Semantic Web Interest Group IRC Chat Logs for 2003-04-23

06:06:09 <tav> tav is now known as tav|offline

08:44:25 <sbp`-> sbp`- is now known as sbp`

11:35:03 * nb is away: afk

12:14:22 <arnarl> hi

13:33:24 * libby_ demos IRC to Ben

13:36:21 <MorbusIff> MorbusIff is now known as Morbus

13:44:33 <dajobe> dajobe has changed the topic to: Semantic Web wish factory, http://rdfig.xmlhack.com/ 'When you go meta, don't forget to come back' | RDF Calendar meeting 2003-04-23 1600 UTC +90mins

14:06:00 <DanC> high irony: I just got a surface-mail ad from the C/C++ users journal. "Q: Quick! What's wrong with this code? class Ff { ... }" is the teaser on the envelope

14:06:08 <DanC> my answer: It's written in C++!

14:06:25 <Morbus> hey, i got that too.

14:06:31 <Morbus> and you know what, that's *exactly* what I said.

14:06:37 * Morbus ^5's DanC.

14:06:50 * DanC returns ^5

14:08:21 <deltab> hehe

14:09:49 <jang> I have to grudgingly agree

14:09:56 <jang> a language should not be a logic problem

14:10:27 <jang> C++ has too many different things that look alike.

14:11:17 <DanC> the irony is so intense... presumably a C/C++ journal would tout the *benefits* of a language.

14:11:40 <DanC> but they're using the *problems* of the language to get the attention of prospective subscribers.

14:12:04 <deltab> I read something that claims C++ is popular because it gives programmers something to fiddle with while they're thinking about hard problems

14:12:18 <jang> C++ has a refreshingly start attitude to inheritance ambiguity: it's an error

14:12:26 <jang> s/start/stark

14:12:36 <jang> however, there's a rgeat deal of linguistic ambiguity there

14:13:29 <jang> it comes from all over the shop: the attempt to keep C's neat "declaration looks like use" syntax

14:13:57 <jang> but when the number of ways to use things expand, the number of potential things that a statement/declaration might be goes up

14:14:31 <jang> it's pleasant to use C++ to write "scripting" tasks: _using_ class libraries that are well-designed (eg, the boost stuff) is rather good

14:14:42 <jang> writing the bloody things in the first place is a nightmare

14:50:13 <DanC> .time

14:50:13 <datum> Wed, 23 Apr 2003 14:50:13 GMT

14:51:25 <DanC>http://lists.w3.org/Archives/Public/www-rdf-calendar/2003Apr/0024.html

14:51:25 <dc_rdfig> A: http://lists.w3.org/Archives/Public/www-rdf-calendar/2003Apr/0024.html from DanC

14:51:41 <DanC> A:|agenda RDF calendar IRC meet 2003-04-23, 1600 UTC

14:51:42 <dc_rdfig> Titled item A.

15:29:27 <shaggy> heya

15:29:41 <shaggy> I've got a quick build question...

15:29:55 <shaggy> I've noticed that there are 2 librdf's built...

15:30:06 <shaggy> one with a T on the end? what's up with that?

15:30:57 <shaggy> that one segfaults on an embedded linux platform I'm working on, the other one doesnt. Just wondering.

15:31:43 <libby_> hm, that looks like something of dajobe's

15:32:03 <libby_> libby_ is now known as libby

15:32:18 <dajobe> eh?

15:32:48 <dajobe> shaggy: you'll need to give a little more context for me

15:33:06 <dajobe> .google librdft

15:33:07 <datum> librdft: sorry, no results were found.

15:33:14 <dajobe> ^- scary

15:35:06 <shaggy> no, check it out..

15:35:25 <shaggy> -rwxrwxr-x 1 buildmas buildmas 372153 Apr 23 15:14 librdf.so.0.0.0

15:35:25 <shaggy> -rwxrwxr-x 1 buildmas buildmas 372089 Apr 23 15:14 librdf.so.0.0.0T

15:35:44 <shaggy> that's the contents (partial) of librdf/.libs

15:35:55 <dajobe> ah libtool

15:36:02 <dajobe> firstly, don't ever look in there ;)

15:36:03 <shaggy> I was just wondering why the T one gets installed

15:36:09 <shaggy> <grin>

15:36:19 <dajobe> what system are you running on? uname -a would help

15:36:27 <shaggy> sure...

15:36:41 <dajobe> I've not seen that and I've built redland on all sorts of strange systems such as OSX

15:36:55 <shaggy> Linux buildmaster 2.4.7-10 #1 Thu Sep 6 17:27:27 EDT 2001 i686 unknown

15:37:07 <dajobe> oh, regular linux x86

15:37:19 <shaggy> this is our build machine. eventually the binaries go onto an embedded SBC

15:37:27 <dajobe> ah, cross compiled

15:37:34 <dajobe> I suspect T=target binary then

15:38:11 <shaggy> ahhh...

15:38:35 * dajobe guessing

15:38:40 <shaggy> ok, I was just wondering. one of my co-workers said you were very helpful and available via IRC.

15:38:48 <shaggy> keep up the good work. cheers!

15:38:56 <dajobe> thanks, a user, yay

15:39:06 <qmacro> qmacro is now known as q|eve

15:39:23 <shaggy> later all.

15:47:16 <dajobe>http://www.guardian.co.uk/online/news/0,12597,942024,00.html

15:47:16 <dc_rdfig> B: http://www.guardian.co.uk/online/news/0,12597,942024,00.html from dajobe

15:47:41 <dajobe> B:|Battle of the blog builders - new movable type hosting services

15:47:41 <dc_rdfig> Titled item B.

15:47:54 <dajobe> B:including photos, list of favourite sites, books, ...

15:47:54 <dc_rdfig> Added comment B1.

15:48:49 <danbri> B:ooooh, that's interesting...

15:48:49 <dc_rdfig> Added comment B2.

15:52:15 * danbri has rsi, might not type much this meeting :(

15:53:06 <danbri> B:whee, "Other features include real-time statistics, posting by email, and automatic creation of Friend of a Friend data - instantly taking an experimental standard and taking it to the mainstream."

15:53:07 <dc_rdfig> Added comment B3.

15:54:05 <libby> :)

15:56:01 <libby> 5 mins till meeting - I'd better get chumping.....

15:56:38 <libby>http://www.w3.org/2002/12/cal/

15:56:39 <dc_rdfig> C: http://www.w3.org/2002/12/cal/ from libby

15:56:51 <libby> C:|RDF Calendar Workspace

15:56:51 <dc_rdfig> Titled item C.

15:57:37 <libby> C:meeting here [http://www.timeanddate.com/worldclock/fixedtime.html?day=23&month=4&year=2003&hour=16&min=0&sec=0&p1=0|2003-04-23 1600UTC]

15:57:38 <dc_rdfig> Added comment C1.

15:57:51 <DanC> libby, see A

15:57:58 <libby> C:[http://lists.w3.org/Archives/Public/www-rdf-calendar/2003Apr/0027.html|revised agenda]

15:57:59 <dc_rdfig> Added comment C2.

15:58:02 <libby> oops

15:58:24 <libby> heyho

15:58:51 <DanC> I'd rather not reduce the old business to one agendum.

15:59:02 <DanC> I like to review actions in the context of the agendum in which they're assigned.

15:59:13 <DanC> so let's put x-properties and namespaces on the agenda

15:59:17 <DanC> etc.

15:59:24 <libby> ok

15:59:44 <DanC> scheduling the next meeting is non-trivial, to me.

15:59:54 <DanC> wanna do that toward the beginning or toward the end?

16:00:15 <libby> BLURB: RDF calendar agenda item (D) x-properties and namespaces

16:00:15 <dc_rdfig> D: RDF calendar agenda item (D) x-properties and namespaces from libby

16:00:24 <libby> we normally do it at the end....

16:00:30 <DanC> ok

16:00:52 <libby> yeah, actually I don;t think I can make the next one as originally scheduled

16:01:04 <DanC> D:|x-properties and namespaces (agenda item D)

16:01:04 <dc_rdfig> Titled item D.

16:01:19 * DanC plays with agenda notations

16:01:41 <libby> BLURB: cal/geo usecase 1: opening hours of a thing that has a place (agenda item E)

16:01:42 <dc_rdfig> E: cal/geo usecase 1: opening hours of a thing that has a place (agenda item E) from libby

16:02:32 <DanC> this one next? http://esw.w3.org/topic/EventDiscovery

16:02:49 <DanC>http://esw.w3.org/topic/EventDiscovery

16:02:49 <dc_rdfig> F: http://esw.w3.org/topic/EventDiscovery from DanC

16:03:07 <DanC> F:|EventDiscovery, sources of calendar data (agenda item F)

16:03:08 <dc_rdfig> Titled item F.

16:03:23 <libby> BLURB: useful vocbularies for events (e.g. location, topic...) (agenda item G)

16:03:24 <dc_rdfig> G: useful vocbularies for events (e.g. location, topic...) (agenda item G) from libby

16:03:54 <DanC> hmm... G seems a little open-ended. we really want to talk about how to put our ical vocab into RSS, no?

16:04:09 <libby> BLURB: authoritativeness and completeness of data for scheduling purposes (agenda item H)

16:04:10 <dc_rdfig> H: authoritativeness and completeness of data for scheduling purposes (agenda item H) from libby

16:04:29 <libby> yeah, G is open ended, need not just be RSS vocab though, could be dublin core for example

16:04:50 <DanC> well, it could be; but it's most likely gonna be RSS.

16:04:53 <DanC> no?

16:05:00 <DanC> i.e. let's focus. let's make real test cases.

16:05:12 <DanC> if the test case ends up being dublin core in stead, very well.

16:05:28 <libby> BLURB: round-tripping RDF ical tests, toIcal.py (agenda item I)

16:05:29 <dc_rdfig> I: round-tripping RDF ical tests, toIcal.py (agenda item I) from libby

16:05:37 <libby>http://esw.w3.org/topic/InterpretationProperties

16:05:38 <dc_rdfig> J: http://esw.w3.org/topic/InterpretationProperties from libby

16:05:43 <danbri> did we start, you folks want to hack agenda a bit more?

