into German?

10:57:48 <mdupont> sure

10:57:58 <mdupont> i can translate some stuff for you.

10:58:11 <mdupont> just mail me mdupont777@yahoo.com

10:58:26 <mdupont> i might need a week for it

10:58:33 <mdupont> but i can give it a try

10:58:56 <mdupont> mail-to:mdupont777@yahoo.com

10:59:06 <mdupont> smtp://mdupont777@yahoo.com

10:59:14 <mdupont> hmm, that is not working

10:59:19 <mdupont> mail-to://mdupont777@yahoo.com

10:59:24 <mdupont>mailto://mdupont777@yahoo.com

10:59:28 <mdupont> err

10:59:37 <mdupont> anyway, chump dont like smtp

10:59:44 <mdupont> gotta run

11:01:55 <danbri> ok no rush, will be in touch. cheers.

11:04:28 <deltab>mailto:mdupont777@yahoo.com

11:05:44 <mdupont> well, i just tried to get this Chumper to recognize it

11:10:42 <danbri> morning Jim

11:14:52 <JibberJim> morning, what would a temporary tatoo be of, "666" ?

11:16:47 <danbri> "Semantic Web developer (until something better comes along)"

11:17:20 * danbri will definitely do foaf ones, but only after spec's more finished.

11:24:05 <libby> B:[some ideas|http://swordfish.rdfweb.org/~libby/2002/11/promos/] - my personal ideas, not endorsed by SWAD-E etc etc

11:24:06 <dc_rdfig> Added comment B6.

11:25:00 <mdupont> JibberJim: my safe email is mdupont 666 + 111 at yahoo dot com

11:25:45 <danbri> hi dave

11:26:05 <mdupont> hey dajobe

11:26:10 <dajobe> hello

11:28:10 <mdupont> hey dajobe i thinx aux is a reserved word on dos

11:28:22 <mdupont> if i say, ls aux, it hangs

11:28:44 <dajobe> lol

11:28:52 <dajobe> here is the fix:

11:28:54 <dajobe> format c:

11:29:25 <mdupont> mdupont@PI ~

11:29:30 <mdupont> HEHE

11:29:33 <mdupont> lol

11:29:50 <deltab> in some versions of Windows and IE you can use AUX etc. as the source for images in HTML

11:30:00 <dajobe> i can rename it, np

11:30:11 <dajobe> probably to misc

11:30:54 <danbri> yeah <img src="file:///dev/mouse"/> or similar used to wreak havoc. linux too i think...

11:31:07 <dajobe> excellent

11:32:31 <dajobe> hi danja, nice to see you dip in

11:35:20 <danja> hi dajobe - all well?

11:35:48 <dajobe> yeah

11:36:00 <danja> I'm hanging about waiting for XF86 to download

11:36:07 <dajobe> could be a while then

11:36:29 <danja> aye

11:36:41 <dajobe> i see you noticed our ESW weblog, our first trackback

11:36:58 <dajobe> I was wondering if we should turn that on for all items

11:37:05 <danja> heh - movable type is wonderful

11:37:15 <dajobe> oh, maybe I should add you to the sidebar as other sw blogs

11:37:35 <danja> and yours

11:37:51 <dajobe> do you use the categories much?

11:38:14 <danja> only just started with mt, still getting used to it

11:38:34 <danja> though really want to map to Wordnet/odp categories asap

11:38:52 <dajobe> did you see the ESW semantic blogging thing we're doing?

11:39:22 <danja> I think I've seen most of the docs - cool

11:39:36 <dajobe> HP is doing as one of their demonstrators: http://www.w3.org/2001/sw/Europe/reports/chosen_demos_rationale_report/hp-applications-selection.html#sec-app1

11:40:18 <dajobe> but you read the public-esw list so I guess this isn't news

11:40:36 <dajobe> which BTW is great that you are participating in

11:41:02 <danja> the e-sw stuff seems like a real shot in the arm - things happening!!

11:41:34 <danja> keeping danbri out of trouble anyway

11:41:41 <dajobe> heh

11:42:32 <dajobe> don't bother him today, he's got to help with rdf core doc pub stuff ;)

11:42:53 <danja> aw

11:44:03 <danja> I've made semantic blogging the priority for my Ideagraph app btw

11:44:34 <danja> once the bits are in place for basic rss items etc

11:44:51 <dajobe> oh, you do SVG too. have you seen the recent #foaf SVG stuff?

11:45:10 <dajobe> well danbri chumped some pics here yesterday IIRC

11:45:17 <danja> not looked for a few weeks

11:45:20 <mrlc> hi

11:45:39 * danja goes chump-hunting

11:47:28 * JibberJim is pretty sure that danja's seen foafnaut.

11:48:06 <danja> ah - dajobe, while you're there, the scratchpad page could do with a link to the RSS version - took me a while to find it the other day

11:48:14 <dajobe> ok

11:48:50 <danja> JibberJim - yep, but it keeps improving ;-)

11:48:59 <danja> I love it

11:49:40 <dajobe> indeed, and it is worth keeping on telling people about it

11:50:06 <JibberJim> I got it vaguely working in Batik btw.

11:50:13 * danja blogs again

11:50:37 <danja> JJ - that's cool, Batik is great fun

11:50:49 <danja> how?

11:51:08 <JibberJim> by lots of try/catch around the errors :-)

11:51:36 <dajobe> lol

11:51:50 <dajobe> try { Java } catch { Out of memory }

11:51:54 <danja> heheh

11:53:35 <danja> ported your script, or used Jena (or something)?

11:53:46 <JibberJim> oh no, all script.

11:53:55 * JibberJim doesn't use java!

11:54:09 <JibberJim> and there's no rdf parser used in foafnaut... it just uses XML...

11:54:24 <JibberJim> but my javascript RDF parser does run under Batik.

11:54:45 <danja> ah, novel...

11:56:30 <danja> hmm - have you tried talking between the script and Java?

11:57:03 <JibberJim> no, one day it might be nice if I can find a java speech component.

11:57:18 <danja> ooh - that's be good

11:57:34 <danja> give the 'nauts a voice

11:58:19 <danja> hmm, foafnaut.org is timing out - probably this ftp I've got running

11:58:28 <JibberJim> well I did talking svg with ASV on windows.

11:59:26 <danja> wow - what did you have it doing?

11:59:39 <JibberJim> reading out an RDF description of an image...

11:59:54 <danja> marvellous

12:00:03 <JibberJim> - http://jibbering.com/2002/9/talking.html

12:00:17 <JibberJim> which was nice, as it brings us back on-topic :-)

12:01:12 <danja> cool - can't remember seeing this before (though had seen dpawson's bits)

12:04:06 <danja> ok folks, I've just got a new 'office' server, what are the must-have RDF-related services/apps?

12:04:29 <JibberJim> a foafnaut scutter...

12:04:55 <danja> yep!

12:09:19 <dajobe> is this a silly Q, why isn't http://ideagraph.net/2002-12/status.htm just a part of your blog? Or do you want to split that off from your personal stuff?

12:09:51 <dajobe> I'm wondering at what point you mix/split this stuff. If there were categories, you could present different views?

12:10:34 <danja> pretty arbitrary at present

12:11:14 <danja> categories would definitely be great solution - save on domain names for a start...

12:12:27 <danja> thanks for the reminder - must update the status page

12:13:14 <danja> (nothing much new, just bugfixing really)

12:13:54 <danja> and trying to set up this server with an XF86-unfriendly onboard video chip...

12:15:05 <dajobe> that is sadly getting more common

12:15:40 <danja> I naively thought it would work right away, being an ATI chip

12:15:52 <dajobe> aren't they nvidia these days?

12:16:03 <danja> no idea

12:16:22 <danja> ah - another driver to try...

12:19:53 <danja> I'm trying Debian this time, btw

12:20:43 <dajobe> B:"I <heart> cwm" ?

12:20:43 <dc_rdfig> Added comment B7.

12:21:09 <dajobe> I guess irony doesn't work too well on IRC

12:22:08 <dajobe> so I've added an rss1.0 feed link to the rdf interest chump

12:22:18 <dajobe> should I add one of those <link> rss feed ptrs too?

12:29:22 <danja> thanks dajobe - not sure about the <link>, can't hurt I 'spose

12:35:29 <danja> hehe - only just seen where you put the <heart>

12:36:05 <danja> "My other web's Semantic"

12:37:05 <dajobe> bbl

12:37:34 <none> none is now known as mrlc_

12:39:56 <danja> nope, nv driver didn't work either

12:44:05 <danja> dajobe - is there a way of specifying a generic (S)VGA card?

13:10:22 <dajobe> danja: you need the vesa driver I guess

13:11:17 <danja> thanks

13:13:01 <danja> woo-hoo!!! it worked!!! thanks dajobe!

13:13:31 <dajobe> it'll be slow though, no acceleration

13:13:34 <danja> ah, bombed out on mouse - did have a graphic screen though

13:14:15 <danja> graphic speed not really a big issue, I mostly want it for a server

13:29:48 <libby> BLURB:http://www.w3.org/2002/12/cal/

13:29:48 <dc_rdfig> D: http://www.w3.org/2002/12/cal/ from libby

13:31:16 <libby> D:|RDF Calendar workspace

13:31:16 <dc_rdfig> Titled item D.

13:31:52 <libby> D:chat about RDF calendaring tonight at 5pm GMT [invitiation|http://lists.w3.org/Archives/Public/www-rdf-calendar/2003Jan/0016.html]

13:31:52 <dc_rdfig> Added comment D1.

13:32:05 <libby> D:please add agenda topics here

13:32:05 <dc_rdfig> Added comment D2.

13:32:33 <libby> D:1. Andrea Campi/Benjamin Sonntag have said they'll talk about a specific example of CAP in use for facilitating a meeting.

13:32:33 <dc_rdfig> Added comment D3.

13:44:30 <libby> D3:no agenda items as yet

13:44:31 <dc_rdfig> Replaced comment D3.

13:57:52 <danja> got XF86 now - thanks dajobe, bye y'all

14:17:06 <danbri> hi charles

14:17:09 * danbri heads off to reboot

14:33:26 <libby> D3:a question from maxf: what tools support RDF calendar output that don't support icalendar?

14:33:26 <dc_rdfig> Replaced comment D3.

