Semantic Web Interest Group IRC Chat Logs for 2001-03-13

00:00:19 <AaronSw> yeah, i know -- i've wanted to do this for years..

00:00:34 <SeanP> Ugh, my line will be disconnecting soon...

00:00:45 <SeanP> What, trawl through lists?

00:00:53 <AaronSw> rdf mial

00:00:56 <AaronSw> i mean mail

00:00:57 <SeanP> Oh

00:01:04 <SeanP> Do it then!

00:01:05 <AaronSw> see my.theinfo.org

00:01:16 <AaronSw> and the notes on aaronsw.com

00:01:56 <SeanP> Yep

00:02:02 <SeanP> XML-RPC

00:02:19 <AaronSw> no -- not xml-rpc, eeww

00:02:53 <SeanP> Sorry, I'm back and forth in a frenzy here... I know I'll be disconnected soon

00:03:15 <AaronSw> no probl.

00:03:34 <SeanP> SeanP takes a closer look

00:03:45 <SeanP> XML-RPC

00:03:45 <SeanP> my.info is dedicated to Internet standards like XML-RPC

00:03:52 <SeanP> there

00:04:04 <AaronSw> i should probably take that off...

00:04:08 <SeanP> Yes!

00:05:20 <AaronSw> hehe! cool!

00:05:30 <SeanP> ahem

00:06:13 <SeanP> oh no"

00:06:17 <SeanP> lol

00:06:47 <AaronSw> for those of you reading the log, and wondering what we're talking about, that was a...um...typo. yeah, that's it.

00:07:14 <SeanP> SeanP cough, cough

00:09:25 <SeanP> "microsoft.com"

00:12:14 <SeanP> SeanP takes a peek

00:12:17 <SeanP> oops

00:22:18 <SeanP> I've said too much

00:22:24 <SeanP> left channel

00:22:28 <AaronSw> i haven't said enough

00:41:33 <AaronSw> is dc ok?

01:04:42 <edd> logger, listen

03:57:38 <dirkEddel> left channel

08:58:47 <edd> edd waves

08:59:08 <danbri> hi Edd!

08:59:48 <danbri> I'm in a two day meeting (medcertain -- rdf/xml medical content rating) but will be lurking here too

09:00:26 <edd> I hope you find them productive! I've got the usual pre-publication rush on for XML.com, so won't be as talkative as y'day either.

11:45:18 Disconnected from irc.openprojects.net (ERROR :Closing Link: logger[tatooine.ilrt.bris.ac.uk] by forward.openprojects.net (Ping timeout for logger[tatooine.ilrt.bris.ac.uk]))

11:45:18 Attempting to reconnect

11:45:30 Users on #rdfig: logger dc_rdfig edd

12:20:42 Disconnected from irc.openprojects.net (Connection reset by peer)

12:20:42 Attempting to reconnect

13:47:06 Users on #rdfig: logger dajobe dc_rdfig edd

14:48:23 <dajobe> http://www.jan-winkler.de/dev/e_rdfe.htm

14:48:50 <dajobe> A:|RDFedt - RDF Editor for MS Windows

15:54:01 <jhendler> left channel

16:03:37 Disconnected from irc.openprojects.net (ERROR :Closing Link: logger[tatooine.ilrt.bris.ac.uk] by lackey.openprojects.net (Ping timeout for logger[tatooine.ilrt.bris.ac.uk]))

16:03:37 Attempting to reconnect

16:04:05 Users on #rdfig: logger dajobe dc_rdfig edd

17:04:34 Disconnected from irc.openprojects.net (ERROR :Closing Link: logger[tatooine.ilrt.bris.ac.uk] by barnes.openprojects.net (Ping timeout for logger[tatooine.ilrt.bris.ac.uk]))

17:04:34 Attempting to reconnect

17:04:37 Users on #rdfig: logger dajobe dc_rdfig edd

17:17:33 <DanC> er... no ops here? shall we all leave and come back?

17:17:49 <DanC> er... I can't set the topic.

17:17:55 <dajobe> danbri / Aaron get auto-operatored on entry

17:18:03 <DanC> ah.