16:05:52 <libby> J:|modelling timezones

16:05:53 <dc_rdfig> Titled item J.

16:05:54 <DanC> we're starting, yes.

16:06:03 <libby> yep, sorry, bit behind with the chumping

16:06:09 <DanC> we're starting by building/reviewing agenda. oh... let's do some attendance too...

16:06:21 <DanC> A:[DanC|http://www.w3.org/People/Connolly/] attending

16:06:21 <dc_rdfig> Added comment A1.

16:06:45 <libby> A:[Libby Miller|http://ilrt.org/people/libby] attending

16:06:45 <dc_rdfig> Added comment A2.

16:06:47 <eikeon> A:[eikeon|http://eikeon.com/] attending

16:06:47 <DanC> DanC has changed the topic to: RDF Calendar meeting 2003-04-23 1600 UTC http://rdfig.xmlhack.com/

16:06:48 <dc_rdfig> Added comment A3.

16:07:08 <DanC> I nominate libby to chair (noting that turn-about is fair play, and she's free to nominate me)

16:07:14 <danbri> Re RSS 'vs' DC, RSS statements (title/description/uri) _imply_ DC statements, via subPropertyOf in the RSS 1.0 schema.

16:07:29 * danbri nominates libby

16:07:33 <DanC> I'm also willing to fly chair-less

16:07:55 <libby> ok, I'll chair, and pass the floor to Danc for the first few items on the agenda if that's ok

16:08:00 <mattb> A:[Matt Biddulph|http://www.hackdiary.com/] attending

16:08:00 <dc_rdfig> Added comment A4.

16:08:16 <libby> i.e. D and E

16:08:17 <DanC> J:|InterpretationProperties, esp as applied to timezones (agenda item J)

16:08:17 <dc_rdfig> Titled item J.

16:08:32 <libby> thanks for coming along mattb, eikeon

16:08:39 <DanC> great, libby.

16:09:07 * DanC is tempted to put SUMO update on the agenda... seems over-full already though.

16:09:08 <libby> ok, danc, want to talk about x-properties and namespaces (agenda item D)?

16:09:35 <DanC> right... well, we decided that my --xnames command-line option sucks ;-)

16:10:07 <DanC> we sketched a design (using hashes and such), and libby took an action to give it a go in code/tests. Any progress?

16:10:26 <libby> no, I'm sorry, that one got away

16:10:42 <libby> I think I got distracted into building the schema using cwm

16:10:45 <danbri> A:[Dan brickley|http://www.w3.org/People/DanBri/] attending

16:10:45 <dc_rdfig> Added comment A5.

16:10:55 <libby> we can carry it forward

16:10:59 <DanC> D:continued from [26 Feb|http://rdfig.xmlhack.com/2003/02/26/2003-02-26.html#1046279854.884486]

16:10:59 <dc_rdfig> Added comment D1.

16:11:35 <DanC> D:ACTION libby: to build/test what we decided on 26Feb, continued

16:11:35 <dc_rdfig> Added comment D2.

16:11:51 <DanC> I'll say that the --xnames hack continues to suck.

16:12:22 <DanC> actively. i.e. it continues to cost me. So I'm very pleased that you're willing to pick up work on it.

16:12:44 <DanC> I might find a round tuit meanwhile, but I dunno.

16:12:49 * libby hopes it's not too much work .... ;)

16:12:50 <DanC> Maybe I can get ryan lee to do it.

16:13:05 <DanC> anything else on x-properties?

16:13:36 <libby> not from me - shall we move to cal/geo usecase 1: opening hours of a thing that has a place (E)?

16:13:59 <DanC> E:continued from [9Apr|http://rdfig.xmlhack.com/2003/04/09/2003-04-09.html#1049903200.904995]

16:14:00 <dc_rdfig> Added comment E1.

16:14:10 <DanC> any progress on "ACTIOn libby write a converter from chefmoz opening housr format to rdf calendar format (or icalendar format maybe?)"?

16:14:22 <libby> no, still nothing... :(

16:14:32 <DanC> would you prefer to continue or withdraw?

16:14:37 <libby> I stil think it should be relatively simple, still not done it

16:14:43 <libby> continue is fine

16:14:47 <danbri> woudl you like help?

16:14:52 <libby> sure :)

16:14:55 <DanC> yeah, any volunteers?

16:15:06 <danbri> I can do everything except the function that takes the chefmoz property and emits the RDF...

16:15:19 <danbri> ie download scrub and emit the data, for some subset...

16:15:22 <DanC> elaborate, danbri?

16:15:39 <DanC> "function that takes the chefmoz property..." ... example, please?

16:15:49 <danbri> Libby writse a perl/ruby/java/python function that takes the funny opening hour strings from chefmoz, and returns the appropriate rdf/xml

16:16:10 <libby> all of them!? ;)

16:16:15 <danbri> i'll wrap that in something that actually emits a bunch of RDF/XML restaurant reviews with this bit added in.

16:16:20 <DanC> I've been writing N3 rules using string:scrape to do string munging of that sort, danbri. interested?

16:16:25 <libby> that'd be great danbri

16:16:52 <danbri> I think that'd be a steep learning curve for me DanC... (though I'm learning a bunch from studying Cwm's APIs)

16:17:17 <danbri> we want one owner for the action (libby hopefully) but do name me in the action writeup so i remember it.

16:17:19 <DanC> E:ACTION libby: write a converter from chefmoz opening housr format to rdf calendar format (or icalendar format maybe?). continues, with danbri helping

16:17:19 <dc_rdfig> Added comment E2.

16:17:44 <DanC> learning: right. you guys seem to have a handle on it.

16:17:54 <mattb> re zool

16:18:16 <DanC> what news on "ACTION danbri_fr to have a draft of the 'concentric circles' within100m schema/vocab for next week"?

16:18:29 <libby> hey zool

16:18:33 <DanC> I think i saw progress on that...

16:18:42 <zool> re mattb libby

16:18:49 * DanC considers indicating the current agenda item in the /topic

16:19:18 <libby> re danbri's action, tehre's http://esw.w3.org/topic/GeoOnion

16:19:26 <libby> but there's also some more discussion somewhere

16:19:38 <eikeon> Are there pointers to the chefmoz bits... how to get the string that needs converting.

16:20:00 <DanC> E:re 'concentric circles', see [esw:GeoOnion|http://esw.w3.org/topic/GeoOnion]

16:20:01 <dc_rdfig> Added comment E3.

16:20:19 * danbri saves http://esw.w3.org/topic/GeoOnion

16:20:25 <DanC> danbri, do you claim victory on 'concentric circles'? I guess it's done to my satisfaction, but I wasn't there 9Apr

16:20:42 <danbri> I submit that I'm stuck. I just added some more explanatory prose, but the scale isn't really usable.

16:21:04 <danbri> I feel blocked and mathematically clumsy. If someone could help with calculating a better sscale, I'd take further actions on this.

16:21:12 <DanC> how about writing some examples?

16:21:13 <danbri> But for now I think I should declare victory.

16:21:22 <danbri> I just started to ;)

16:21:40 <DanC> { :hotelWhereImStaying :within100metersOf :sandwichShop } # walking distance

16:21:51 <danbri> OK, I'll paste that in :)

16:22:05 <DanC> strike the {}s

16:22:12 <DanC> put . before #

16:22:28 <DanC> (perhaps link to ["ShorthandRDF"])

16:22:43 * DanC checks time; 20min into 90min slot

16:23:04 <libby> was that all the actions for review danc?

16:23:11 * libby thinks so

16:23:35 * danbri updates http://esw.w3.org/topic/GeoOnion

16:23:40 <zool> eikeon, re chefmoz, grubstreet latest work here http://un.earth.li/~kake/cgi-bin/wiki.cgi?action=rss&feed=chef_dan&id=Calthorpe_Arms%2C_WC1X_8JR if not seen

16:23:56 <DanC> yes, there are no more actions, but discussion of "opening hours of a thing that has a place " is in order.

16:24:21 * danbri ooohs at http://un.earth.li/~kake/cgi-bin/wiki.cgi?action=rss&feed=chef_dan&id=Calthorpe_Arms%2C_WC1X_8JR/

16:24:29 <danbri> ...though not sure if changed since i last looked

16:24:37 <libby> danc, where's that from?

16:24:41 <eikeon> zool: thank you.

16:24:42 <danbri> I feel foaf:topic is working out rather nicely, btw

16:24:47 <libby> ok got it

16:25:16 <DanC> yes, foaf:topic++ (except for SlashVsHash ;-)

16:25:24 <libby> danc, anything in particular re "but discussion of "opening hours of a thing that has a place " is in order."

16:25:29 <libby> ?

16:25:48 <danbri> Oh also should mention that DanC reworked the GeoInfo page rather nicely, which re 'things with a place' might interest various folks here -- http://esw.w3.org/topic/GeoInfo

16:26:29 <DanC> somebody this E thing on our agenda 9Apr. actions were assigned, which causes the item to be carried forward until they're complete or withdrawn. We've reviewed the actions, but we can continue to discuss the item until folks want to move on.

16:26:32 <eikeon> SlashVsHash looks good to me ;)

16:26:36 <DanC> somebody put...

16:26:53 <danbri> An idea for soemthing re things, place, opening hours... identify something that is in both ChefMoz and Grubstreet, and use that to show augmentation (via merging or whatever) of data from the two sources.

16:27:19 <DanC> logger_1, pointer?

16:27:19 <DanC> See http://ilrt.org/discovery/chatlogs/rdfig/2003-04-23#T16-27-19

16:27:33 <libby> are you volunteering danbri? sounds like an excellent idea...

16:27:46 <DanC> E:[irc log|http://ilrt.org/discovery/chatlogs/rdfig/2003-04-23#T16-27-19] includes more pointers and discussion

16:27:47 <dc_rdfig> Added comment E4.

16:27:49 <danbri> ah, there is geo stuff in http://un.earth.li/~kake/cgi-bin/wiki.cgi?action=rss&feed=chef_dan&id=Calthorpe_Arms%2C_WC1X_8JR now, i forget it wasn't there before.