14:33:53 <LotR> what time was the meeting again?

14:34:03 <libby> 5pm gmt

14:34:16 <libby> 2 1/2 hours-odd

14:34:40 <libby> LotR, anything you'd like to discuss? we're a bit short on agenda items...

14:35:44 <chaalsMEL> lib, any thoughts about how to deal with the hejri calendar?

14:35:46 <libby> B:I like danja's idea: 'my other web's semantic'

14:35:46 <dc_rdfig> Added comment B8.

14:36:06 <libby> hi chaalsMEL. not really....sorry...

14:37:29 <chaalsMEL> the tricky issue is that the date mapping depends on where in the world you are, because lunar visibility isn't the same day in all places, so a month that has 29 days in australia may have 30 in the uk.

14:38:22 <LotR> libby: not really. I'm just looking along to see if there's any interesting ietf:ical bits that come up. no interest in rdf, so..

14:38:28 <libby> wow. that is tricky

14:39:05 <LotR> hejri == islamic?

14:39:10 <chaalsMEL> I think it is like recurrent events that occur in a non-GMT timezone - for example W3C has a meeting that is at 0800Z-0400 during the time that boston is on summer time, and is at 0800Z-0500 when boston is at standard time.

14:39:19 <chaalsMEL> which amounts to the same problem.

14:39:24 <libby> I don;t know anything much about it chaals. if it helps to keep the gregorian in, amkes sense

14:39:37 <chaalsMEL> yes, is islamic calendar.

14:40:32 <libby> recuurence is an issue we've not really tackled. it lookslike it might be doable with some form of rules so maybe cwm might be useful.

14:40:41 <libby> so chaals, it's about conversion?

14:41:12 <LotR> libby, did you see my recurrence tutorial?

14:41:23 <libby> man, isn;t it a bit late in Melbourne?

14:41:29 <chaalsMEL> I think the appraoch I want to follow up is publishing a set of schemas at http://some.org/calendars/gregorian/boston/2002/02/11 and so on, and similar ones at http://some.net/calendars/hejri/melbourne/1453/03/12 and so on that explain how they are related to each other.

14:41:39 <libby> no, LotR, I ddn;t, can you poit me to it?

14:42:03 <libby> challs, have you got any data at all?

14:42:03 <LotR> libby, http://tipi.sf.net/doc/recur/

14:42:09 <chaalsMEL> very late. I am going to bed soon and not going to make the calendar meeting (trying to talk a little to the dublin core accessibility folk during their face to face)

14:42:21 <libby> right

14:42:25 <LotR> it's very much in progress

14:42:25 <libby> thought so

14:42:28 <libby> thanks LotR

14:42:32 <libby> can I chump it?

14:42:39 <libby> chaals, when are you back in europe?

14:42:50 <LotR> libby: I guess

14:42:57 <libby> ta

14:43:00 <libby>http://tipi.sf.net/doc/recur/

14:43:00 <dc_rdfig> E: http://tipi.sf.net/doc/recur/ from libby

14:43:04 <chaalsMEL> dunno. Prolly after W3C tech plenary in march.

14:43:21 <libby> E:|Recurrence tutorial by Martijn van Beers

14:43:21 <dc_rdfig> Titled item E.

14:43:39 <libby> E:"This document is a short tutorial on iCalendar recurrence, and how to use this in your Perl programs."

14:43:39 <dc_rdfig> Added comment E1.

14:43:44 <libby> bummer chaals

14:43:53 <LotR> E: very much in progress

14:43:53 <dc_rdfig> Added comment E2.

14:44:04 <libby> maybe we shuld try andscheule it better for you

14:44:20 <libby> gotta run, back later. night chaals

14:45:23 <chaalsMEL> LotR, interesting stuff...

14:46:32 <LotR> chaalsMEL: thanks :) let me know if anything is not clear/correct/etc

14:46:42 <chaalsMEL> E: This is why I want to provide some RDF way of dealing with occurrences according to different calendars (every easter, every ramadan are two things that I want to know for practical purposes but can't fit into the existing iCal approach)

14:46:42 <dc_rdfig> Added comment E3.

14:47:28 <JibberJim> Can you define when Easter is with iCal ?

14:47:45 <chaalsMEL> well it seems clear. From that it seems to me that I can't describe a recurrent event according to non-gregorian calendaring.

14:48:00 <LotR> you need to know the moon cycle for determining easter, no? so I don't think so

14:49:06 <chaalsMEL> first sunday after the first full moon after the first northern spring equinox that isn't passover (for the western churches - I don't know how the orthodox churches do it but they come up with different results most years)

14:49:55 <LotR> hmm. /me didn't know about the 'that isn't passover' part

14:50:17 <chaalsMEL> the basis for determining Hejri dates is knowing when the new moon is visible in a given place.

14:50:57 <DanC> chaalsMEL, if you want to describe such events, that bit of english seems to work. I don't see how your problem relates to E:. Do you want/expect machines to get involved somehow?

14:51:14 <chaalsMEL> So the moon is the key in both these cases (and can be calculated)

14:51:35 <chaalsMEL> danC Yes. I don't know when the moon comes up, although it is a predictable event.

14:51:55 <DanC> how do you want/expect machines to get involved?

14:52:08 <chaalsMEL> millions of people who are affected by easter don't even know how to calculate it, but know that it will occur every year.

14:52:14 <JibberJim> I don't think it's unreasonable to schedule in your calendar that you'll spend every Easter with your family.

14:52:21 * DanC can't see why just doing an HTTP GET from some place that's trusted to know when easer occurs doesn't suffice.

14:52:40 <chaalsMEL> in most european countries and australia you get a holiday for easter...

14:53:13 <chaalsMEL> it does, if I can set a recurring event that says "whenever this trusted place says so"

14:53:40 <LotR> so can you reliably calculate the new moon date x months from now?

14:53:53 <chaalsMEL> it does - an HTTP get from a trusted place suffices...

14:54:16 <DanC> I do that sort of thing ("whenever this trusted place says so") with RDF (specifically, with cwm and log:includes and string:scrape) all the time.

14:54:26 <chaalsMEL> there are programs that claim to calculate it pretty reliably

14:54:32 <DanC> If you want iCalendar tools to grok, you're out of luck.

14:54:39 <chaalsMEL> for hundreds of years...

14:55:09 <DanC> well, one could make a little tool that consulted a trusted site and published the results in iCalendar format, of course.

14:56:06 <chaalsMEL> iCalendar tools: Right. This is where I think it is valuable to extend what iCal tools grok, so it includes pointing to recurrence at apparently arbitrary points.

14:56:42 <chaalsMEL> I am not sure what the best approach is - for the hejri calendar there are plenty of useful points through the year (likewise for the jewish clendar I imagine)

14:56:42 * DanC has no intent/hope of changing the scope of iCalendar

14:57:27 <chaalsMEL> I don't want to change iCalendar. I also don't see much value in rewriting iCal tools to do RDF if they don't provide some of the things that I can't get with iCal tools anyway.

14:58:02 <DanC> who's rewriting iCal tools?

14:58:38 <LotR> well, I'm rewriting Net::ICal.. ;)

14:58:43 <chaalsMEL> for the chinese and vietnamese calendars the only important event I know of is new year (Tet, for those old enough to recall what Americans think it means). At one per year there are only a few things in my life

14:59:07 <chaalsMEL> that count, so I can happily rely on http get. (If I have tools that actually deal with that properly).

14:59:41 <chaalsMEL> by building tools that convert it to iCal single events, and thinking my own way around those events.

15:01:13 <LotR> I think there's holiday calendars on sites like icalshare.com. doesn't that suffice?

15:02:00 <chaalsMEL> can I say "this is a recurring event, whose frequency is relative to hanukkah"?

15:02:22 <DanC> well, you can say it, but you can't expect iCal tools to grok.

15:02:36 <chaalsMEL> (That's my real issue - ultimately I don't think it matters where the information comes from).

15:03:23 <DanC> anybody wanna play pathcross? i.e. you encode your travel schedule in RDF and make it available via HTTP, then we try to get cwm to notice intersections in our travel plans.

15:03:31 <chaalsMEL> so if I have an RDF-ised version, that treats recurring events as things that can be in any namespace,

15:03:33 * DanC updates http://www.w3.org/People/Connolly/home-smart.rdf

15:04:03 <DanC> (cwm or Euler or whatever)

15:04:20 <LotR> DanC: I don't travel enough (at all) :)

15:04:39 <chaalsMEL> and knows how to relate to gregorian namespaces, and groks the idea that some RDF description of an event might be able to describe the event as a sequence of dates it understands,

15:05:10 <DanC> hmm... well, LotR, you could just record where you are/live (to the resolution of the nearest airport, to balance privacy with usefulness of info).

15:05:18 <chaalsMEL> then I can say "go find out what you can about the namespace I created yesterday for some event, relate that to your internal calendar (which I presume for most tools is curently gregorian)

15:05:34 <DanC> go for it, chaalsMEL.

15:05:52 <chaalsMEL> and show me how my religious holidays are going to intersect with my work calendar".

15:06:31 <chaalsMEL> before going for it I am trying to figure out if my logic is sound (I am not sure...)

15:06:47 <DanC> hard to say until you try it, in my experience.

15:07:02 <setre> is logger possible to use with dircproxy?

15:07:18 * LotR bets there won't be many people making a stop in EHEH while traveling

15:07:38 <LotR> maybe if I list EHAM as the airport

15:09:11 <DanC> hmm... take attendance today using foaf/about-me RDF files?

15:09:16 <DanC> s/files/pointers/?

15:09:35 <LotR> chaalsMEL: you could have recurring events relative to hanukkah (how many ways to spell that are there anyway) if you know when that is in gregorian terms. Guess it is a problem if you want it for hanukkah x years from now

15:09:48 <chaalsMEL> iata:AMS isn't that far away...

15:09:58 <chaalsMEL> LotR, right.

15:10:29 * JibberJim will play pathcross, I already intend to mark up my travel plans just as soon as I get round to it.

15:10:44 <chaalsMEL> but a recurring event relative to hanukkah is like 4 days after hanukkah this year, 4 days after hanukkah next year, 4 days after hanukkah the year after...

15:10:52 <JibberJim> also have lots (well 20 odd) peoples nearest airports from peoples foaf data.