17:18:14 <dajobe> I'm sure others can be annointed

17:21:55 <DanC> so... is there a bugathon scheduled? has anybody started thinking about times?

17:23:02 <dajobe> not yet, waiting for rdf tests to be sorted out - ArtB is working on that, danbri to announce this place and other things, me to tidy my demo

17:26:57 <dajobe> http://www.cse.ogi.edu/~shawn/rpe/bowersrpe.pdf

17:32:25 Disconnected from irc.openprojects.net (ERROR :Closing Link: logger[tatooine.ilrt.bris.ac.uk] by king.openprojects.net (Ping timeout for logger[tatooine.ilrt.bris.ac.uk]))

17:32:25 Attempting to reconnect

17:34:22 Users on #rdfig: logger DanC dc_rdfig edd

17:38:56 Disconnected from irc.openprojects.net (ERROR :Closing Link: logger[tatooine.ilrt.bris.ac.uk] by sagan.openprojects.net (Ping timeout for logger[tatooine.ilrt.bris.ac.uk]))

17:38:56 Attempting to reconnect

17:42:21 Users on #rdfig: logger DanC dc_rdfig edd

18:27:16 Disconnected from irc.openprojects.net (ERROR :Closing Link: logger[tatooine.ilrt.bris.ac.uk] by herbert.openprojects.net (Ping timeout for logger[tatooine.ilrt.bris.ac.uk]))

18:27:16 Attempting to reconnect

18:27:19 Users on #rdfig: logger DanC dc_rdfig edd

18:43:45 <AaronSw> Hello everyone!

18:43:51 <AaronSw> I've noticed we're in need of a topic. ;-)

18:43:55 <AaronSw> Any suggestions?

18:44:34 <SeanP> Hi Aaron

18:44:39 <SeanP> How about...

18:44:47 <SeanP> "mmmmm.....Semantic Web"

18:44:54 <DanC> pls put a pointer to the RDF IG home page in the topic

18:44:59 <SeanP> Yep

18:45:10 <SeanP> &/or SW dev

18:45:36 <DanC> FYI... Dan's trip to KT2001 in Austin, Mar 2001

18:45:36 <DanC> http://www.w3.org/2001/03dc-aus/trip

18:45:50 <AaronSw> C:|Dan's trip to KT2001 in Austin, Mar 2001

18:45:59 <DanC> nifty!

18:46:08 <SeanP> To coment, C:comment blah, blah

18:46:27 <DanC> er... that thing export to HTTP somehow, no?

18:46:30 <SeanP> It appears on http://rdfig.xmlhack.com/

18:47:04 <DanC> what's the scope of the 'C' indentifier? how long does it last?

18:47:11 <SeanP> Forever

18:47:18 <AaronSw> No, I think about a day...

18:47:21 <SeanP> THe C identifier ties a comment to the URI that you enter

18:47:28 <SeanP> Well, it's all stored isn't it?

18:47:33 <SeanP> In XML on Edd's site

18:47:36 <AaronSw> Yeah, but the identifier doesn't last.

18:47:46 <SeanP> Oh, right

18:47:49 <DanC> DanC wonders if he knows xgavin

18:48:12 <DanC> lemme try this thing out... it seems to have some WikiNature...

18:48:21 <DanC> http://www.w3.org/Architecture/NOTE-ioh-arch

18:48:23 <AaronSw> Dan, that trip link is ACLed.

18:48:31 <AaronSw> Can I look at it?

18:48:31 <SeanP> Yep... Edd called it a "Wiki for IRC"

18:48:46 <DanC> D:|An Evaluation of the World Wide Web with respect to Engelbart's Requirements

18:48:59 <SeanP> I read that evaluation...

18:49:15 <SeanP> You want to release it as a note?

18:50:18 <DanC> er... it's already released, in a way (at the ECommerce conference in austin). I guess I could re-relase it to get more folks to look at it, but I'd want to update it first.

18:50:55 <SeanP> it looks like you've been working on it for a long time

18:51:20 <DanC> grumble... no xmlns="http://www.w3.org/1999/xhtml" in http://rdfig.xmlhack.com/index.html

18:51:26 <SeanP> lol

18:51:33 <AaronSw> Dan, how do I get access to http://www.w3.org/2001/03dc-aus/trip ?