16:28:05 <danbri> anyone know how many lat/longs in there? is it for every pub/restaurant?

16:28:17 * DanC gets wierd HTTP foo at earth.li

16:28:36 <DanC> no, got wierd HTTP foo somewhere else...

16:28:50 <danbri> I volunteer but won't take an action beyond the 'help you' one above, will just try to use examples that merge with Grubstreet coverage.

16:28:57 <mattb> i may have scuttered it, can check...

16:29:06 <libby> ok, cool danbri

16:29:34 <libby> ok, I propose that we move to the next agenda item, as we have many

16:29:35 <DanC> E:|cal/geo usecase: opening hours of a thing that has a place (item E)

16:29:35 <dc_rdfig> Titled item E.

16:29:44 * DanC doesn't think the usecase number (1) is useful

16:29:54 <DanC> right, moving on

16:30:18 <libby> so, agenda item F, EventDiscovery

16:30:30 <danbri> wikinames might work well for use cases, as they're meeting independent. Using Chump ID, 'E:' etc is confusing over several meetings...

16:30:54 <danbri> logger_1, pointer?

16:30:54 <danbri> See http://ilrt.org/discovery/chatlogs/rdfig/2003-04-23#T16-30-54

16:31:14 <DanC> yes, wiki topic work well for agenda that last for more than one meeting

16:31:26 <DanC> topic

16:31:27 <DanC> topics

16:31:32 * DanC note to self: learn to type

16:31:45 <libby> re EventDiscovery, who here has a harvester?

16:31:57 * eikeon started eikster.com in the spirit of "and to encourage it to be put out there!" and started an events section.

16:32:33 * eikeon does

16:32:33 <DanC> well, I don't have a harvester, but I did some work to connect my foaf "home KB" to my travel schedule. I think that was a sort of wishlist (not quite action) as of a few meetings ago.

16:32:54 <mattb> i have http://www.hackdiary.com/archives/000030.html

16:33:01 <DanC> I'd like to know if my travel schedule and itineries are getting scuttered

16:33:05 <eikeon> DanC: Are these some of your events showing up: http://eikster.com/2003/events/

16:33:08 <libby> that the travel tools page danc?

16:33:25 * libby has a scutter too

16:33:25 <danbri> I have a harvester, now bundled in RubyRdf that can scutter rdfs:seeAlso links. I've not yet found any calendar 'table of contents' page to point it at, though.

16:33:31 <mattb> DanC: i got about 1700 triples off one of your rdf URLs earlier today

16:33:38 <mattb> somethignsomething/smart/something.rdf

16:33:43 * mattb would have to check logs

16:33:50 <mattb> event-related i think

16:33:57 <DanC> F:[danc's foaf home KB|http://www.w3.org/People/Connolly/home-smart] or [home-smart.rdf|http://www.w3.org/People/Connolly/home-smart.rdf]

16:33:57 <dc_rdfig> Added comment F1.

16:33:58 * danbri updates http://esw.w3.org/topic/EventDiscovery to point to this chat, and to Matt's tool and RubyRdf...

16:34:13 <libby> that's cool eikeon :)

16:34:19 <DanC> I enhanced my XSLT homepage scraper to put seeAlso links to my travel schedule and itineraries

16:34:29 <danbri> oh cool

16:34:38 <libby> eikeon, how can information in your database be accessed?

16:34:49 * DanC enourages folks to put cool stuff in the weblog

16:34:56 <danbri> oh nearly forgot

16:35:30 <libby> I'm interested in mattb's ideas about subscribing to certain kinds on information in a RDF db

16:35:35 * DanC looks to see if http://eikster.com/2003/events/ has any of my stuff...

16:35:40 <eikeon> All the information in the events page is available in rdf/xml... the rdf_flyer icon at the top of the page.

16:35:47 <mattb> yeah, i don't see everyone running their own harvesters

16:35:48 <danbri> F:I started notes in [http://esw.w3.org/topic/AnnounceOMatic|ESW:AnnounceOMatic] towards harvesting event data from W3C mailing lists, specifically conference announcments...

16:35:48 <dc_rdfig> Added comment F2.

16:36:00 <libby> mattb, have you got a minute or two talk about your ideas?

16:36:03 <DanC> bingo! "AMERICAN AIRLINES #1549 from KANSAS CITY INTL to DALLAS FT WORTH"

16:36:08 <mattb> libby: i'll give it a go

16:36:19 <libby> thanks mattb

16:36:19 <danbri> re AnnounceOMatic, Caroline Meek at ILRT is going to have a rummage back through archives for a few months, look for common patterns...

16:36:23 <mattb> i expect lots of domain-specific applications to turn up out there that would need to crawl the whole rdfweb to find the bits they need

16:36:44 * danbri nods

16:36:49 <eikeon> Looks like AA #1549 etc events are coming from: http://www.w3.org/2003/05dc-bud/bud25Mar.rdf

16:36:52 <mattb> also, no point wasting all that bandwidth

16:37:03 <mattb> and harvesters should be well-behaved, respecting robots.txt, using ETags, etc

16:37:15 <libby> yup

16:37:19 <mattb> see RSS-hoovering bandwidth problems over the last few months as rss goes through a renaissance

16:37:22 * danbri whistles, looks at the floor, walls etc guiltily

16:37:28 <libby> mine isnlt well behaved

16:37:32 <mattb> also, a harvester should run continuosly

16:37:37 <mattb> not re-spidering once a day/week/whatever

16:37:49 <mattb> my current code remembers lots about what it saw during its plan

16:37:56 <mattb> and can now say whether a URL is updated since last run or not

16:37:59 <DanC> PROPOSED(F1): that all attendees of this meeting who build robots shall respect robots.txt and use ETags

16:38:08 <libby> does it record what it knows in RDF mattb?

16:38:14 <mattb> libby: yup

16:38:14 <libby> e.g. in a scutterplan?

16:38:20 <libby> neat, got an example?

16:38:28 <mattb> <url> <scutter:etag> "'foobarblah'" .

16:38:39 <mattb> <url> <scutter:lastModified> "2003-04-03 12:00" .

16:38:41 <mattb> etc

16:38:45 <libby> danc, I'll take an action to sort mine out re robotos.txt

16:38:50 <danbri> Seconded. I'll take an action to fix RubyRdf scutter in that regard.

16:38:50 <eikeon> My harvester is doing much the same... guilty on not paying attention to robots... but to ETag and pay attention to 404s etc.

16:39:02 <DanC> PROPOSED(F2): that all attendees of this meeting who build robots should respect robots.txt and use ETags or at least explain why not

16:39:10 <libby> cool eikeon

16:39:11 <mattb> i can dump out the scutterplan rdf bits this evening, put them up as example

16:39:20 <libby> thanks, that'd be great

16:39:29 <mattb> so, knowing what's new at least on a per-URL level, it could send out notifications about what it's found

16:39:32 * DanC hopes folks will say: I second F2.

16:39:36 * mattb seconds F2

16:39:40 * danbri does too

16:39:52 <libby> yep, sounds good to me

16:39:55 <danbri> respec!

16:39:56 * eikeon is interested in a schema where such information could be shared between harvesters.

16:39:57 <mattb> it could also know what apps are looking for what kinds of triples, and notify them only when they come along

16:40:02 * DanC asks that the chair put the question by saying "any objections to F2?"

16:40:07 <libby> ok

16:40:10 <mattb> yeah, scutterplan-sharing sounds sensible

16:40:12 <libby> any objections to F2?

16:40:20 <mattb> more than just the current form of a file full of rdfs:seeAlsos

16:40:20 <libby> that's be good mattb

16:40:35 <mattb> there should be stuff as in my examples above at least

16:40:50 <eikeon> So that if harvester A just cheched and document B has not changed harverster C can get that info from harvest A type of thing.

16:40:51 * DanC asks that the chair announce that the motion carries, at which point I'll tell dc_rdfig

16:41:05 <mattb> i have a prototype notification system that POSTs to registered URLs with an rdf document saying "here's some new stuff, headline info about it, and where you can get it if you want it"

16:41:12 * libby wasn;t sure how long to wait danc

16:41:22 <DanC> how long you wait is up to you, libby

16:41:24 <mattb> in other words, clients do the selection of what to download, notifications are broadcast to everyone

16:41:34 <libby> ok, F2 carries since there are no objections

16:41:46 <DanC> F:RESOLVED that all attendees of this meeting who build robots should respect robots.txt and use ETags or at least explain why not

16:41:47 <dc_rdfig> Added comment F3.

16:41:48 * mattb braindumping into channel, sorry to keep chatting via important process issues :)

16:41:56 * mattb at work...

16:42:01 <danbri> scutterplan sharing is very useful... also I'd like to provoke some collaboration on smarter heuristics for rdf crawlers. At the moment, the traverse-everything methodology re rdfs:seeAlso works, cos the Semantic Web is tinyish. But not for long...

16:42:03 <mattb> s/via/during/

16:42:06 <danbri> keep going matt :)

16:42:06 <DanC> braindumping is in order. we love IRC. ;-)

16:42:17 <libby> yeah, mattb, interesting stuff

16:42:22 <danbri> yes, agree re sharing amongst harvesters

16:42:22 <mattb> yeah, i'd like scutterplans to be graphs too

16:42:24 <mattb> not just lists

16:42:34 <mattb> x -> y -> (a -> b, c -> d) -> etc

16:42:36 <libby> like typed links?

16:42:43 <mattb> with annotations on why they were followed

16:42:52 <danbri> I think (mentioned this in #foaf a few times) the old Harvest project (C+Perl) worked out a lot of this for us, including Squid (which I think was theres) for cache sharing...

16:42:53 <mattb> particularly if there were hints in the origin document along with the seeAlso

16:42:53 <DanC> what are the ()s in x -> y -> (a -> b, c -> d) -> etc?

16:43:00 <danbri> Encouraging typed links in seeAlsos:

16:43:01 <DanC> oh... like n3 ;

16:43:02 <mattb> (boss)

16:43:11 <danbri> (hide!)