15:11:01 <DanC> JibberJim, do you publish your travel plans in HTML? I have an XSLT scraper (I think... testing to see if it has rotted...)

15:11:20 * chaalsMEL doesn't trust foafdata to know my nearest airport

15:11:20 <LotR> chaalsMEL: that's one variety of relative to hanukkah

15:11:21 <JibberJim> no, I don't publish them at all yet, I'll start with RDF!

15:11:35 * chaalsMEL doesn't always trust chaals to knw the nearest airport...

15:11:57 <chaalsMEL> LotR: yep. It's the variety I am intersted in.

15:12:11 <LotR> *nod*

15:12:46 * JibberJim is intending to update mine as and when to be reasonably accurate, despite travelling around a lot.

15:13:15 <LotR> chaalsMEL: well, ietf:ical does have a CALSCALE property. I guess you can propose an extension for the jewish calendar

15:14:04 <LotR> JibberJim: if you keep an accurate calendar, with travelling info, you can have your computer do it for you :)

15:14:25 <JibberJim> I don't have that structured travel plans...

15:16:54 <chaalsMEL> LotR thanks. That's something I had missed.

15:17:53 <LotR> chaalsMEL: the islamic calendar sounds like something you can't really codify tho

15:19:11 <chaalsMEL> there are about a billion people whose lives are substantially ordered by it. And it is codified in a program compiled for DOS/Windows machines...

15:19:53 <chaalsMEL> which claims to be always correct from AH 1 until now, and "reliable enough" "into the future"

15:20:06 <LotR>http://cvs.sourceforge.net/cgi-bin/viewcvs.cgi/reefknot/Net-ICal/test-data/rfc2445

15:20:06 <dc_rdfig> F: http://cvs.sourceforge.net/cgi-bin/viewcvs.cgi/reefknot/Net-ICal/test-data/rfc2445 from LotR

15:20:27 <LotR> F:|iCalendar data test suite

15:20:27 <dc_rdfig> Titled item F.

15:21:26 <LotR> F:far from complete, but perhaps a nice start for a complete set of iCalendar fragments to test rfc compliance

15:21:26 <dc_rdfig> Added comment F1.

15:21:42 <chaalsMEL> (The algorithm is more complicated that 11 months of fixed length and one of variable length, dependent on simple frequency rules - an extra day every 4 years except for every hundred that is not also a 400 year repeat.

15:22:08 <LotR> chaalsMEL: so it's basically a historical database. which isn't *that* great for calendaring/scheduling purposes

15:22:27 <chaalsMEL> Where the days starts at a time based on the difference between disappearance of the sun and its reappearance, adjusted for arbitrary geographic boundaries).

15:22:27 <JibberJim> DanCon, did you invent the various foaf date bits (like foaf:yyyy) for your events smart?

15:23:34 <chaalsMEL> LotR, no, the program will produce a calendar for any given lat/long (or many city names) for history/future until astronomical time.

15:24:29 <chaalsMEL> (the length of a month varies according to geographic location, with days starting at the predictable phenomenon of sundown)

15:26:17 <LotR> and people claim the gregorian calendar system is a nightmare :)

15:26:23 <chaalsMEL> it's a calculator that figures out where the sun and moon and earth are, and the answers questions like "how long is shawwal in AH 1672?" or what gregorian date was AH:567-05-09 in jerusalem.

15:27:42 <DanC> foaf date bits... I think maybe I did invent them... I think I'm uninventing them now...

15:28:14 <chaalsMEL> is there a generally recognised namespace for encoding dates according to the W3C Note on the topic?

15:28:17 <JibberJim> I looked up and you were the only user I could find, so don't worry, you've not polluted anyone else...

15:28:36 * chaalsMEL thinks he used one but doesn't recall if he invented it or not

15:29:45 <chaalsMEL> LotR: it is if you just watch the moon every few weeks to give you an idea of which month you are in now.

15:30:40 <chaalsMEL> The gregorian calendar coincides nicely with the seasons, and keeps the equinox/solstice at about the same date every year.

15:30:48 <LotR> chaalsMEL: if you do that, I doubt you're likely to be using a computer. let alone use it for scheduling purposes :)

15:31:16 * DanC uses @prefix dt: <http://www.w3.org/2001/XMLSchema#> for w3c dates

15:31:28 <DanC> e.g. <#iswc2003> k:endingDate [

15:31:28 <DanC> dt:date "2003-10-23" ];

15:31:28 <DanC> k:eventOccursAt [

15:32:10 <chaalsMEL> which is what it was designed to do. This is good for a bunch of planning, like making it easy to figure out when easter will be ahead of time.

15:33:52 <chaalsMEL> actually the seasons aren't that regular, because they are influenced by other factors. Bushfire season is february in Oz (hence the term february dragon) but they don't always cooperate...

15:37:03 <chaalsMEL> LotR - depends on your scheduling needs. Except when travelling there are relatively few dates I need to worry about - things recur on cycles of 7 days (meetings) or birthdays (even my muslim friends celebrate those on the gregorain calendar) or religious holidays when people are busy (which split about even across gregorian-based ones, islamic ones, jewish ones, and pagan-based ones where there are multiple conflicting calendars

15:37:48 <chaalsMEL> And lunar cycles - affects the tides, the fishing, menstrual cycles, and in my honest belief based on years of observation when it is a good time not to go to the pub.

15:37:56 * DanC is struggling with semweb 101... how do I publish the facts I want the world to know about me?

15:38:35 * DanC wishes chaalsMEL would take his energy and put it into coding.

15:38:40 * DanC or even writing an article

15:39:04 <JibberJim> - http://jibbering.com/foaf/identify.1?nick=DanCon is most of what I know about you, well and can expose again in a meaningful way.

15:39:55 * DanC takes a look at http://jibbering.com/foaf/identify.1?nick=DanCon thru N3 glasses

15:40:00 <JibberJim> - http://jibbering.com/foaf/foaf.sha1.6.1?94b6eb0c835f928c5ed565dc3ed1a355ac1b41e5 I know all of that too.

15:40:31 <DanC> hmm... foaf:homepage "http://www.w3.org/People/Connolly/" that should be foaf:homepage <http://www.w3.org/People/Connolly/>, right?

15:40:54 <JibberJim> yes...

15:41:07 <JibberJim> sorry about that.

15:41:26 <DanC> ooh... tagboard.jpg> img:hasPart ... cool!

15:42:09 <DanC> ok, I committed http://www.w3.org/People/Connolly/travel-sched.rdf

15:42:26 <DanC> but that just says a bunch of events exist. It doesn't actually say that I plan to attend them.

15:44:02 <DanC> and it gives place names, but not lat/long.

15:44:29 <DanC> I do have rules for correlating those place names to lat/long; maybe I'll mix those in...

15:45:29 * DanC thinks with http://www.w3.org/2000/10/swap/pim/usPlace2LatLong.n3

15:46:27 * LotR thinks eventOccursAt is named strangely for its content

15:46:45 * DanC didn't make up eventOccursAt; cyc guys did

15:47:31 <LotR> I wasn't trying to put the blame on anyone

15:49:02 <LotR> so far it looks easy to turn it into iCalendar :)

15:50:44 <DanC> ah... yes, I can write rules for that too...

15:56:05 * DanC commits travel-sched.rdf 1.2 with lat/long info

15:56:29 <DanC> JibberJim, what properties do you use for lat/long in your SVG stuff?

15:56:36 <DanC> or do you just key off airport codes?

15:57:07 <JibberJim> at the moment I just get it from airport codes, I should extend it sometime.

15:57:24 <JibberJim> I found out why you couldn't see the lines btw.

15:57:42 <DanC> hmm... so I could get a picture if I made a route from MCI<->SNA for my trip to the TAG ftf, say.

15:57:58 <DanC> ah... I have that route in http://www.w3.org/2003/02dc-sna/itin1.rdf

15:58:03 <DanC> gotta mix that in.

15:58:27 <JibberJim> yeah it should work straight out the box on your itin stuff, work for the geo lat/lon namespace will be added real soon now.

15:58:49 * DanC hunts for geo namespace...

15:59:17 * DanC owes a reply to www-rdf-calendar inquiry about contact namespaces...

15:59:31 <LotR> now it needs a NOT_TO_BE_USED_BY_HUMANS warning :)

15:59:32 <JibberJim> - http://www.w3.org/2003/01/geo/wgs84_pos

15:59:50 <DanC> it=?, LotR?

16:00:03 <LotR> that travel-sched.rdf

16:00:41 <DanC> oh. I pretty much always put on N3 glasses when I actually want to read .rdf files

16:01:43 <DanC> or glasses like this: http://www.w3.org/2003/02dc-sna/itin1.png

16:02:03 <JibberJim> hmm..

16:02:09 <DanC> (oops; date label in itin1.png is bogus)

16:02:09 <JibberJim> - http://jibbering.com/2002/9/routes.svg?url=http%3A%2F%2Fwww.w3.org%2F2003%2F02dc-sna%2Fitin1.rdf

16:02:37 * DanC fires up win4lin to use fancy SVG features...

16:03:48 <JibberJim> that's better, fixed it... it was looking for the wrong long predicate.

16:04:01 <LotR> DanC: 'n3 glasses' ?

16:04:16 <DanC> n3 glasses: python cwm.py --rdf foo.rdf --n3 | less

16:04:38 <DanC> where cwm.py is from http://www.w3.org/2000/10/swap/

16:05:51 * LotR ponders installing python

16:06:36 * DanC notes timbl's python endorsement http://www.w3.org/2000/10/swap/#L88

16:07:52 * LotR is a perl fan :)

16:08:03 * DanC has a love/hate relationship with perl

16:08:37 <DanC> I still have my perl 3.x quick reference card, given to me by tchrist@convex.com when I sat in the office next to his in ~1990.

16:09:38 <DanC> I often (a) start with a text processing problem, (b) get a solution-for-today written in perl in half an hour, (c) add a few features, (d) rewrite in python for manageability after it reaches 100 lines or so.

16:10:37 <LotR> so the problem is step b :)

16:11:11 <DanC> you're saying I shouldn't bother with perl at all? ;-)

16:11:56 <LotR> no, that you apparently lack the will/discipline to start out writing managable code

16:12:31 <DanC> well, when I'm feeling disciplined, I start in python.