18:51:41 <SeanP> Have any of the requirements been met in the time in which you've been developing the document?

18:52:07 <SeanP> IRC integration and so on has been better I suppose...

18:52:12 <DanC> er, just clicking doesn't work, AaronSw? I thought I makde it world-readable...

18:52:27 <SeanP> "11. Every Object Addressable"

18:52:33 <AaronSw> I get W3CACL realm...

18:52:48 <AaronSw> Unauthorized.

18:53:05 <DanC> indeed... hadn't chacl'd it; try again.

18:53:16 <AaronSw> Great! Thanks.

18:53:22 <SeanP> That's a continual problem with Team stuff at the W3C, the team sometimes don't even seem to realise that there are ACLs on them

18:53:26 <DanC> DanC tries to decide whether to go to 1pmCT telcon or not.

18:53:47 <AaronSw> Yeah, we were just talking about ACL problems the other day.

18:53:56 <dajobe> dajobe catches up

18:54:06 <DanC> I've got a pile of stuff that I scribbled into another IRC channel that would really benefit from this rdfig.xmlhack.com thingy (does it have a WikiName, btw?)

18:54:13 <dajobe> scope of A: etc. identifiers is till midnight GMT

18:54:29 <AaronSw> It's named the Daily Chump -- edd is packaging it up

18:54:45 <DanC> ah. Cool.

18:55:00 <AaronSw> He said he'd release it Thursday night or so.

18:55:22 <DanC> I hope it scales beyond "daily"; I'd like to use it as an annotated bibliography.

18:55:40 <SeanP> It's be neat if the identifers were scalable...

18:55:46 <dajobe> I'm going to modify my irc->rdf stuff to have proper IDs for each statement, then use those in HTML and RDF/XMl. Then each chat event here will have a URL#fragment ref

18:55:47 <SeanP> It's = It'd

18:56:08 <DanC> no, temporary identifiers are fine; what I want long-term access to is the title etc.

18:56:18 <DanC> title, date, who-posted-it, etc.

18:56:36 <DanC> RDF/XMI? did you mean RDF/XML?

18:56:37 <SeanP> But you'd still need to reidentify that, no?

18:57:00 <AaronSw> The archives stay -- look at http://pants.heddley.com/ for a chumpt that's been running for months.

18:57:06 <DanC> re-identify: why? the temporary id was based on a URI in the first place, no?

18:57:09 <dajobe> [that's a lower case:l rather than upper: L]

18:57:34 <SeanP> Yes, but what if the title changes etc.?

18:57:35 <DanC> it's one thing for the archives to not go poof; it's another to be able to efficiently navigate them when they get big.

18:58:25 <dajobe> see RDF/XML [L, not l] logs at http://ilrt.org/discovery/chatlogs/rdfig/ and I still have to tweak the schema, add ChatEvent types. Maybe topic/title changes would be possible too.

18:58:25 <AaronSw> Very true.

18:58:26 <DanC> DanC finds following a link to a search page tedious; a little search field/box would be 1000% easier.

18:59:02 <AaronSw> Alright, lunch is over -- got to get back to class.

18:59:10 <AaronSw> Thanks for the chat!

18:59:12 <DanC> phtphpht. search gizmo violates the "when to use GET vs. POST" principle.

18:59:14 <SeanP> I've gotta go to

18:59:27 <SeanP> See y'all in about an hour or so...

18:59:33 <dajobe> SeanP: another TV show?

18:59:34 <AaronSw> DanC: tell edd about it.

18:59:44 <AaronSw> he'll fix it

18:59:48 <AaronSw> see you all later!

18:59:51 <SeanP> Yes! The Simpsons

18:59:55 <SeanP> Heh, heh

18:59:56 <DanC> can I tell edd by annotating the thing somehow, I wonder...

19:00:04 <SeanP> bye

19:00:07 <SeanP> left channel

19:06:10 <dajobe> B:|A Generic Approach for Representing Model-Based Superimposed Information

19:07:29 <dajobe> B: Comparing RDF, RDF Schemas, Topic Maps and XML representations by superimposing the model, schema and instance information into one layer

19:15:17 <dajobe> logger, listen

19:16:42 <dajobe> to get to logs follow URL in topic to get to http://ilrt.org/discovery/chatlogs/rdfig/

19:17:51 <bwm> Ah - whats this topic thing - do I have to press a button or something?