16:43:38 <danbri> <rdf:Description foaf:name="Dan Brickley"><rdfs:seeAlso rdf:resource="http:etc"/></rdf:Description> is useful.

16:43:46 <zool> mmm

16:43:56 <danbri> <foaf:Person foaf:name="Dan Brickley"><rdfs:seeAlso rdf:resource="http:etc"/></foaf:Person> is more useful.

16:44:09 <DanC> <x> s:seeAlso [= <y>; s:seeAlso [= <a>; s:seeAlso <b>], [= <c>; s:seeAlso <d>]].

16:44:24 <danbri> <foaf:Person foaf:name="Dan Brickley"><rdfs:seeAlso><foaf:BiblioDoc rdf:about="http://"/></rdfs:seeAlso></foaf:Person> is EVEN more useful.

16:44:37 <danbri> ie. you can mention types at either end of the seeAlso arc.

16:44:44 <libby> I like taht a lot

16:44:52 <libby> e.g. here's a calendar associated with me

16:44:54 <danbri> To bring it back to this meeting's goals, we could have foaf:CalendarDoc...

16:45:05 <libby> or maybe Vcalendar

16:45:07 <danbri> I've not really looked into a design for that...

16:45:12 <eikeon> And preferably at both ends of the arc, perhaps ;)

16:45:19 <DanC> why foaf:CalendarDoc when we already have ical:Vcalendar? oh yeah... not all Vcalendars are docs

16:45:22 <danbri> Yes, I think i asked previously if Vcalendar class was appropriate, but were some issues.

16:45:31 <danbri> yep...

16:45:45 <danbri> this would be a subclass of writeme:RDFXMLDocument...

16:45:58 <libby> vcalendar might be ok if there's at least one vcalendar at the other end

16:46:00 <libby> maybe

16:46:02 <danbri> the subclass of RDFXMLDocuments that describe a Vcalendar, perhaps

16:46:06 <libby> seems a shame to double up

16:46:27 <DanC> ok, so I connected my foaf home to my travel itinerary and eikon's harvester noticed. Now the geek in me is very pleased, but... ummm... how can impress my grandma with that? i.e. what actual value are we after?

16:46:28 <danbri> yup... I feel we're only slightly blocked from something quite useful... :(

16:46:43 <libby> I was particularly interested in the idea of subscribing to certain sorts of data from a database, e.g. via a query

16:46:59 <danbri> Didn't Jim Ley do some more pathcross in SVG stuff? SVG's nice for impressing the non-geek...

16:47:04 <danbri> sorry for the vague ref...

16:47:12 <DanC> can this harvester stuff help me play PathCross?

16:47:27 <eikeon> Something useful might be to get out meeting events into RDF/XML and have our bots be able to remind us and/or set the topic etc.

16:47:42 <libby> it mkight be able to if you could pull out potentail crosses via a query danc

16:47:45 <danbri> It can get you the data that allows you to ask questions that'll find the subset of data that'll help w/ pathcross...

16:47:46 * libby likes queries

16:47:55 <DanC> has anybody else published, formally (i.e. in RDF/XML) the fact that they plan to attend WWW2003?

16:48:16 <danbri> eg. people do dumb/blunt stuff in SQL aggregates, then invoke Cwm directly on selected URIs identified... (possible deployment combo)

16:48:27 <danbri> not me. Tell me what to type to assert that...

16:48:52 <eikeon> Is there a VEvent for WWW2003 somewhere?

16:48:55 * DanC wonders if eikon's scutter pulled WWW2003 out of my travel schedule; I don't think I declared it to be an ical:Vevent

16:48:59 <mattb> (unboss)

16:49:04 * mattb reads back

16:49:26 <libby> danc, I have in icalendar, but not RDF :(

16:49:46 <mattb> i have a feeling the "here's my query, tell me when you have new matches" problem is at least vaguely hard

16:49:50 <DanC> where, libby? how about you run ical2rdf on it right quick?

16:49:56 <mattb> certainly if full generality of rdql or similar is offered

16:50:10 <mattb> it's gotta be something you can bulk-match, not run N queries for N people on every new URL/changed URL

16:50:15 <libby> eikeon, I want to do that and have written a bot to do that, but I don;t know the best way to get the right data for it fromsomewhere

16:50:36 <libby> danc, yeah, ok

16:50:39 <DanC> hard in what way, mattb? implementations of graph-match-strength query engines are pretty easy to come by, around here. And euler can even do better than that.

16:50:48 <eikeon> libby: I leverage my existing harvester ;)

16:50:54 <libby> :)

16:50:59 <libby> yeah

16:51:02 <mattb> DanC: partly gut feeling

16:51:12 <eikeon> ... and could extend eikco to do it, if people like.

16:51:18 <zool> libby yeah what is the right data... are you taking bot/REST approach a la mudlondon, picbot

16:51:21 <DanC> now if libby runs ical2rdf on her ical file, that'll assert that WWW2003 exists. Will it connect libby with that event?

16:51:29 <mattb> but i don't think it'd be possible to accept a new subgraph into a store and say that some of its triples matched a query without checking again the existing complete graph

16:51:30 <libby> eikeon, I (and I'm sure others can ask a query of my db and get ntriples back

16:51:33 <libby> yep zool

16:51:38 <eikeon> Else maybe pull data from: http://eikster.com/2003/events/?content_type=application/rdf%2Bxml

16:51:44 <mattb> in the case where the existing graph has some of the triples req'd and the subgrpah completes the graph required for results

16:51:50 <libby> ok, handy, thanks eikeon :))

16:51:50 <DanC> hmm... how to say "I'll be there!"

16:51:53 * mattb defers to those wth more query-optimising knowledge

16:51:58 <libby> attendee?

16:52:12 <DanC> you mean ical:attendee?

16:52:17 <libby> do you have an code for the subscribe thing mattb?

16:52:19 <danbri> mattb, yup its a belief-update-problem, when you learn new stuff, not clear which other conclusions need revising... so you end up requerying everything

16:52:21 <libby> yeah...

16:52:28 <danbri> maybe ok, can do in batch or round-robin...

16:52:41 <mattb> libby: yes, for the basic announce-and-clients-do-what-they-like-with-it

16:52:45 <mattb> i have java server, perl client

16:52:49 <mattb> the client bit is pretty trivial

16:53:00 <DanC> I've been using { ?E cyc:socialParticipants ?WHO }. now why haven't I been using { ?E ical:attendee ?WHO}? hmm...

16:53:09 <mattb> i see it as a separate rest/rdf messaging system that the scutter could use as a toolkit

16:53:22 <mattb> it's an implementation of some stuff posted to the rest-discuss list, with the xml rdf-ized

16:54:36 <DanC> some part of me hesitates to say that WWW2003 is an ical event. ical events have exactly one organizer who has exactly one mailbox, and I dunno if WWW2003 works that way. or something.

16:54:36 * danbri wonders where we go from here... we have people, we have events, we have harvesters, but but but...

16:55:00 <danbri> I'm willing to run with cyc:socialParticipants or whatever folk recommend, just looking to do what someone tells me ;)

16:55:23 * eikeon proposes action that would result is more events finding there way to/from rdf/xml

16:55:28 <mattb> i'd like to build some little demo apps that use a harvester and pubsub mechanism

16:55:29 <eikeon> ... whatever that may be.

16:55:40 <zool> ical:attendee sounds good, though odd ns

16:55:51 <mattb> i would also like to split a harvester into a url-discover-checker-and-downloader-and-notifier and a storage system

16:55:52 <zool> forkage, event:participant

16:56:06 <mattb> iow i don't think harvesters should necessarily aggregate the entire graph of what they find

16:56:24 <zool> what side will the smushing go on mattb

16:56:42 <mattb> smushing would happen in a smushing store subscribed to the crawler i think

16:56:58 <mattb> it could do arguably-dodgy things like bnode-labeling to help simple clients

16:56:59 <libby> ok, there's my calendar data danc - not sure if it came out right: http://ilrt.org/people/libby/calendar/exportrred20030414.rdf

16:57:00 * eikeon is currently smushing at query time.

16:57:33 <mattb> i would like it if the crawler's only state was its scutterplan and annotations

16:57:34 <DanC> F:[Upcoming Events discovered on the Semantic Web|http://eikster.com/2003/events/], a service by eikon, found my flights.

16:57:36 <dc_rdfig> Added comment F4.

16:57:43 <zool> sounds lush; so i need smushing on the aggregator end; i still dont have an algo i'm happy with

16:57:54 <danbri> this is very much the harvest gatherer/broker and cacheing design (I'm not complaining -- it's an obvious pattern)...

16:58:03 * mattb steals all the best ideas

16:58:14 <eikeon> libby: Added... and more events show up :) http://eikster.com/2003/events/

16:58:20 * DanC perks up at "obvious pattern" ... esw wiki topic, pls?

16:58:28 <mattb> reminds me of discussion of federated, specialised stores and query-systems discussed with zool and yoz last summer

16:58:31 <danbri> I'd be willing to repackage a .tar.gz of re-serialized scuttered RDF (though I worry about privacy laws etc...)

16:58:51 <DanC> libby, now can you connect that calendar from your foaf home via rdfs:seeAlso?

16:58:54 * danbri nods to DanC, but I can't think of a good wiki topic

16:59:01 <mattb> i have a demo of bnode-labelled (for smushing) reserialized RDF notification, but it keeps the re-generated model in memory

16:59:10 <mattb> so it's conceptually a transitory message with a time-to-live

16:59:18 <libby> so, I'm thinking we shoudl try to come up with some action-type things to end discussion of this tpoic for now, as we have many things to get through and 30 mins to go...

16:59:19 <mattb> i wouldn't like the idea of permanently republishing other people's RDF with extra stuff on

16:59:24 <libby> danc, okey dokey

16:59:27 <danbri> I wish I'd gotten to go to http://www.w3.org/Search/9605-Indexing-Workshop/ -- looked v interesting

16:59:37 <DanC> F:a wiki topic around "the harvest gatherer/broker and cacheing design" seems to be brewing... near pubsub

16:59:38 <dc_rdfig> Added comment F5.