16:13:11 <LotR> s/disciplined/submissive/ ;-P

16:13:38 <DanC> I learned to write nasty perl. I learned to do software engineering in python. (well... I built my habits in C/C++, while reading Modula-3 books... modula-3 thinking applies easily to python)

16:14:28 <DanC> i learned perl before use strict, modules, objects, all that stuff.

16:14:31 <LotR> I think if you really grok 'software engineering' it doesn't really matter what language you write in

16:14:55 <LotR> DanC: ah, so you're writing perl4 instead of perl5 :)

16:15:09 <DanC> well, I'm competent in c/perl/python/shell/lisp/scheme and a few others, but I find it really hard to do software engineering in perl.

16:15:34 <LotR> lots of people seem to feel that way

16:17:01 <DanC> dude, I'm still pretty much writing perl3. "Perl 3.0 Reference Guide -- Copyright (c) 1989, 1990 Johan Vromans. Rev 3.0.18.1". Old habits die hard. Though I have learned to use my() rather than local().

16:17:35 <LotR> I'm too young for perl3 :)

16:17:48 <LotR> well, not really, but I didn't run into perl till perl4

16:19:22 * LotR should go learn a lisp or scheme

16:20:18 * DanC recommends reading the scheme48 source code

16:20:31 <DanC> ... and papers. And the modula3 papers, and the m3 library code.

16:21:17 <DanC> where was I... ah yes... travel schedule... relating to ical vocab, geo stuff.

16:22:05 * LotR goes vote

16:22:23 <LotR> well, after I find my voter card

16:23:25 <libby> maan, the curse of zetland rd, poor danbri\'s machine died now

16:26:40 <DanC> are the various foaf scutterplans all maintained by hand? I've got an idea for decentralizing it: allow folks to get in the scutter plan by making two links from some page google is likely to know about (1) to their foaf.rdf thingy, (2) to the foaf project page.

16:26:56 <DanC> we can find such pages with google's back-link support.

16:27:06 <DanC> and it'll give the foaf project page more google-kharma

16:27:15 <JibberJim> foaf discovery is mainly done by google in any case.

16:27:28 <DanC> really? how so, JibberJim?

16:27:43 <JibberJim> google, the rdfweb wiki, and manual submissions to any one of the scutter services.

16:27:59 <JibberJim> a google search for foaf filetype:rdf

16:28:29 <DanC> "about 1,550". wow!

16:28:33 <JibberJim> provided by pixel at http://www.perceive.net/xml/googleScutterNoChatlogs.rdf

16:28:52 <JibberJim> cron'd every morning I believe.

16:29:30 * JibberJim scutters 1424 urls.

16:30:00 <JibberJim> of which only 898 give me any triples.

16:30:55 <DanC> ooh... cool... http://rdfweb.org/foaf/examples/edd-foaf.jpeg

16:30:59 <DanC> how is that done?

16:31:16 <JibberJim> that looks like rdf-author

16:31:52 <DanC> oops... contact:nearestAirport leaked out into foafdom. I guess I better add it to the contact schema!

16:32:20 <JibberJim> yep, we draw a map of people at http://jibbering.com/2002/8/foaf-people-map.svg

16:34:55 * DanC crosses fingers, takes a look thru linux SVG viewer

16:35:40 <DanC> wow!

16:41:28 <libby> F:great, this looks useful

16:41:29 <dc_rdfig> Added comment F2.

16:43:18 <LotR> libby: I was going to post that last meeting, but cvs.sf.net was down

16:43:36 <libby> ah yes

16:44:03 <libby>ftp://ftp.metamatrix.se/pub/software/mimedir-parser/

16:44:03 <dc_rdfig> G: ftp://ftp.metamatrix.se/pub/software/mimedir-parser/ from libby

16:44:12 <mdupont> hey dajobe , this aux directory is not going to work on windows

16:44:17 <libby> G:|mime-directory parser in java

16:44:17 <dc_rdfig> Titled item G.

16:44:29 <mdupont> really silly, that

16:44:29 <libby> G:by Daniel Resare

16:44:29 <LotR> what is mime-directory?

16:44:30 <dc_rdfig> Added comment G1.

16:45:03 <libby> G:mime-directory is the format for icalendar

16:45:03 <dc_rdfig> Added comment G2.

16:45:06 <libby> I think

16:45:15 <dajobe> mdupont: I already said I'd rename it, earlier today

16:45:24 <mdupont> oh

16:45:27 <mdupont> cool

16:45:32 <dajobe> yes, to nul

16:45:33 <dajobe> or con

16:45:38 <mdupont> sorry, i missed that

16:45:43 <mdupont> or lpt1

16:45:45 <mdupont> ?

16:45:47 <dajobe> sure

16:45:50 <mdupont> LOL

16:46:02 <mdupont> dajobe: you are too funny

16:46:19 <mdupont> i cant chat with you at work, i will get in trouble for laughing!

16:46:23 <libby> G: the test document seems to work fine, so I'm going to try and replicate Dan Connolly's [ical2rdf.pl|http://www.w3.org/2002/12/cal/ical2rdf.pl] in java using it

16:46:23 <dc_rdfig> Added comment G3.

16:46:39 <libby> G:but I havn't got around to it yet

16:46:39 <dc_rdfig> Added comment G4.

16:46:40 <dajobe> well, you gotta laugh at someone polluting the filename space with that sillyness ;)

16:47:01 <dajobe> libby: maybe announce the upcoming calendar chat in the topic?

16:47:22 <mdupont> dajobe: about the calendar,

16:47:33 <libby> yeah, I don;t know how to chnge it...can you?

16:47:39 <mdupont> will you be able to address any times at all?

16:47:39 <libby> good idea

16:47:49 <mdupont> like all mondays

16:48:09 <mdupont> or weekedays from 9-5?

16:48:30 <libby> thanks, bt what do I do :)

16:48:40 <dajobe> use the /topic command

16:48:45 <libby> ta

16:48:46 <LotR> libby: /topic foo

16:49:02 <DanC> foafnaut is way cool.

16:49:34 <libby> libby has changed the topic to: RDF calendaring chat today 5pm-6pm - all welcome

16:49:35 * DanC hunts for "view source" link from foafnaut... er... I guess i can just use the browser's view source?

16:49:53 <DanC> libby, add timezone info

16:49:59 <libby> libby has changed the topic to: RDF calendaring chat today 5pm-6pm GMT - all welcome - http://www.w3.org/2002/12/cal/

16:50:02 <libby> oops yep

16:50:03 <DanC> I suggest writing it 1700Z

16:50:31 <DanC> s/today 5pm-6pm/Wed, 22 Jan 1700Z/

16:50:31 <libby> libby has changed the topic to: RDF calendaring chat today 1700Z-1800Z - all welcome - http://www.w3.org/2002/12/cal/

16:50:38 <DanC> well, today is good enough.

16:50:43 <libby> ble

16:51:10 <libby> yeah, foafnaut is way cool

16:51:10 <JibberJim> you should be able to use the browsers view-source for foafnaut, for the server side portion, (or just download the tarball for offline use that contains the source)

16:51:19 * mdupont restarts to get the calender

16:51:53 * LotR wonders how long he'll be able to attend the meeting

16:51:54 <JibberJim> for the client side portion I mean, the serverside is the scutter which you can download and jibbering.com/foaf/foaf.sha1.6.1.txt

16:53:34 <mdupont> the mozilla calender is cool

16:53:46 <mdupont> bit slow

16:54:32 <DanC> foafnaut is way cool... I wanna navigate my family photos this way

16:55:27 <DanC> EricM? foafnaut is the tool we need for organizing groups and such in the next phase of the semantic web activity.

16:55:38 * mdupont looks for foafnaut

16:55:44 <DanC> hmm... aspect ratios get smushed

16:55:56 <JibberJim> yeah, an unfortunate bug in the SVG Viewer.

16:56:00 <DanC> I started here to get to foafnaut: http://jibbering.com/2002/8/foaf-people-map.svg

16:56:03 <JibberJim> (and a bug in the spec IMO)

16:56:41 <ericm> DanC, when you were talking about spring relationships and social networks, foafnaut what was i was thinking of as well...

16:57:14 <ericm> i was thinking more work had to be done in this however before it would be useful

16:57:21 <ericm> then i saw one of the latest images...

16:57:23 * ericm hunts

16:57:35 <mdupont> hmm

16:57:42 <mdupont> i get a scripting error

16:57:46 <JibberJim> The self organising map part of foafnaut doesn't work unfortunately.

16:58:04 <mdupont> ok, it works with ie6

16:58:08 <DanC> libby, which deely is the agenda?

16:58:19 <ericm> -> http://rdfweb.org/foaf/images/foafnaut-screenshot-path.jpg was quite impressive

16:58:42 <mdupont> cool!

16:58:43 <libby> er, D,. I think

16:58:51 <mdupont> looks a bit like touchgraph

16:58:52 * ericm puts head back down and goes back to work... hoping to have time for calendaring later

16:58:58 <ericm> mdupont, agreed

16:59:22 <mdupont> i have started to clean up vcg yesterday

16:59:26 <DanC> D:I'd like to talk about pathcross, travel schedules

16:59:26 <dc_rdfig> Added comment D4.

16:59:30 <libby> crap, we're about to start

16:59:33 <libby> pathcross

16:59:37 <mdupont> that is a graphing tool

17:00:03 <ericm> thinks foafnaut implemnting some of xerox's graphical fish-eye view techniques would be *very* cool

17:00:41 <mdupont> the vcg supports fisheye

17:00:44 <JibberJim> yeah that was one of the ways the designer wants to take it.

17:00:49 <mdupont> and layout

17:00:56 <mdupont> what is it written in?

17:01:44 * mdupont looks for the source of foafanaught

17:01:55 <libby> --------------- rdf calendaring chat ----------------

17:02:25 <libby> sorry jim....

17:02:38 <mdupont> ok

17:02:40 <mdupont> have fun

17:02:47 <libby> you could always move to #foaf for a bit....

17:02:49 * mdupont leaves the experts to discuss calenders

17:03:02 <LotR> lots of people not here it seems

17:03:11 <libby> yeah

17:03:23 <libby> I'm not sure if andrea and benjamin are coming along

17:03:33 <libby> Ol's here - hi Ol!