19:18:10 <bwm> din dins approaches

19:18:21 <dajobe> not sure on irc terminology, but basically a <title> for irc channels

19:18:58 <dajobe> aah food. US jellybeans from Texas went down a treat in office today

19:19:22 <bwm> You are back in uk. How was texas, apart from large.

19:19:54 <dajobe> dajobe tries to sum it up in one word

19:20:02 <dajobe> big

19:20:28 <bwm> Is 'big' a big word?

19:22:06 <bwm> I need go.

19:22:55 <edd> edd wakes up

19:23:27 <edd> edd wonders how the search can violate GET/POST principle when he hasn't implemented it yet.

19:24:28 <edd> edd reads back further and guesses DanC meant the htdig stuff from the other Daily Chump

19:25:06 <edd> i'm afraid that won't change quickly until i find something else that's as quick to set up and pleasant to use.

19:27:12 <DanC> other chump: yes.

19:31:41 <edd> edd notes that there's also a command-line driver for the Chump engine.

19:35:01 <xgavin> xgavin can't wait for the FCS of the chump code.

19:35:25 <edd> edd currently tearing his hair out over configuring wireless lan

20:07:34 <SeanP> http://dublincore.org/documents/2001/03/09/dcmi-namespace/

20:07:48 <SeanP> E:|DRAFT Namespace Policy for the Dublin Core Metadata Initiative (DCMI)

20:08:22 <SeanP> Grr, I don't like the bit "Declarations for individual elements are constructed by adding #entity_name, for example:- http://purl.org/dc/elements/1.1/#title"

20:09:37 <dajobe> it's meant to be easy to understand for non-RDFers

20:10:08 <SeanP> Well, I just mean that up until now we never usd to add the #

20:10:16 <SeanP> THe namespace redirect at PURL did it for us

20:10:25 <SeanP> And now they're saying it should be changed?

20:10:50 <SeanP> How many examples already leave out the "#" - that's a lot of out-of-date namespaces

20:11:00 <dajobe> that looks like a mistake

20:11:17 <SeanP> MAybe

20:11:23 <SeanP> I'll mail DCMI Arch

20:11:25 <SeanP> E:Form the pre-amble "The use of URI-based namespaces for formal, machine-processable declarations of metadata entities is a convention intended to support web-addressable concepts that can be shared across applications, and hence promote the possibility of shared semantics."

20:11:44 <SeanP> oops, form = from...

20:12:15 <SeanP> Ugh, I have to go again...

20:12:23 <SeanP> left channel

20:26:51 <dajobe> you quoted #rdfig on your dc-architecture posting to the 'public' before its launch ... and without mentioning what it was

20:27:12 <dajobe> I need to post myself

20:27:34 <SeanP> Come again? It's mentioned on the RDF IG page somewhere

20:27:41 <dajobe> I said launched

20:28:03 <SeanP> Well, it's been mentioned before :-)

20:28:41 <SeanP> No one told me about it - I found it from the W3C site somewhere

20:29:45 <SeanP> SeanP goes looking for the reference

20:31:05 <dajobe> don't waste time doing that, #rdfig will be announced formally by Friday

20:32:15 <SeanP> Good

20:32:29 <SeanP> Why hasn't it been announced already?

20:32:57 <dajobe> patience!

20:33:06 <SeanP> :-)

20:33:12 <SeanP> SeanP is always impatient

20:33:13 <dajobe> logs, stable set up, moderators, collaborative workspace

20:33:26 <dajobe> s/moderators/operators/

20:33:51 <SeanP> Yes, but that's been done already...

20:34:03 <SeanP> We have a log and an IRC Wiki!

20:34:13 <dajobe> yes, we got it on *Monday* - i.e. yesterday

20:34:38 <SeanP> O.K. then, I'll let you off :-)

20:38:33 <SeanP> left channel

21:01:23 <jhendler> w/respect to the chat logs, for some reason I've worked on a couple of machines today that cannot resolve ilrt.org - perhaps the www.ilrt.bris.ac.uk could be used as well - or at least a note put somewhere? I realize the brain damage is probably on my side, but thought I'd mention it...