16:59:39 <danbri> ...maybe we should have another?

16:59:54 <danbri> (big job though)

17:00:21 <DanC> eikon, http://eikster.com/2003/events/ sparks my curiousity... I'd like a "view source" sorta link... what query is it running? how often?

17:00:51 <DanC> can you add people and places to the query somehow?

17:01:11 <DanC> events do seem to be ordered by date; that's cool.

17:01:45 <eikeon> DanC: query is done on page request. And source is at: http://eikster.com/2003/hypercode

17:02:13 <DanC> F:[source of eikster.com upcoming events|http://eikster.com/2003/hypercode]

17:02:13 <dc_rdfig> Added comment F6.

17:02:20 <libby> eikeon, cool!

17:03:34 * danbri visits http://eikster.com/2003/events/ sees "staff meeting at ILRT"

17:03:50 <eikeon> Um... code is a bit hard to follow with the inner/outer hackage going on to share code between html and rdf/xml view of events page.

17:03:57 <danbri> i'm reminded of skical, i think they've been further than most in national event gathering... maybe some lessons re implicit contexts

17:04:09 <danbri> eg. "Mother's Day" <- in which country?

17:04:15 <danbri> (US vs UK varies)

17:05:03 <DanC> re "in which country", the cool think about 2003/event/s is it's got the "oh yeah?" button next to each event. so you can follow the link to http://eikster.com/2003/US32Holidays.rdf

17:05:35 * DanC awards eikon 200 points for implementing an "oh yeah?" button

17:05:37 <libby> ok, I've linked my rdf calendar witha typed seealso from http://swordfish.rdfweb.org/people/libby/rdfweb/webwho.xrdf

17:06:09 <libby> hehe, re ILRT meeting

17:06:15 <danbri> eikeon, this is very cool! sorry if i leapt straight in with questions/nitpics. nice work :)

17:06:37 * libby looks at clock, agenda

17:07:14 * DanC doesn't mind if this item consumes the rest of the meeting

17:07:17 <danbri> does anyone have code that goes from rdfical back to ical? I'd like to use that now I have harvester happily working and a Mac w/ iCal to view .ics files on...

17:07:17 <eikeon> danbri: The icon to the right of the event label links to the source... in the case of mothers day the source url provides a nice clue ;) http://eikster.com/2003/US32Holidays.rdf

17:07:19 * danbri neither

17:07:24 * danbri just noticed :)

17:07:29 <DanC> back to ical: yes...

17:07:45 <DanC> ... http://www.w3.org/2000/10/swap/pim/toIcal.py

17:08:43 <DanC> PROPOSED(F.a): that folks should be encouraged to link their calendars from their foaf home thingies via rdfs:seeAlso (@@e.g. danc's, libby's)

17:09:02 * libby seconds F.a

17:09:13 <DanC> PROPOSED(F.att):that folks should use ical:attendee to relate themselves to events that they plan to attend.

17:09:23 * DanC isn't sure about F.att

17:09:45 <DanC> PROPOSED(F.att.cyc):that folks should use cyc:socialParticipants to relate themselves to events that they plan to attend. (@@e.g. danc's travel schedule)

17:09:46 <libby> it's tricky. attendee points to an email address?

17:09:48 <danbri> funny, I was studying http://www.w3.org/2000/10/swap/pim/toIcal.py recently, but just for the API, i didn't notice what it was _for_ :)

17:10:06 * danbri adds it to EventDiscovery (by hand, wonders about installing wikibot here)

17:10:15 <libby> danc, how do you refer to yourself in that file?

17:10:41 <danbri> re F.a, 2nd'd.

17:10:52 <DanC> in an itinerary: <r:Description r:about="bud25Mar.n3#thisTrip">

17:10:52 <DanC> <k:firstSubEvents r:resource="bud25Mar.n3#_gflt1549_3"/>

17:10:52 <DanC> <k:lastSubEvents r:resource="bud25Mar.n3#_gflt1319_14"/>

17:10:52 <DanC> <k:passengers r:resource="http://www.w3.org/People/Connolly/home-smart#me"/>

17:11:28 <danbri> re F.att*, do those properties carry the meaning of anticipated attendence?

17:11:41 <danbri> ie. we can't use them to determine who actually attended

17:11:42 <DanC> phpht. in http://www.w3.org/People/Connolly/travel-sched , I don't connect myself to events. :-{

17:12:08 * libby checks rfc2445

17:12:19 <DanC> any offers to shout loudly in the foaf community re F.a?

17:12:40 <zool> is there a definitive-enough reference file for attaching events to myself?

17:12:46 <libby> I wonder how easy it is to connect yourself to an event using calendar tools

17:12:57 <mattb> how do you pick the object of such a triple?

17:13:03 <mattb> canonical URL for www2003? or literals?

17:13:09 * mattb apologises if this has been overdiscussed

17:13:13 <libby> we would have to be able to recommend using mbox or mbox_sha1sum

17:13:17 <libby> re foafers

17:13:27 * eikeon will add a page to view an individual event showing attendees... when we know what property(ies) we want to use... and there is some data to test with.

17:13:49 <libby> could use foaf:homepage to connect event and cononical url

17:13:54 <DanC> I expect GUI calendar tools to keep the association with the user implicit. I'd use a command-line option on ical2rdf.pl to say "make it explicit that I, <http:.../#me>, plan to attend all these events"

17:14:06 * DanC has an example somewhere...

17:15:11 <eikeon> Or I, <http:.../> ;)

17:15:45 <DanC> ah... in my WWW2003 makefile: 05dc-bud/Makefile: $(PERL) $(GROKI) --aboutMe $(ABOUTWHO) --localName $(WHO) < $< >$@.nt

17:15:55 <libby> rfc 2445: 4.8.4.1 Attendee....Value Type: CAL-ADDRESS; "This property MUST

17:15:55 <libby> NOT be specified in an iCalendar object when publishing the calendar

17:15:55 <libby> information"

17:16:25 <DanC> yeah... grubby protocol constraints like that are why I hesitate to say that WWW2003 is an ical:Vevent

17:16:39 <libby> [[

17:16:40 <libby> Formal Definition: The value type is as defined by the following

17:16:40 <libby> notation:

17:16:45 <libby> cal-address = uri

17:16:46 <libby> ]]

17:17:01 * DanC challenges eikeon to a duel for the honor of HashURIs ;-)

17:17:03 <danbri> <DanC> any offers to shout loudly in the foaf community re F.a?

17:17:14 <danbri> sure though not sure quite what to shout yet

17:17:28 <libby> well www2003 could still be a vevent, but use the cyc property not attenddee

17:18:32 <DanC> my WWW2003 makefile: http://www.w3.org/2003/05dc-bud/Makefile

17:19:08 <DanC> meetingbot, pending proposals?

17:19:26 <DanC> PROPOSED(F.att.cyc):that folks should use cyc:socialParticipants to relate themselves to events that they plan to attend. (@@e.g. danc's travel schedule)

17:19:34 * libby enthused to work on calendar bot via eikeon's harvester

17:19:35 <DanC> PROPOSED(F.att):that folks should use ical:attendee to relate themselves to events that they plan to attend.

17:19:45 <DanC> PROPOSED(F.a): that folks should be encouraged to link their calendars from their foaf home thingies via rdfs:seeAlso (@@e.g. danc's, libby's)

17:20:08 <DanC> yeah, but meetingbot, how many 2nds for each of those?

17:20:09 <danbri> what's the range of: cyc:socialParticipants ?

17:20:13 <libby> F.a I seconded

17:20:18 <DanC> cyc:Agent, I think, danbri

17:20:20 <libby> F.att is tricky

17:20:22 <eikeon> DanC: :) -- I recently jump from the Hash side of the fence to Slash for my own usage. As I kept bumping into cases where I wanted a URIRef with a hash to make it to the server... but the hashes do not make it ;(

17:20:40 <DanC> so it doesn't look like we'll close on F.att. shall I follow up in email?

17:20:49 <libby> that'd be great danc

17:21:23 <DanC> F:ACTION DanC: follow up in email about how folks should relate themselves to events they plan to attend. ical:attendee? cyc:socialParticipants?

17:21:25 <dc_rdfig> Added comment F7.

17:21:31 * libby wonders about harvesters and repeating events

17:21:35 * eikeon seconds F.a

17:21:37 <DanC> (meetingbot notes that F.att.* are withdrawn)

17:21:45 <libby> any objections to F.a?

17:22:36 * danbri 3rds F.a (for RDF docs only, not .ics...)

17:22:36 <libby> an expander from repeating events to non-repeating version has been suggested before, but non-trivial I think

17:23:30 <DanC> as the meeting closes, note that you all have about 10 hours to decorate the (scratchpad/weblog part of the) meeting record with stuff that you want to be able to find without poring over the IRC log.

17:23:47 <DanC> i.e.

17:23:48 <DanC> .time

17:23:49 <datum> Wed, 23 Apr 2003 17:23:48 GMT

17:24:00 <DanC> the chump bot rolls over and loses its marbles at 24:00Z

17:24:09 <libby> PROPOSED(F.a): that folks should be encouraged to link their RDF calendars from their foaf home thingies via rdfs:seeAlso (@@e.g. danc's, libby's)

17:24:25 <eikeon> What happens after 10 hours... does it get frozen? /me curious.

17:24:29 <libby> any objections to ammended F.a?

17:24:30 <DanC> F:RESOLVED: that folks should be encouraged to link their RDF calendars from their foaf home thingies via rdfs:seeAlso (@@e.g. danc's, libby's)

17:24:31 <dc_rdfig> Added comment F8.

17:24:37 * DanC oops

17:24:51 <libby> I think ok danc, just checkoing the RDF bit

17:24:53 <danbri> seems agreed-enough to me

17:25:05 <DanC> best if somebody followed up on F.a with an article or something, though I suppose it's not critical

17:25:22 <mattb> eikeon: at midnight GMT the frontpage goes into the archive

17:25:22 <libby> an article?