17:03:45 <libby> garyfreder is here :)

17:03:51 <garyfreder> howdy

17:03:57 <libby> howdy

17:04:04 <LotR> oh right. missed Ol's join

17:04:20 * DanC was thinking of taking roll with pointers to foaf.rdf documents, got sidetracked into a zillion fun foaf toys

17:04:29 <Ol> Hi

17:04:30 <libby> ok, well, let's start a participants chump

17:04:35 <libby> heh, that would be cool

17:04:49 <libby> hey gk

17:04:52 <DanC> participants: nah; the last one is still fresh enough, no?

17:04:56 <gk> Hi Libby

17:05:15 <libby> ok, just wonndering who was here specificall for the chat

17:05:27 <DanC> D:I'd like to talk about [value=duration and value=uri|http://lists.w3.org/Archives/Public/www-rdf-calendar/2003Jan/0022.html]

17:05:28 <dc_rdfig> Added comment D5.

17:05:44 <timbl_> You could ask. I am specifically here to lurk the calednaring.

17:05:58 <timbl_> (libby could ask who is here)

17:06:09 <libby> ok: who is here specifically for calendaring chat/lurking?

17:06:10 * DanC hasn't studied swad-eu stuff http://lists.w3.org/Archives/Public/www-rdf-calendar/2003Jan/0005.html ... considers asking for it to be on the agenda

17:06:21 <amy> I am here for calendaring

17:06:30 * maxf too

17:06:33 <LotR> libby: me

17:06:36 <libby> hi amy, glad you could make it :)

17:06:42 <libby> cool LotR

17:06:45 <amy> hi Libby :)

17:06:46 * gk too, but in a lurking kind of way

17:06:46 <DanC> and if you're here for calendaring, any particular agenda requests? (use D: if you like)

17:07:03 <garyfreder> I'm here for calendering and would like to talk about what DanC suggested

17:07:06 <libby> I can talk about swad-e calendaring, but all demos are down due to disk crash :(

17:07:13 <DanC> ew.

17:07:23 <libby> hey terryP

17:07:24 <maxf> awww

17:07:33 <libby> hehe

17:07:34 <TerryP> Hello All

17:07:42 <libby> agenda items here: http://rdfig.xmlhack.com/2003/01/22/2003-01-22.html#1043242188.201788

17:07:49 <libby> please add more if interested

17:07:55 <DanC> ... by typing D: ...

17:08:25 <DanC> did somebody do a howto?

17:08:35 <DanC> ah yes... http://tipi.sourceforge.net/doc/recur/

17:08:41 * LotR thinks we need to do the 'decide on agenda' bit earlier next time

17:08:53 <gk> d:My recent interest is in recurring schedule, for access control

17:08:53 * DanC nominates LotR to chair next time

17:08:57 <libby> D:Bill kearney just brought up a problem importing some kinds of apple ical files into outlook (on email though)

17:08:58 <dc_rdfig> Added comment D6.

17:09:01 * garyfreder seconds

17:09:16 <libby> gk, you need a capital D

17:09:28 <gk> D:My recent interest is in recurring schedule, for access control

17:09:28 <dc_rdfig> Added comment D7.

17:09:34 <libby> sorry about that - would have pinmged people ealier but for aforementioned diskcrash

17:09:36 <gk> thanks Libby

17:11:05 <garyfreder> can we start with "DanC: I'd like to talk about value=duration and value=uri "

17:11:16 <libby> sounds like a plan

17:11:25 <libby> DanC?

17:11:29 <DanC> yes...

17:11:44 <DanC> well, the question and the answer I chose are there in http://lists.w3.org/Archives/Public/www-rdf-calendar/2003Jan/0022.html

17:12:46 <gk> I don't understand the problem that's being solved here

17:13:03 <DanC> it feels wierd to me that ACTION:AUDIO turns audio into a symbol/URI, but ATTACH;VALUE=URI:Ping turns Ping into a string.

17:13:42 <DanC> er... the problem isn't very subtle; it's just: how do we want to represent ATTACH;VALUE=URI:Ping in RDF?

17:13:53 <libby> ble

17:13:54 <DanC> is iCalendar really a specialization of mimedir?

17:14:02 <timbl_> Could you fill me in on what aattach:value=uri:ping is supposed to mean?

17:14:07 * DanC doesn't grok "ble"

17:14:11 <libby> what does the colon mean?

17:14:17 <LotR> did anyone confirm the ATTACH:Ping is valid iCalendar?

17:14:31 <libby> good point

17:14:36 * gk goes off to look at iCalendar spec...

17:14:38 <DanC> er... well, ATTACH;VALUE=URI:Ping was spit out by apple:iCal, I gather.

17:14:40 <Ol> The iCalendar definition is

17:14:41 <Ol> attach = "ATTACH" attparam ":" uri CRLF

17:14:43 <libby> yep

17:14:56 * timbl_ wonders whether libby wrote it or generated it

17:15:02 <libby> generated it

17:15:36 <DanC> I think ATTACH;VALUE=URI:Ping along with ACTION:AUDIO and TRIGGER;VALUE=DURATION:-PT5M means "play the contents of <Ping> 5 minutes before this event"

17:15:36 <Ol> another example from the spec ATTACH:CID:jsmith.part3.960817T083000.xyzMail@host1.com

17:15:51 <libby> ah, that makes sense

17:16:02 * DanC never understood why iCalendar syntax has URI schemes in uppercase... CID: , MAILTO:

17:16:07 <LotR> yes. doesn't the CID make it an uri?

17:16:21 <libby> CID = calendar ID?

17:16:42 <DanC> no, CID = content ID. http://www.w3.org/

17:16:48 <DanC> er... http://www.w3.org/Addressing/schemes#cid rather

17:17:21 <LotR> 4.3.13 has uri = <as defined by any IETF RFC>

17:17:29 <garyfreder> gk: did you find it in the spec

17:18:04 * DanC notes path from http://www.w3.org/2002/12/cal/ to http://www.imc.org/rfc2445 seems too long

17:18:20 <gk> sort of, but still checking

17:18:21 * DanC never mind... it's in the 1st para

17:19:05 <gk> DanC, wouldn't http://www.ietf.org/rfc/rfc2445.txt be more definitive?

17:19:07 * DanC reads 4.8.1.1 Attachment in http://www.ietf.org/rfc/rfc2445.txt

17:19:17 <DanC> yes, that's the one we cite, gk

17:19:32 <LotR> so I'd say ATTACH:Ping is not valid (the VALUE=URI bit is just repeating the default)

17:20:13 <DanC> you don't suppose relative URIs are alowed, huh?

17:20:35 <DanC> Ol, any clues on what apple:iCal meant by ATTACH:Ping?

17:20:41 <LotR> how would you ever know relative to what

17:20:52 <DanC> relative to the file you found it in, as usual.

17:20:53 <Ol> lotr: to the ics file itself ?

17:21:10 <Ol> I don't think 2445 specifically forbids relatives

17:21:11 <LotR> there's probably a system sound called Ping

17:21:13 <timbl_> There seems to be some assumption that there is a base URI which has near it a bunch of usual sounds.

17:21:25 <DanC> ew.

17:21:58 <Ol> danc: I'll check with who write this, but I think it's relative to the base of system sounds

17:22:04 <Ol> wrote

17:22:09 <LotR> Ol: so the ping sound gets exported if you publish your calendar?

17:22:31 <DanC> please do check. that'll be great. Meanwhile, I think I'll change ical2rdf.pl to complain about relative URIs.

17:22:43 <DanC> relative URI references, that is.

17:22:56 <Ol> lotr: exported as this, not as binary attachment

17:23:03 <garyfreder> ATTACH:Ping is valid 2445, just is unclear where it points

17:23:04 <dc_rdfig> Label ATTACH not found.

17:23:04 <Ol> unless you choose to filter alarms

17:23:14 <libby> I seem to recall that moz cal understood that there was an alarm from that apple:ical file

17:23:40 <Ol> libby: yes, I think so, but the question is how they interpret the attachment

17:23:44 <DanC> I read it as invalid. 2445 say "URI as defined in IETF RFCs" and the IETF RFCs (2396, in particular) say that the term 'URI' doesn't include relative references.

17:24:02 <LotR> maybe it is valid to the letter of the rfc, but I doubt this was the intention

17:24:22 * DanC reads 4.6.6 Alarm Component

17:24:56 <timbl_> Sounds like some std body should host a standard base directory for a bunch of souds.

17:24:59 <DanC> [[ When the action is "AUDIO", the alarm can also include one and only

17:25:00 <DanC> one "ATTACH" property, which MUST point to a sound resource, which is

17:25:00 <DanC> rendered when the alarm is triggered. ]]

17:26:01 * garyfreder wonders if Ol wants to ask the ietf calsch group about DanC s comment or if I should

17:26:19 * timbl_ Oh for a namespace. appleSounds:Ping

17:26:20 <Ol> gary: I can do that

17:26:27 <garyfreder> good :-)

17:26:35 <Ol> timbl: probably better, yes

17:27:37 <libby> shall we leave that agenda item then?

17:27:42 <timbl_> (We could slip in a few sounds to really get you ouit of bed! - ping doesn't do it for me! ;-)

17:27:43 * Ol thinking about hosting a ics file on http somewhere, with a file attachment in an event... maybe relative uri would make sense there

17:27:45 <garyfreder> duration?

17:27:50 <libby> danC, are you happy with that?

17:27:59 <DanC> test case in question now spits out: relative URI reference; not clear that this is allowed/interoperable at ical2rdf.pl line 317, <> line 21.

17:28:10 <libby> danbri had an idea about recording various alarmclock sounds in foaf files...

17:28:16 <Ol> timbl: or a file:/// URI to you mp3 library :>

17:28:20 <timbl_> Ol, any idea whether iCal would actually pick up a different sound from a URi if one were givem?

17:28:24 <gk> RFC2445 cites very old URI specs ... I recall relative URIs were clarified somewhat later

17:28:38 <LotR> Ol: isn't that what cid is for? so you can point to a different mime part

17:28:58 <DanC> happy with what, libby?

17:29:15 <Ol> lotr: if everything is in the same file, right

17:29:43 <libby> DanC: that Ol takes the question about relative uris (?) to the calsch group

17:30:08 <Ol> timbl: I guess the implementation just get the sound with the given name through NSSound, I'll check

17:30:11 <DanC> while I'm hacking the code... should it be <attach>http://...</attach> or <attach rdf:resource="http://..."/>. That latter appeals to me, but it's more than 45 seconds of work.

17:30:25 <gk> cid: is to point an MIME body part identified by Content-Id header. It's covered in the specs dealing with multipart/related (or friends thereof)

17:30:34 <DanC> yes, we owe the calsch group a problem report, and I'm happy for Ol to do it.

17:30:40 <Ol> :)

17:30:44 <libby> You wanted to talk about value=duration too

17:30:57 <timbl_> The code should probably have a concept of safe retreival for security's sake. But it would be nice to have custom sounds by URI.

17:30:57 <garyfreder> DanC: and your duration comment

17:31:03 <DanC> I'm not quite done with value=uri. should it become rdf:resource? or stay a string?

17:31:25 <libby> why not a resource?

17:31:42 <DanC> no good reason; just cuz the code doesn't grok that yet.

17:31:50 <timbl_> +1 for resource in general when a URI is given.

17:31:56 <garyfreder> is a uri a rdf:resource?

17:31:57 <DanC> i.e. all the other value= thingies turned into interpretation properties.

17:32:09 <DanC> hmm... I could use log:uri. That's contorted, though.

17:32:11 <garyfreder> I think it is

17:32:15 <timbl_> When people give URIs they mean the thing identified by it except in strange cases.

17:32:33 <DanC> ok...

17:32:54 <DanC> RESOLVED: value=uri should become rdf:resource="...". ACTION DanC: fix code, test case.

17:32:54 <dc_rdfig> Label RESOLVED not found.

17:33:13 <DanC> now, onto duration...

17:33:25 <LotR> DanC: you're aware that URI is the default value type for ATTACH, no?

17:33:29 * DanC is distracted by the smell of something really yummy here...

17:33:43 <DanC> yes, just coded that default: 'attach', 'uri', # 4.8.1.1 Attachment

17:33:54 <LotR> ok, just checking :)

17:34:42 * DanC surfs back to http://lists.w3.org/Archives/Public/www-rdf-calendar/2003Jan/0022.html

17:35:03 <DanC> timbl, can you see the bit about duration there? maybe you'd like to explain why I got it right?

17:35:05 <garyfreder> danc ponders duration

17:35:21 <DanC> TRIGGER;VALUE=DURATION:-PT5M becomes...

17:35:26 <DanC> <trigger rdf:parseType='Resource'>

17:35:26 <DanC> <duration>-PT5M</duration>

17:35:26 <DanC> </trigger>

17:36:01 <garyfreder> duration is the 'type' of the trigger

17:36:04 <DanC> in N3, that's { :someEvent ical:trigger [ ical:duration "-PT5M" ] }.

17:36:28 <libby> danc, what's your reason for not using RDF datatypes/xml schema datatypes?

17:36:52 <DanC> I think of ical:duration as an interpretation property; i.e. it relates the duration to its string representation "-PT5M"".

17:36:55 <DanC> s/""/"/

17:37:32 <DanC> i think this is the way RDF datatypes should work. The RDF Core WG cooked up something way more complex than is worthwhile, IMO.

17:38:05 * DanC was hoping timbl would join the discussion of interpretation properties. phpht. oh well, he wrote it up in http://www.w3.org/DesignIssues/InterpretationProperties

17:38:22 <libby> ok; guess we might run into problems later if everything suports datatyping and we don't take advantage of that.

17:38:30 <DanC> basically, I don't have any code that supports the RDFCore WG's new datatype syntax.

17:38:38 <garyfreder> the resource is a 'trigger' and the value is PT5M which is type duration

17:39:15 <gk> I can't find "VALUE=DURATION" in RFC2445. I do see things like "TRIGGER:-PT30M"

17:39:24 <DanC> hmm... the way I've represented it, trigger is a property that relates an event to a duration.

17:39:56 <LotR> gk: DURATION is the default valuetype for TRIGGER

17:40:07 <libby> was this example from the apple:ical file again DanC?

17:40:10 <garyfreder> duration is an attribute of that trigger (in 2445 speak)

17:40:17 <DanC> libby, we might someday run into trouble if "everything suports datatyping and we don't take advantage of that". but today, I'm certain I have a problem in that the RDF parser I use doesn't support rdf:datatype.

17:40:44 <LotR> garyfreder: no, VALUE is. DURATION is the value of the attribute :)

17:40:47 <DanC> no, garyfreder, in 2445 speak, duration is the type of the value of the trigger property

17:40:57 <garyfreder> 4.3.6 Duration

17:40:59 <garyfreder> Value Name: DURATION

17:41:00 <garyfreder> Purpose: This value type is used to identify properties that contain

17:41:02 <garyfreder> a duration of time.

17:41:21 <Ol> danc: trigger is a property that relates to the alarm component within an event

17:41:22 * libby find sthe languiage of rfc 2445 hard to understand and model in my head

17:41:28 * DanC invites timbl to jump in and explain how iCalendar value types are best seen as interpretation properties

17:41:52 * TimBL just rejoined after client crash

17:42:12 * LotR hands out stable clients :)

17:42:18 <TimBL> One sentence on interpretation properties:

17:43:00 <TimBL> if you have a weird new data type which is a particular way of interpreting a simpler type, liek string, then you make a single arc between the two.

17:43:15 * DanC should know that bad-mouthing RDF Core datatypes would cause the chair to show up. ;-)

17:43:27 <libby> heheh

17:43:37 <bwm> my ears were burning

17:44:03 <bwm> shall I leave so you chaps don't feel constrained?

17:44:06 * TimBL stablest irc client for os-x, lotr? i'm using moz chat 1.2b

17:44:35 <gk> Nah... we know you can always peek at the logs

17:44:38 <LotR> surely there's some console clients that compile on OS X?

17:44:40 * DanC doesn't feel constrained

17:44:45 <libby> no, stay bwm :)

17:44:58 <DanC> I note " We do not use RDF's datatyping mechanism in the vocabulary's schema to note that the rdfs:range of the lat,long and alt properties are XML Schema float datatypes. Instead, from RDF's point of view, the properties are simply strings." -- http://www.w3.org/2003/01/geo/

17:45:04 * gk notes he forgot to add smiley

17:45:08 * bwm was thinking the idea of him constraining anyone on RDFCore is a bit laughable

17:45:12 <DanC> and I'm starting to wonder who *is* using rdf:datatype

17:45:24 <libby> well, someone has to be first right?

17:45:32 <TimBL> So what is the current problem with this duration which is the delta between the reminder and teh event?

17:45:33 <DanC> no, not just for the sake of it.

17:45:57 <gk> I think PatrickS may be (using datatypes). Art has proposed CCPP variant that does.

17:45:58 <DanC> duration: I don't think there's a problem.

17:46:33 <DanC> But I think it's a somewhat subtle design decision, and I'd like folks to understand it and confirm that it's reasonable.

17:46:58 <TimBL> I wantto put numeric types into cwm - just a few oethr things to do to.

17:47:16 * DanC would like rdf:nodeID queued before numeric types in cwm, pls.

17:47:52 <LotR> so did we just finish the value=duration point? :)

17:47:54 * TimBL thinks that rdf:nodeID was in the sax parser which is DanC's code mostly, no? ;-)

17:48:10 <garyfreder> I am not clear but willing to defer

17:48:44 * DanC oh yeah; I guess I could add that.

17:48:47 * TimBL didn't find a clear statement of the subtly design decision let a lone a note of which way it went

17:48:52 <gk> Ditto ... (I didn't catch what the subtle design point was)

17:48:56 <garyfreder> till another time the value=duration and have DanC see how his stuff works

17:49:27 <DanC> rewinding, pls see http://lists.w3.org/Archives/Public/www-rdf-calendar/2003Jan/0022.html starting with TRIGGER;VALUE=DURATION:-PT5M

17:49:28 * garyfreder is glad I was the only one that was not subtle

17:50:34 <libby> I guess someone (me I guess) could also come up with the rdf:datatype version.

17:50:49 <gk> You mean the bit about following interpretation property style?

17:50:53 <libby> was the subtle part making it a property rather than an object?

17:51:10 <garyfreder> The xCal for the example is <trigger value='duration'>-PT5M</trigger>

17:51:16 * libby notes that lots of people have to leave in 10 minutes

17:51:25 <garyfreder> (xCal is not a standard)

17:51:39 <DanC> so the alarm component looks like this in N3: [ a i:Valarm; i:attach <Ping>; i:trigger [ i:duration "-PT5M]; i:action i:audio ]. # assuming uri decision above

17:52:09 <Ol> and you should be also able to encode that a trigger can be an absolute time (VALUE=DATETIME) or a duration (which can also be relative to event start or event end through an optional param).

17:52:10 <DanC> yes, I mean the bit about following interpretation property style.

17:52:14 <garyfreder> can you express that in XML?

17:52:23 <gk> GK contemplates <ical:trigger rdf:value="-PT5M" ical:value="duration" /> ... not seriously

17:52:43 <DanC> in XML, see http://www.w3.org/2002/12/cal/test/20030122mtg.rdf ... er.. frap; didn't I check that in?

17:53:06 <libby> frap? ;)

17:53:18 * DanC says frap cuz I have little kids

17:53:18 <TimBL> I think someone should replace <Ping> with <http://standards.apple.com/sounds/iCal1.0/Ping>

17:53:37 <LotR> f is also near enough to c to be a typo :)

17:53:43 <libby> heh

17:53:57 <DanC> using some of RDF's more obscure XML syntax, the N3 I gave above looks like <i:trigger i:duration="-PT5M"/>.

17:54:03 <gk> I'm content with interpretation property style

17:54:24 <libby> that's nicely short

17:54:45 <libby> shall we spend 5 mins talking about next meeting?

17:54:46 * DanC makes a mental note to add support for that obscure syntax in cwm's parser

17:54:59 <DanC> RESOLVED: interpretation property style OK for duration etc.

17:55:00 <dc_rdfig> Label RESOLVED not found.

17:55:16 <garyfreder> <i:trigger i:datetime="20030122"/>

17:55:31 <garyfreder> resolved?