17:25:27 <mattb> and the chump starts again at A:

17:25:34 * DanC likes the fact that if anybody didn't think it was resolved, they could scribble it out.

17:25:43 <eikeon> mattb: ty

17:25:50 <libby> ok, we have 5 minutes, shall we discuss time and date of next meet?

17:26:16 <DanC> yes (and record it under A)

17:26:17 * danbri nods, suggests fortnightly working ok for now...

17:26:31 * DanC tries to compute now() + 1 fortnight

17:26:33 <libby> it should be on 7th May, but I will be at XMLEurope, notsure about internet access yet

17:26:40 * mattb at xmleurope also

17:26:45 <mattb> prolly no net access

17:26:47 <mattb> if it's like last year

17:26:55 <mattb> speakers' notes say no net connection for speakers

17:27:00 <libby> !

17:27:16 <mattb> etcon it is not

17:27:35 <DanC> <#event_-1515880520> k:temporallyIntersects <#theDay2003-05-07>;

17:27:35 <DanC> rdf:value """RDF Calendar """; is foaf:topic of <http://calendar.sidekick.dngr.com/event?id=486&event-date=2003-05-07>.

17:27:54 <DanC> ^ i.e. it's already in my calendar

17:28:15 <DanC> and there aren't any conflicts.

17:28:21 <libby> we did say every 2 weeks till www2003

17:28:35 <libby> so, we can do that

17:28:42 <DanC> I guess I'm willing to convene a meeting on 7 May... regrets from libby?

17:28:53 <libby> great, thanks danc

17:29:02 <libby> yep, regrets from me

17:29:10 <DanC> just a sec...

17:29:40 <DanC> PROPOSED(A.next7): to meet next on 7May 1600Z, per every-other-week agreement. danc to call the meeting. regrets from libby.

17:29:44 * DanC would like a 2nd

17:29:56 * eikeon wonders if next meeting is finding its way into rdf/xml ?

17:30:17 <DanC> it's in RDF/N3 above ;-)

17:30:26 <danbri> 2nd'd, but hmm i'm considering xml europe attendance...

17:30:27 <libby> it's in with the time incorrect in my file as a repeating event

17:30:42 <libby> eikeon, but I normally create an rdf file for each meet, will do.

17:30:59 <DanC> can anybody else confirm for 7May?

17:31:07 <eikeon> libby: Nice :)

17:31:20 <mattb> danbri: i think 7th may is when we are all doing our semweb talks

17:31:25 <mattb> me, celia, zool, uche, etc

17:31:34 <mattb> so you gotta come see us ;)

17:31:39 <DanC> PROPOSED(A.cxl7): to cancel 7May meeting.

17:32:32 * danbri nods re cancel

17:32:35 <DanC> I withdraw my offer to convene a meeting 7May

17:32:39 * libby seconds A.cxl7

17:32:51 <libby> ok, so next meeting...?

17:32:57 <DanC> so what's 7May+2 weeks? are we in Budapest by then?

17:32:58 * mattb heads for home

17:33:02 <mattb> interesting chat, cheers

17:33:03 <libby> yep

17:33:07 <libby> mattb, thanks

17:33:16 <libby> yep re budapest

17:33:22 <libby> we could do 14th, in between

17:33:36 <DanC> oops; we assigned actions before we figured out when we'd have a chance to review them.

17:33:47 * DanC checks 14May

17:34:08 <DanC> no conflicts.

17:34:36 <DanC> PROPOSED(A.next14): to meet next on 14May at 1600Z.

17:34:49 <danbri> 2nd'd.

17:34:53 * libby seconds A.next14

17:35:04 <libby> any objections?

17:35:14 <libby> can anyone else make it?

17:35:24 * danbri believes so

17:35:55 * DanC notes folks were excused at xx:30, the scheduled end time

17:36:08 <libby> true

17:36:12 <libby> we're 6 mins over

17:36:20 <libby> maybe next time do scheduling first....

17:36:21 <DanC> A:next meeting looks like 14May at 1600Z, though we didn't formally decide.

17:36:22 <dc_rdfig> Added comment A6.

17:36:30 <DanC> A:RESOLVED: to cancel 7May

17:36:30 <dc_rdfig> Added comment A7.

17:36:49 <libby> A:[http://www.timeanddate.com/worldclock/fixedtime.html?day=14&month=5&year=2003&hour=16&min=0&sec=0&p1=0|2003-05-14, 1600UTC]

17:36:50 <dc_rdfig> Added comment A8.

17:37:00 <libby> thanks all, sorry for overrunning

17:37:06 <DanC> not at all.

17:37:09 <libby> very intersting chat today

17:37:13 <DanC> quite!

17:37:19 <libby> :)

17:37:27 <q|eve> q|eve is now known as qmacro

17:38:26 * libby looks at http://www.xmleurope.com/2003/wednesday.asp

17:38:35 <libby> wishes there was an RDF view...

17:39:00 <libby> does anyone know if there are official RDf versions of the WWW2003 schedule?

17:40:05 <libby> I did these: http://swordfish.rdfweb.org/discovery/2003/04/www2003/ but looking at the now, there are no timezones or anything

17:45:37 <eikeon> libby: Think I am going to rework how the rdf/xml views of pages on my sites work... will impact aggregated events rdf/xml... but not much.

17:45:53 <libby> cool

17:46:31 <libby> so much to do, so little time.....

18:06:42 <sandro> What breaks if you think of QNames as URIs XMLNS as providing a local-override URI Scheme? Anything?

18:06:53 <sandro> What breaks if you think of QNames as URIs, and XMLNS as providing a local-override URI Scheme? Anything?

18:44:14 * libby wonders what to suggest for the (late) rdf query testcases agenda

18:45:11 * danbri fails to understand sandro's proposal

18:45:11 <danbri> like file://dc/title vs xmlns://dc/title? or dc:title with dc: being the uri scheme?

18:46:50 <sandro> slashes are out, but why not just view dc:title as being itself a kind of relative-URI (in this case relative to some ns declarations instead of a base). ? It's just such a pain that

18:47:17 <sandro> "x:y" could be a qname OR a URI and the semantics are ALMOST the same, and maybe could be the same.

19:20:22 <libby> cheers for all the discussion mattb earlier, very interesting

19:20:30 <mattb> ta libby

19:20:37 <mattb> helps to talk it over

19:20:42 <mattb> ideas been in back of head for a while

19:21:07 <libby> :)

19:21:11 <mattb> haven't updated my foaf file for a while

19:21:20 <mattb> be interesting to do some event-attendance in there

19:21:31 <libby> yeah, definitely

19:21:50 <mattb> i was thinking about an index of pieces on my site using foaf:Document too

19:22:01 <libby> ooh

19:22:03 <mattb> should come up with a mapping from my movabletype categories into some useful ontology

19:22:18 <mattb> it'd just be a simple MT template probably, perhaps with a plugin

19:22:29 <libby> I've not looked at MT really

19:22:36 <DanC> what breaks you can't take dc:title and hand it to somebody else as a meaningful symbol. Yes, it works like a relative URI reference (timbl's n3 grammar has a uriref2 production that includes qnames) , but not like a URI.

19:22:38 <mattb> every article has an rdf index already

19:22:43 <mattb> eg http://www.hackdiary.com/archives/000030.rdf

19:23:23 <Morbus> mmmm. MT.

19:23:44 <libby> ooh, neato

19:24:22 <dajobe> that seems to use xlink

19:24:33 <mattb> yeah, random rdf i made a while back

19:24:43 <mattb> i wanted to ask earlier how best to encode the code of weblinking in rdf

19:24:47 <mattb> for my scutterplan really

19:24:52 <mattb> but we'd moved on in the agenda

19:24:54 <mattb> forgot about it

19:25:15 <mattb> dc:relation might work in the hackdiary case above i suppose

19:25:33 <mattb> "i found X on the web and it linked to Y" is the sort of statement i want to make in the scutterplan case

19:25:53 <dajobe> using the xlink terms might be confusing, people (software?) might interpret the XML links rather than leave them as rdf properties

19:25:59 <mattb> true

19:26:00 <dajobe> html:a ?

19:26:08 <mattb> good idea

19:26:11 <sandro> Thanks, DanC. That doesn't sound so bad -- we already use a lot of relative URIs, and understand how they are just syntactic sugar. We don't have an algorithm for turning a qname + NS-declarations into an absolute-URI, though. (and if we did, some people would probably get upsef....)

19:26:16 <dajobe> or maybe dc:relation ...

19:26:20 <mattb> it's actually scraped straight from <a href="foo"> in the original html

19:26:28 <dajobe> I expected it did

19:26:53 <dajobe> at least dc:relation has a schema around, no html rdfs I know

19:27:07 <mattb> i used dc:relation in something the other day

19:27:19 <mattb> oh, the upmystreet postcode<->latlong service

19:27:38 <dajobe> probably just as much use as rdfs:seeAlso hmm

19:27:54 <mattb> as in ihttp://www.picdiary.com/~mattb/rdf/e97qp.rdf

19:28:46 <mattb> the intention was to aggregate all the articles and say "((A linksto X) and (B linksto X)) implies (A relatedto B)"

19:34:25 <DanC> [[ We don't have an algorithm for turning a qname + NS-declarations into an absolute-URI, though. (and if we did, some people would probably get upsef....) ]]

19:34:35 <DanC> ^ that's a TAG issue, owner yours truely... discussion welcom.

19:35:00 <DanC>http://www.w3.org/2001/tag/ilist#rdfmsQnameUriMapping-6

19:35:01 <dc_rdfig> K: http://www.w3.org/2001/tag/ilist#rdfmsQnameUriMapping-6 from DanC

19:35:11 <DanC> K:|rdfmsQnameUriMapping-6 : Algorithm for creating a URI from a QName?

19:35:13 <dc_rdfig> Titled item K.

19:38:03 <DanC> K:hmm... my favorite is pfx = ns[pfx]; if xmlnamechar(pfx[-1]): pfx = pfx + '#'; uri=pfx + lname

19:38:04 <dc_rdfig> Added comment K1.