17:55:42 <DanC> well, to my satisfaction, it's resolved.

17:55:44 <LotR> garyfreder: the value there isn't a datetime :)

17:55:48 <libby> well if gk says it's ok, then it's ok :)

17:56:01 <TimBL> (Obscure but I think it may be supported already in cwm)

17:56:12 <garyfreder> it's as best as I can do - look at the intent, not the value of the attribute

17:56:18 <DanC> it was proposed (by me), 2nded (by gk) and I didn't see anybody object (though I didn't ask ;-).

17:56:19 <libby> we can take this to the amiling list I think, if we need to argue further.

17:56:27 <garyfreder> I objected

17:56:32 <garyfreder> sorry I was slow

17:56:34 <DanC> oops.

17:56:41 <mdupont> ok guys

17:56:45 <mdupont> see you

17:57:01 <TimBL> Objection?

17:58:08 <libby> garyfreder?

17:58:08 <garyfreder> it does not represent what a duratio is

17:58:18 * DanC has no strong feelings about next meeting...

17:58:18 <garyfreder> was looking for stuff in 2445

17:59:08 <garyfreder> I would be happy if it was ok with the caveat it may be revised if garyfreder can come up with an example

17:59:58 <gk> I can see the term "duration" is a bit odd, but AFAICT of a quick skim, RFC2445 uses it in this context to mean an offset

17:59:59 <garyfreder> I think there is a subtle difference in using 2445 as it was intended and 2445 as we can in rdf

18:00:14 <garyfreder> it's a type

18:00:28 <gk> Would it be better to use an interpretation property called "offset" of similar in this case?

18:00:34 <gk> s/of/or/

18:00:38 <garyfreder> 2445 uses value

18:00:45 <garyfreder> value=duration

18:01:19 <DanC> I read "value=" as punctuation

18:01:21 <garyfreder> <trigger value='duration'>-PT5M</trigger>

18:01:43 <gk> An interpretation property works a bit like a type

18:02:11 <DanC> garyfreder, you're familiar with the way datatypes relate values to string-representations, yes?

18:02:16 <garyfreder> no

18:02:19 <DanC> hm

18:02:33 <garyfreder> which is why I was ok with what you said

18:02:42 <garyfreder> till I do become familiar :-)

18:02:48 <DanC> well, in XML Schema datatypes, they do. the decimal datatype related, e.g. the integer 10 to the string "10".

18:02:57 <DanC> s/related/relates/

18:03:23 <DanC> garyfreder, are you familiar with N3? (RDF/XML is really klunky in IRC)

18:03:29 <garyfreder> no

18:03:38 <garyfreder> but I have used 2445 :-)

18:04:03 <garyfreder> I have not had a chance to look into it

18:04:04 <DanC> any RDF experience at all, garyfreder?

18:04:11 <garyfreder> almost none

18:04:15 <DanC> ah.

18:04:27 <garyfreder> but I can trick people that are not well versed

18:04:33 * DanC is kinda stumped...

18:04:37 <libby> gary, you wrote the rdf export for moz cal right?

18:04:44 <DanC> could I ask you to skim the N3 primer real quick? (pointer pending...)

18:04:50 <gk> garyfreeder, you said "it's a type" above ..I assume you meant the DURATION in VALUE=DUIRATION

18:04:51 <garyfreder> go ahead and I will use yor stuff and then get back to you

18:04:55 <garyfreder> yes

18:04:56 <LotR> hmm. dinner bell. /me will read the log from here on

18:05:00 <DanC> Primer: Getting into RDF & Semantic Web using N3 http://www.w3.org/2000/10/swap/Primer

18:05:04 <garyfreder> based on danc's stuff

18:05:11 <garyfreder> just plug and chug

18:05:12 <libby> bye LotR

18:05:20 <garyfreder> ok

18:05:25 <garyfreder> next week

18:05:29 <DanC> ok, great.

18:05:30 <gk> ... then the interpretation property <duration> in DanC's proposal is one way of expressing that intent in RDF

18:05:42 * garyfreder thinks I should mention I am out of time --- naaa

18:05:57 <DanC> yes, we're sorta adjourning/petering out.

18:06:03 <libby> ok

18:06:11 * DanC apologizes for monopolizing the meeting

18:06:15 <libby> no, useful danc

18:06:21 <garyfreder> It's one of my talents ;-)

18:06:31 <garyfreder> and me

18:06:35 <libby> danbri passes on regrets (shoudl have done that at begiining)

18:06:46 <DanC> hard to believe one little test case, with 2 little questions, could consume the whole meeting!

18:06:50 * gk thinks the discussion was interesting, but wonders if in this medium it's important to be clearer, upfront, about exactly what points are being debated?

18:06:57 <libby> any ideas about next meeting? 1 week? 2 weeks?

18:07:09 <libby> in what way, gk? better agenda?

18:07:11 * DanC leans toward 2 weeks

18:07:31 <gk> I'm away next week -- (workshop presentation on RDF for network management)

18:07:31 <libby> yeah, 1 week comesa round too fast

18:07:34 <garyfreder> 2 weeks !!!

18:07:40 * Ol agrees

18:07:46 <libby> what gary?

18:08:06 <libby> ok, what about same time, 2 weeks time

18:08:15 * DanC checks calendar for 2 weeks hence...

18:08:43 <DanC> logger, pointer?

18:08:43 <DanC> See http://ilrt.org/discovery/chatlogs/rdfig/2003-01-22#T18-08-43

18:09:29 <libby> is that 5th Feb?

18:09:32 * DanC adds 1700Z (= 11am Chicago/KC time, right?) Wed, 22Jan to his calendar. http://calendar.sidekick.dngr.com/event?id=298

18:09:47 <DanC> phpht. yes, RDF Calendar Wednesday, February 5, 2003

18:09:58 <libby> how's that for other peopel?

18:10:03 <garyfreder> good

18:10:10 <libby> cool

18:10:33 <Ol> good

18:10:43 <libby> next ime I will bug people for agenda items sooner

18:10:44 <gk> libby, about being clearer; e.g. today, to point to the message and also highlight the questions therein up for discussion.

18:10:45 <Ol> just a quick update about the icalendar data test suite

18:10:45 <libby> cool

18:10:58 <Ol> I think we should be able to contribute some of our test suite

18:11:00 <gk> Time OK

18:11:08 <libby> good gk

18:11:16 <garyfreder> Ol: !!!

18:11:21 <libby> Ol, you're welcome to :)

18:11:26 <DanC> test data: yum yum!

18:11:29 <libby> what do you mean?

18:12:30 <gk> Libby, was that aimed at me?

18:12:48 <libby> no, Ol....

18:12:57 <Ol> oh, sorry

18:13:02 <libby> I kinda thought danC did point to his message though...

18:13:35 <libby> ol, we'd love more test data, what's the best way of you getting it to us? would it be ok to put it ojn the w3c site with th rest?

18:13:45 <Ol> libby: coverage of the properties / components we support through ics files

18:14:14 <Ol> libby: sure

18:14:14 <gk> Indeed he did, and so much was great, but it wasn;t clear to me until we were well into the discussion what particular design elements of his message were being debated. I can quite accept that's my problem, coz I'm not up to speed on iCalendar subtleties, but I got the sense I wasn't the only one.

18:14:17 <Ol> or a tarball

18:14:23 <garyfreder> regression data? Ol

18:14:32 <libby> D:[next meeting|http://www.timeanddate.com/worldclock/fixedtime.html?day=5&month=2&year=2003&hour=17&min=0&sec=0&p1=0]

18:14:32 <dc_rdfig> Added comment D8.

18:14:52 <libby> gk, ok, they were tricky I agree.

18:15:00 <libby> icalendar subtleties I mean

18:15:17 <Ol> gary: no 'real' complete regression data, but part of our test cases

18:15:32 <Ol> gary: i.e. it won't be a complete test suite for 2445 :)

18:15:53 <libby> Ol, want to put it up somewhere or email it to me and I'll put it on teh w3c site?

18:15:58 <Ol> libby: ok

18:16:07 <libby> thanks :))

18:16:08 <garyfreder> I am not up on iCalendar subtleties - anyone else here that can invite calsch gurus?

18:16:23 <gk> Oh yes... I have a sense that this work of mapping it to RDF could well have the additional effect of clarifying or fixing a meaning for the original. When I did my access control work, I relied as much on your hybrid schema as I did on the original spec.

18:16:35 <libby> really! :)

18:16:41 <libby> I totlly agree though gk.

18:17:00 <libby> so have mozilla cal got lots of testcaes as well?

18:17:17 <libby> (I noticed last week that mozilla has lots of RDF testcases, which is interesting)

18:17:24 <garyfreder> I don't think Moz has a lot of test cases

18:17:39 <gk> Yes, in the sense that in the RDF it was absolutely clear how the pieces relate to each other: like Tiggers and bouncing, isn't that what RDF schemas do best?

18:17:40 <garyfreder> they have lots of timezone data tho

18:18:19 <garyfreder> make that Moz calendar does not have a lot of test cases

18:18:44 <libby> gk, heheh. but you're right..one problem with the icalendar rfc is that it seems very syntactic (at least to me - that could be a difference in world view)

18:18:55 <libby> gary, timezone data?

18:19:14 <libby> gk, i.e. like XML, the syntax mattered, not the model.

18:19:18 <garyfreder> specifies how to 'parse' :-) a timezone

18:19:25 <gk> libby, I think you could be dead right there.

18:20:42 <gk> I figured (to some limited level) recurrence in iCalendar by looking for an example close to what I wanted, and copying that. I don't like working that way.

18:20:52 <libby> gary, that sounds useful....is that from an 'official' source

18:21:07 <garyfreder> contributions

18:21:15 <libby> gk, that's how I always work :)

18:21:19 <garyfreder> lots of contributions

18:21:47 <libby> gary, icalendar contributions? can we borrow them?

18:22:03 <garyfreder> they are probably mpl

18:22:10 <gk> I probably worry topo much, I worry what happens if things aren't just like the example.

18:22:14 <garyfreder> and they are just useful to test timezones

18:23:34 <libby> gk,I agree, it's a worry. and it's also teh reason we are working dfrom testcases to schema...because the icalendar syntax has multiple p[ossible representations in RDF because it's ambiguous.

18:23:43 <libby> gary, mpl?

18:23:44 <garyfreder> http://www.mozilla.org/projects/calendar/ look2/3 way down

18:23:44 <dc_rdfig> H: http://www.mozilla.org/projects/calendar/ from garyfreder

18:23:53 <garyfreder> Mozilla Public License

18:24:03 <libby> oh, ok, yep

18:24:52 <libby> so teher are loads and loads of holiday files of course

18:24:57 <gk> Libby, yes.

18:25:05 <garyfreder> this is NZ timezone ics http://www.mozilla.org/projects/calendar/caldata/NewZealandPublicHolidays.ics

18:25:35 <libby> ah, cool

18:28:01 <libby> I'm going to head off and get some food. thanks all for coming :)

18:31:16 <dajobe> libby: can you change the topic back, cheers

18:37:22 <libby> libby has changed the topic to: RDF and Semantic Web Hacking party, 24x7 http://rdfig.xmlhack.com/

19:14:03 * DanC can't read my own code in ical2rdf.pl ... seriously considering recoding in python

19:37:00 <DanC> hmm... calAddress... where did I get that from?

19:37:11 <DanC> did anybody look at the way I represented the attendee stuff? [[

19:37:13 <DanC> <attendee rdf:parseType='Resource'>

19:37:13 <DanC> <calAddress rdf:resource='mailto:www-rdf-calendar@w3.org'/>

19:37:13 <DanC> <cn>RDF calendar Mailing list</cn>

19:37:13 <DanC> </attendee>

19:37:13 <DanC> ]]

19:46:03 <DanC> ah... calAddress is cal-address in camelcase... it's an interpretation property... hmm...

20:00:39 * LotR reads backlog

20:04:23 <LotR> libby: can you mail me with an url when you put the apple test tarball up?

20:07:19 <LotR> garyfreder: doesn't mozilla use libical?

20:07:28 <garyfreder> yep

20:07:41 <garyfreder> as does Evolution

20:07:44 <garyfreder> and others

20:07:58 <LotR> so the mozilla timezone data is the libical timezone db?

20:08:07 <garyfreder> yes

20:08:14 <garyfreder> uses the same ical

20:10:05 <LotR> what do you mean when you say mozilla has lots of contributed timezone data then?

20:11:07 <garyfreder> look here ...

20:11:36 <garyfreder>http://www.mozilla.org/projects/calendar/

20:11:37 <dc_rdfig> I: http://www.mozilla.org/projects/calendar/ from garyfreder

20:11:47 <garyfreder> about 2/3 the way down

20:12:18 <garyfreder> is a list of calendar holiday files

20:12:26 <garyfreder> people have contributed

20:14:05 * LotR wonders what's up with the X-MOZILLA-*-DEFAULT properties

20:15:13 <garyfreder> LotR: look at http://bugzilla.mozilla.org/show_bug.cgi?id=129660

20:16:42 <soc> are there any RDF projects active that use trust metrics in a social context?

20:20:00 <LotR> garyfreder: I'm wondering why they are in there at all, not if they are valid (the ones I've seen seem to be)

20:20:30 <garyfreder> ? Peple are contributing .ics files for Holidays

20:21:00 <garyfreder> there is some conversation on the ietf calsch list about the timezone files

20:21:16 <LotR> I mean, X-MOZILLA-RECUR-DEFAULT-UNITS, I really don't see what the purpose of that is

20:21:22 <garyfreder> oh

20:21:29 <garyfreder> Don't know

20:22:11 <garyfreder> I never do anything that happens more than once --- (I have not looked at what they do)

20:22:53 <soc> birthday?

20:22:55 * soc ducks

20:23:18 <LotR> garyfreder: well, there's also X-MOZILLA-ALARM-DEFAULT-UNITS. same issue

20:23:36 <garyfreder> I think I remember

20:23:48 <garyfreder> OEOne used those internally

20:23:48 <soc> what about somebody elses?

20:23:59 <LotR> what's OEOne?

20:24:00 <garyfreder> to persist stuff

20:24:17 <garyfreder> OEone did most of the work on Mozilla calendar

20:24:40 <LotR> garyfreder: but it seems silly to store what seems to be default settings in each event

20:25:02 * soc is playing with mozcal

20:25:09 <garyfreder> soc: I don't schedule things more than once

20:25:19 <garyfreder> they were getting thing up

20:25:40 <garyfreder> made decisions that will return to haunt them ;-)

20:25:47 <soc> garyfreder: but isn't a birthday a recurring event?

20:26:04 <garyfreder> but I don't schedule it - it is thrust upon me

20:26:35 <soc> how do you remember when to give gifts to people then?

20:26:52 <LotR> maybe he doesn't :)

20:27:08 <garyfreder> I suggested again on the ietf calsch mailing list that someone that actually knows the spec get on the next chat

20:27:10 <LotR> or only when he gets invited

20:27:34 <garyfreder> yep - no gifts, unless they are for me

20:28:40 <garyfreder> argh - locked the browser up

20:28:50 <soc> mozcal?

20:29:29 * soc experiences mozilla hangs daily, mozcal or not

20:30:11 * LotR happily uses lynx (and galeon when in X)

20:34:15 * LotR hopes garyfreder just left to restart mozilla

20:49:58 <LotR> libby: did you see my request?

20:50:32 <libby> yes but it went immediately out of my brain!

20:50:48 <libby> yes I will.

20:51:27 <libby> wanna send me your email address?

20:53:32 <libby> ta

20:54:57 * LotR wonders if there should be a #icalendar

20:55:03 <libby> ooh, it's all go on hthe ietf calendar list

20:55:08 <libby> probably...

20:55:48 <libby> good of gary to send his message round too

20:56:03 <LotR> yeah

20:56:13 <LotR> I hope Ol sends his query in soon :)

20:56:23 <libby> yeah :)

20:57:46 * LotR cut down the 4000+ ietf-calendar archive down to ~120 messages on recurrence

20:58:12 <LotR> now to (re)read all of those and compress them into some sections for the tutorial

20:58:17 <libby> oouch

20:59:11 <soc> soc is now known as soc|away

21:01:13 * LotR hopes he has the tuits to do a 'kernel cousin' thing for other topics too

21:03:40 <DanC> those round tuits... hard to find.

21:03:55 <LotR> yeah :(

21:05:38 <libby> lotr, a recurrence tutorial will be really really useful

21:08:39 <LotR> libby: I was hoping what I have so far was a good start :)

21:09:00 <libby> yes, sorry, I didn;t mean it wasn;t :)

21:09:36 <libby> I just reaslised how diffiocult people find recurrence, that's all.

21:09:38 <LotR> well, I'm sure there's lots of ways to improve it

21:09:52 <LotR> recurrence *is* hard :)

21:10:10 <libby> yeah...

21:10:40 <libby> I haven;t loked at it in any detail, yet.

21:10:45 <LotR> too bad we didn't get to the recurrence point of the agenda

21:11:26 <libby> well, bagsy a slot now, and you can be number one next time :)

21:12:49 <LotR> I'm not sure I have much to say. I'd be more than happy to (try to) answer any questions people have after reading the tutorial

21:13:49 <libby> so maybe some mails to the list beforehand then....and some testdata

21:14:34 <LotR> I'm not on rdf-calendar. and I don't really want to be

21:15:05 <libby> if you send me stuff, I can forward it.

21:15:10 <DanC> aw, come on; it's a cool place to be... everybody's doing it! ;-)

21:15:22 <mrlc> hi all

21:15:48 <bijan> DanC, you want a photo from me?

21:16:15 <DanC> yes, of the owl.owl award ceremony

21:16:29 * bijan fires up iPhoto

21:16:41 <libby> hi mrlc

21:16:51 <LotR> DanC: I'm sure it's mostly about rdf, which I'm really not very interested in

21:17:37 <DanC> "you will be," he said, in that yoda voice. ;-)

21:18:06 <libby> heheh

21:18:12 <DanC> crud... ruined the joke... in IRC, the way to do that is...

21:18:15 <LotR> DanC: perhaps :)

21:18:16 <DanC> DanC is now known as Yoda

21:18:19 <Yoda> You will be

21:18:20 <Yoda> Yoda is now known as DanC

21:18:25 <libby> heheh

21:18:42 <LotR> or <voice type="yoda">you will be</voice>

21:19:43 * DanC checked in value=uri fix

21:19:53 <DanC> ical2rdf.pl

21:19:53 <DanC> new revision: 1.6; previous revision: 1.5

21:20:59 <LotR> where is ical2rdf.pl again?

21:22:30 <libby> http://www.w3.org/2002/12/cal/test/

21:34:02 <DanC> no, the .pl is in /cal/ not /cal/test/

21:34:10 * DanC updates http://www.w3.org/2001/sw/WebOnt/ftf5 with owl.owl award photo

21:34:42 * LotR sees while (($line = <>) =~ s/^s+//), and ponders if that is correct

21:36:55 * DanC apologizes for nasty perl habits

21:37:16 <LotR> it isn't about the perl syntax

21:37:44 <DanC> oh... it's about iCalendar linebreak conventions?

21:37:49 <LotR> I just think you're not supposed to remove more than just the first space char

21:39:44 <Morbus> mmmmM, perlllL.

21:40:11 <Morbus> gasp, no -w! ;)

21:40:43 <DanC> you're right... "Any sequence

21:40:43 <DanC> of CRLF followed immediately by a single linear white space character

21:40:43 <DanC> is ignored"

21:41:45 <LotR> well, partially right then :) I thought it had to be \ , but \s seems to be correct

21:42:40 <Morbus> is anyoen actively using Apple's iCal? is 1.02 any better?

21:42:57 <Morbus> i'm still using pandocalendar.

21:46:18 * DanC linked reefknot test data from http://www.w3.org/2002/12/cal/

21:46:24 <DanC> libby, wanna link that howto thingy?

21:48:34 <libby> lotr's tutorial?

22:10:59 <mdupont> hi

22:17:08 <DanC> yeah, libby, that thing

22:20:34 * LotR should find out why the stylesheets don't include his email addy in the header

22:22:34 <LotR> DanC: if you link it, please add a note saying it is a work in progress

22:37:54 * DanC adds note... v 1.15 2003/01/22 22:37:39