19:38:58 <DanC> K:satisfies the invertible requirement

19:38:59 <dc_rdfig> Added comment K2.

19:39:18 <DanC> K:to discuss: multiple namespaces in XML schema

19:39:18 <dc_rdfig> Added comment K3.

19:39:46 <DanC> K:cite discussion in (where? rdf-logic?) about using qnames across XSLT and RDF, and how it matters where you split the URI

19:39:46 <dc_rdfig> Added comment K4.

19:40:25 <DanC> K3:to discuss: XML schema naming expressions have more than just ns uri and localname. need to know whether you're referring to a type decl or an element decl etc.

19:40:26 <dc_rdfig> Replaced comment K3.

19:41:50 * danbri gets drawn into uri@w3 thread, probably bad idea...

19:51:31 * DanC was hoping to have that discussion in/around esw:GoodURIs

19:51:54 * DanC might excert some power vested in me as URI CG chair to get that to happen, to save Roy some grief

19:56:08 <danbri> my attempt at FAQ was midly flippant, but some form of Q'n'A about URIs might be a useful thing to explore

19:57:46 <danbri>http://nwalsh.com/java/jpegrdf/

19:57:47 <dc_rdfig> L: http://nwalsh.com/java/jpegrdf/ from danbri

19:57:55 <danbri> L:|jpegrdf

19:57:55 <dc_rdfig> Titled item L.

19:58:12 <danbri> L:From Norm Walsh. "Jpegrdf reads and manipulates RDF metadata stored in the comment section of JPEG images. It can extract, query, and augment the data. Manipulating JPEG images with jpegrdf does not modify the actual image data or any other sections of the file."

19:58:12 <dc_rdfig> Added comment L1.

19:58:24 <Morbus> danbri: i'd like to see a CoolURIs FAQ.

19:58:29 <danbri> L:"Version 1.1 includes the ability to provide decoder classes for proprietary fields in the EXIF data. I've also cleaned up the code a little bit and submitted my jpeg library changes back to the W3C folks."

19:58:29 <dc_rdfig> Added comment L2.

19:58:38 <Morbus> I refer people to the original a lot.

19:58:58 <danbri> Morbus, if you have questions, scribble them in the Wiki, maybe someone'll answer them (hopefully citing evidence/texts/authorities...)

19:59:22 <Morbus> wiki on rdfweb?

20:01:08 <danbri> no, http://esw.w3.org/topic/GoodURIs

20:01:22 <dajobe> dajobe has changed the topic to: Semantic Web wishes can come true http://rdfig.xmlhack.com/

20:01:40 <dajobe> . O O (what was the old topic)

20:01:56 <sbp`> *** Topic is 'RDF Calendar meeting 2003-04-23 1600 UTC http://rdfig.xmlhack.com/'

20:01:56 <sbp`> *** Set by DanC on Wed Apr 23 17:06:47

20:02:02 <Morbus> *** Topic of #rdfig: Semantic Web wish factory, http://rdfig.xmlhack.com/ 'When you go meta, don't forget to come back'

20:02:18 <dajobe> I was close

20:02:39 <sbp`> oh... *old*...

20:02:52 <danbri> that should be the slogan of the entire RDF IG...

20:03:03 <dajobe> dajobe has changed the topic to: Semantic Web wishes can come true http://rdfig.xmlhack.com/ - wiki http://esw.w3.org/topic/FrontPage

20:03:55 <sbp`> when you wish upon a triple...

20:04:19 <DanC> L:one for [http://esw.w3.org/topic/ImageDescription]

20:04:19 <dc_rdfig> Added comment L3.

20:04:32 <DanC> L3:one for [esw:ImageDescription|http://esw.w3.org/topic/ImageDescription]

20:04:32 <dc_rdfig> Replaced comment L3.

20:14:04 Topic now Semantic Web wishes can come true http://rdfig.xmlhack.com/ - wiki http://esw.w3.org/topic/FrontPage

20:14:04 Users on #rdfig: logger_3 esigler mortenf larsbot mattb eikco TrustBot dajobe Dorward golbeck xover Esaj wkearney99 libby jang bijan AndyS Morbus inkel sbp` idoru logger_1 deltab ericP sandro nb jeremiah DanC tav|offline aml danbri kham ilowe eikeon Wack qmacro em_ grault datum xower dc_rdfig

20:14:05 <ChanServ> [#rdfig] This channel is logged and blogged: http://logicerror.com/rdfIRCWelcome

20:14:05 Disconnected from irc.freenode.net (ERROR :Closing Link: logger_3 (Killed (carroll.freenode.net (clarke.freenode.net <- asimov.freenode.net))))

20:15:02 Topic now Semantic Web wishes can come true http://rdfig.xmlhack.com/ - wiki http://esw.w3.org/topic/FrontPage

20:15:02 Users on #rdfig: logger_2 esigler mortenf larsbot mattb eikco TrustBot dajobe Dorward golbeck xover Esaj wkearney99 libby jang bijan AndyS Morbus inkel sbp` idoru eikeon Wack qmacro em_ grault datum xower dc_rdfig ilowe kham danbri aml tav|offline DanC jeremiah nb sandro ericP deltab

20:15:02 <ChanServ> [#rdfig] This channel is logged and blogged: http://logicerror.com/rdfIRCWelcome

20:15:13 <DanC> there's a whole semantic web track and dev day at WWW2003, I think

20:15:28 <mattb> yeah, i shall be there

20:15:42 <mattb> what's the format of developer day?

20:15:50 <mattb> what makes it different from the tracks?

20:15:54 <DanC> .google www2003 developer's day

20:15:56 <datum> www2003 developer's day: http://www.ai.mit.edu/people/dquan/resume.html

20:16:00 <DanC> ?

20:16:11 <mattb> the www2003 site doesn't make it entirely clear

20:16:30 <mattb> i asked eric miller by email but didn't get much response

20:16:38 <DanC> devday stuff is usually less baked...

20:16:46 <DanC> the track is for peer-reviewed stuff.

20:17:58 <DanC> have you been to a WWW conference? maybe the best answer is the #rdfig blog from dev day last year in hawaii...

20:18:06 <mattb> nope, my first time

20:18:17 <danbri> mattb, http://esw.w3.org/topic/RdfQueryTestingBudapestMeeting

20:18:38 <mattb> oh, cool

20:18:38 * danbri just sent mail again today trying to clarify its officially-blessed-ness, get it linked from www2003 site etc

20:19:08 <DanC> dev day 2002: http://www2002.org/devday.html

20:19:52 <danbri> "Join us for a unique opportunity to chat with TimBL, look back upon eight years of DevDays, see the latest Semantic Web demos, and kick around ideas for the Next Big Thing."

20:20:02 * mattb finds http://ilrt.org/discovery/chatlogs/rdfig/2002-05-11.html

20:20:03 <DanC> rdfig blog from devday 2002: http://rdfig.xmlhack.com/2002/05/11/2002-05-11.html

20:20:03 * danbri wonders what's lined up there... demo-wise

20:20:35 <danbri> ah, that was _last_ year. /me dohs

20:20:52 * danbri tries to remember demos at lunchtime, remembers only food, and ppl clamouring for an rdf query WG

20:22:39 <DanC> chuckle... 19:15:42 * dajobe-lap slaps AaronSw with a pointy bracket

20:23:11 <mattb> <DanC> ok, I wanna play with photo metadata tools

20:23:15 <mattb> same again a year later?

20:24:30 * dajobe remembers the speed typing, rubs his wrist

20:24:36 <mattb> good logs dajobe

20:25:22 <DanC> yes, life is a search for photo metadata tools.

20:26:16 <libby> danc, have you got a definition of http://www.w3.org/2000/10/swap/pim/contact#nearestAirport?

20:26:22 <DanC> i.e. what I want to say about my photos is not very much smaller than what I want to say about the world in general. likewise my calendar.

20:26:47 <mattb> yes, my stuff revolves around photos

20:26:53 <mattb> i don't have many words, except on irc

20:27:34 <DanC> nearestAirport... er... it's a property I used to ask WebOnt WG members where they sorta usually are in order to facilitate discussions about where we should hold our ftf meetings.

20:27:45 <DanC> does that suffice?

20:28:04 <DanC> in fact, it's *the* such property.

20:28:13 <danbri> its now an emergent standard for next generation PIM and stalker-support apps.

20:28:21 <libby> I was going to reuse it ... :)

20:28:34 <libby> it's been used in foaf files a fair bit I think

20:28:50 <DanC> the document currently available from http://www.w3.org/2000/10/swap/pim/contact doesn't authorize its use as a property, I don't believe.

20:29:13 <libby> yep, I just realized that

20:29:44 * libby inclined to think it doesn;t matter...

20:30:02 <DanC> it matters, but it matters a whole lot less than a lot of other things.

20:30:07 <qmacro> qmacro is now known as q|zzz

20:31:04 <libby> I wonder why www1.iata.org/codes redirects to the MS IE site?

20:31:04 <mortenf> heh, dajobe, re our rss version talk yesterday: http://use.perl.org/~gnat/journal/7774

20:31:29 * danbri finds 61 uses

20:35:59 <mattb> libby: you were asking about the model my scutter builds for its plan

20:36:05 <mattb> it's at http://www.hackdiary.com/rdf/scutterplan.rdf

20:36:27 <mattb> that's a dump straight from the database, no real polishing or deep thought gone into it

20:37:37 <libby> thanks mattb, that's very useful

20:38:32 <libby> mind if I add it to F in the chump?

20:38:42 <mattb> go ahead

20:39:39 <libby> F:[http://www.hackdiary.com/rdf/scutterplan.rdf|mattb's enhanced scutterplan] with lastmodified among other things

20:39:39 <dc_rdfig> Added comment F9.

20:39:40 <libby> ta

20:44:10 <mortenf> re the internal layered scutter "subscription" issue, i'd thought of specifying a (number of) classes/properties, and simply getting subject uris in return when something matches.

20:44:34 <mortenf> then the subscriber could query the store for actual info

20:44:46 <mortenf> thereby getting the complete graph.

20:46:33 <sandro>http://lists.w3.org/Archives/Public/uri/2003Apr/0062.html

20:46:34 <dc_rdfig> M: http://lists.w3.org/Archives/Public/uri/2003Apr/0062.html from sandro

20:46:47 <sandro> M:|[proposed] New Text for RFC 2396 intro, reframing what URIs are for

20:46:47 <dc_rdfig> Titled item M.

20:47:20 <sandro> (by me)

20:47:47 * sandro dives into the mud again, lured by a glimmer of hope.

20:52:09 <libby> mortenf, I'm not sure what you mean...got an example?

20:53:35 <mortenf> not really, was thinking mostly along the lines of rss, i.e. if a new (or updated) channel resource is added/updated in the triple store, one could request to be notified, receiving the uri of the channel resource.

20:54:28 <mortenf> probably too simple, since an rss channel is seldom updated, the rss:items sequence is...

20:55:14 <sandro> K: Yeah -- hash-vs-slash, how to be invertable, and what to do about people who don't write their XML with this in-mind......

20:55:14 <dc_rdfig> Added comment K5.

20:55:45 <mortenf> but, the major point was that a subscriber should get a uri of some sort, not a graph, in return.

20:58:18 <mattb> mortenf: sounds interesting

20:58:39 <mortenf> so far just an idea, not sure if it would work...

20:58:42 <mattb> doing simple predicate or subject matching would be possible without reference to the entire graph store

20:58:45 <mattb> that's the kind of thing i had in mind

20:59:06 <mattb> my current prototype gives the subscriber a URI

20:59:08 <mattb> seems the best way

20:59:30 <mortenf> yeah, and simple too!

20:59:40 <mattb> you can retrieve it when you like

21:01:42 <mattb> it's very similar to http://groups.yahoo.com/group/rest-messaging/message/3 but with an rdf model

21:04:59 <mortenf> hmm, yeah, but it just occurred to me that we might be talking about different issues - i was thinking about the issue of splitting up scutter/aggregator/etc..., not about the distributed notification system - but actually it seems like there's quite a lot of overlap...

21:11:48 <mattb> yeah, i see them as the same issue really

21:12:08 <mattb> scutter finds things, notifies the aggregator of *everything*, no special subs options available

21:12:20 <mattb> other things use the aggregator/store as a mediator with some query features

21:12:47 <mattb> you could run multiple distributed stores off the same scutter, for example

21:12:51 <mortenf> yep, and the "other things" could subscribe to the aggregator.

21:12:54 <mattb> exactly

21:13:05 <mattb> extra layers just give more capabilities

21:13:10 <mortenf> indeed

21:13:17 <mattb> you plug your app into the layer you like best, best impedance match

21:13:25 <mortenf> methinks there are some nice layered graphics waiting to be made...

21:13:27 <mattb> later layers could do inferencing, smushing

21:13:47 * mattb wishes he didn't have a flat to tidy, got code fingers

21:13:50 <mortenf> the problem is interfaces, and where to split the layers.

21:13:56 <mortenf> heh.

21:14:32 <mortenf> btw, did you invent your own vocab for the http header fields?

21:15:10 <Dorward> mortenf: OK, I'm working my way through this but by bit. <wot:assurance rdf:resource="morten.rdf.asc"/> - does that just point at the ASCII armoured signed version of the foaf file?

21:15:42 <mattb> yeah, didn't think too hard about vocab

21:16:00 <mattb> would be nice to have a general way of describing an http request

21:16:12 <mattb> didn't graham klyne do an rdf model of rfc822?

21:16:12 <mattb> http headers are similar

21:16:27 <mortenf> Dorward: yes.

21:16:36 <Dorward> mortenf: Thanks

21:17:13 <mortenf> i think i saw an incompete http vocab somewhere, i think eikeon was asking about it as well at some point.

21:17:22 <mortenf> s/incompete/incomplete/

21:17:57 <mortenf> Dorward: np, ask away...

21:18:43 <Dorward> mortenf: Is their a suitably lazy way to generate the current time in that ISO format you used for created?

21:19:05 <mortenf> hmm, language?

21:19:57 <Dorward> mortenf: Pretty much any, I'm on a Linux box. Perl by preference, but bash, ruby, java, python, etc are fine.

21:21:35 <DanC> wow... oasis has really gone nutso with TC automation... http://www.oasis-open.org/committees/calendar.php?month=5&year=2003&wg_abbrev=xri#full_calendar

21:21:37 <mortenf> erh, perl: strftime("%Y-%m-%dT%TZ",gmtime($stat->mtime))

21:23:21 <Dorward> mattb: Ta

21:24:11 <Dorward> mortenf: Complains about calling mtime on an underfined value, how do you init $stat?

21:25:08 <mortenf> $stat=stat($file)

21:26:09 <Dorward> mortenf: Ah, thanks. I'll fiddle with it from there.

21:26:36 <mortenf> cool, don't forget to show us when you're done :)

21:26:44 <Dorward> mortenf: :)

21:26:59 <dajobe> perl -e 'use POSIX; print strftime("%Y-%m-%dT%TZ",gmtime),"\n"'

21:27:11 <dajobe> for the current UTC time

21:27:19 <Dorward> dajobe: Handy. Thanks

21:30:20 <Dorward> Any documentation for the <lang: namespace (is namespace the right term?).

21:30:44 <mortenf> yep it is, and the should be some at the url itself.

21:30:51 <mortenf> s/the/there/

21:32:05 <Dorward> mortenf: You mean http://purl.org/net/inkel/rdf/schemas/lang/1.1#? ... Ah!

21:32:17 * Dorward learns how to interpret Schemas

21:32:25 <mortenf> pretty view (plug): http://xml.mfd-consult.dk/ws/2003/01/rdfs/?rdfs=http://purl.org/net/inkel/rdf/schemas/lang/1.1%23

21:33:26 <Dorward> Oh, must prettier :)

21:55:23 * Dorward finally notices that he slipped sideways in to another channel midconversation with mortenf

21:55:46 <mortenf> heh.

21:56:01 <Dorward> Are there any online validatation tools for FOAF (or XML in general)

21:56:23 <mortenf> .google rdf validator

21:56:24 <datum> rdf validator: http://www.w3.org/RDF/Validator/

21:56:39 <mortenf> ... is your best bet.

21:56:43 <Dorward> Ah ha

22:01:36 <Dorward> mortenf: It's giving me a warning about Element node must be qualified for <dateOfBirth> - what namespace does that come from?

22:01:55 <mortenf> that's foaf.

22:02:31 <mortenf> in my file i have xmlns="http://xmlns.com/foaf/0.1/" - the default namespace is foaf.

22:02:42 <Dorward> Ah, I see

22:05:21 * DanC toggles the 3rd of 4 items on my TODO list for the day... feels productive...

22:05:55 <mortenf> congrats, i'm 2/5...

22:06:00 * DanC looks at tomorrow, realizes he has owes WebOnt an action re I18N review that's getting stinky.

22:07:02 <Dorward> Righty, http://david.us-lot.org/foaf.rdf - first public release :)

22:08:59 <mortenf> looks nice. yet another plug/pretty view: http://xml.mfd-consult.dk/foaf/explorer/?foaf=http%3A%2F%2Fdavid.us-lot.org%2Ffoaf.rdf

22:10:47 <Dorward> That is a nice little application :)

22:13:49 * Dorward adds a plug for it at http://stone.thecoreworlds.net/contact/#foaf

22:14:46 <mortenf> nice, thanks.

22:15:01 <DanC>http://www.w3.org/TR/2003/WD-qaframe-spec-20030210/

22:15:02 <dc_rdfig> N: http://www.w3.org/TR/2003/WD-qaframe-spec-20030210/ from DanC

22:15:14 <DanC> N:|QA Framework: Specification Guidelines

22:15:14 <dc_rdfig> Titled item N.

22:15:22 <DanC> N:W3C Working Draft 10 February 2003

22:15:23 <dc_rdfig> Added comment N1.

22:15:29 <DanC> N:in last call, I think

22:15:30 <dc_rdfig> Added comment N2.

22:15:36 <DanC> N:wow... status section is over a screenful/page

22:15:37 <dc_rdfig> Added comment N3.

22:17:08 <DanC> N:odd... SOTD includes notice of open issues. I thought last call meant "all our issues are closed"

22:17:09 <dc_rdfig> Added comment N4.

22:17:10 <danbri> re foaf/rdf and 'validation', note that the site at http://www.semanticplanet.com/sources/58/polls/1661/ rather unfortunately does it wrong. It suggests that RDF data is bad where information is missed out which could be inferred from the RDF vocabulary used.

22:17:28 <danbri> eg "Statement object has wrong type

22:17:28 <danbri> In the statement <http://diveintomark.org/public/foaf.rdf#f8dy> <http://xmlns.com/foaf/0.1/interest> <http://python.org/> ., the object must be of type <http://xmlns.com/foaf/0.1/Document>" instead of

22:17:42 <danbri> ... 'the object can be inferred to be of type ....'

22:18:15 <mortenf> yeah, i think ian knows this.

22:18:21 <mortenf> (now)

22:19:15 <danbri> yes, i think so too. I wrote and lost (crash) a nicer message I wrote explaining the issues... i feel bad that our specs didn't make these distinctions clear enough...

22:20:27 <mortenf> well, it can still be of help sometimes, that approac.

22:20:35 <danbri> hmm, would anyone here be interested in a ScheduledTopicChat on FOAF?

22:20:40 <danbri> I realise we've never done that...

22:20:55 * danbri asking at wrong time of day...

22:21:03 <mortenf> yeah, perhaps to clear up the naming issues, for one?

22:21:36 <mortenf> and one agenda item for "brain storming"...

22:21:39 <danbri> Ah, I pinged some i18n folks at w3c about that earlier... had no easy answers, i'll try to summarise in a mail...

22:21:52 <danbri> brainstorming and stabilising

22:22:08 <mortenf> great, and yes.

22:22:11 <danbri> brainstorming vs stabilising, perhaps

22:22:17 <danbri> ok gotta go! g'nite all...

22:22:23 <mortenf> yeah, nn.