21:49:38 <DanC> JimH, ilrt.org runs in danbri's basement or something; don't be so quick to conclude that the damage is on your side ;-)

21:50:21 <dajobe> ilrt.org is just a domain name pointing at www.ilrt.bris.ac.uk the ilrt.org nameservers are in Germany rather than England; web server is same, and in Bristol

21:52:52 <SeanP> Hi Aaron

21:52:57 <AaronSw> Hi there.

21:53:04 <SeanP> Join the hustle

21:53:15 <AaronSw> How'd we lose the topic again?

21:53:17 <AaronSw> Hmmm.

21:53:27 <SeanP> Its still there!

21:53:33 <SeanP> *** Topic is 'Semantic Web Live! (RDF-IG Chat, see http://rdfig.xmlhack.com/)'

21:53:46 <AaronSw> Odd -- I don't see it.

21:54:08 <SeanP> Maybe you don't get to see it because you set it...

21:57:31 <AaronSw> Is it just me or is the 'hack down?

21:57:44 <SeanP> The 'hack?

21:57:52 <AaronSw> XMLhack

21:58:09 <SeanP> SeanP checks

21:58:12 <AaronSw> which would help explain why edd is gone...

21:58:17 <SeanP> Yes

21:58:22 <SeanP> Ugh, gotta go!

21:58:25 <SeanP> left channel

21:59:55 <dajobe> from here UK -> US network connectivity has been up/down all day but xmlahck.com works for me now; last update 19:22 yesterday

22:00:36 <AaronSw> Odd... that's probably it.

22:02:06 <AaronSw> E:

22:02:24 <AaronSw> Oh, our wiki is down too

22:02:26 <AaronSw> Hmm.

22:02:39 <AaronSw> I mean the chump.

22:02:44 <dajobe> well dc_rdfig isn't here

22:02:54 <AaronSw> Yeah, where'd he go?

23:01:35 <dajobe> dajobe wonders what SeanP was watching from 22:00-23:00 GMT

23:01:53 <SeanP> Frasier, and then Frank Skinner... the Ozzie Osbourne interview, great!

23:02:09 <SeanP> I'm still laughing now

23:02:20 <dajobe> Ahah - Paramount channel

23:02:28 <SeanP> Yep

23:03:43 <SeanP> It's good that Tod has point forward the notion to correct the DCMI FragID thing

23:04:07 <dajobe> I said "probably a mistake" earlier

23:04:25 <dajobe> but the new specs. will probably use fragid

23:04:27 <SeanP> Yes, it's only a draft so that's good

23:04:48 <dajobe> and I used to laugh at the phrase "legacy metadata"

23:04:49 <AaronSw> Sean, what do you think of my comments?

23:05:01 <SeanP> Aaron: Just going through them now

23:05:58 <SeanP> Ah, the quibble with date stamping

23:06:30 <AaronSw> You know me. ;-)

23:06:45 <SeanP> What if there were many versions of one schema produced? Would /1.0/ /1.1/ be able to handle that?

23:07:02 <AaronSw> Why not?

23:07:26 <dajobe> using dublincore.org in the formation of the base URI is a good social indication of who controls the namespace. the dates rather than version numbers indicates that these aren't necessarily alternatives but can be mixed - modular metadata, that's what its about

23:07:35 <SeanP> Just wondering if date stamping is indeed harder to follow than versions, in which case your objections would be very well founded

23:08:00 <AaronSw> wrt dublincore.org, /dc/ does the same thing; and control is likely to change over time

23:08:13 <SeanP> Yes, the less information, the better

23:08:22 <AaronSw> wrt dates, version numbers can be mixed to (i.e. 1.1 implies additions to 1.0)

23:09:01 <AaronSw> (and 2.0 implies something very new)

23:09:06 <AaronSw> (or atl east different)

23:09:14 <AaronSw> I mean at least.

23:09:16 <SeanP> Well, maybe you'd want three or more tiers

23:09:19 <SeanP> 1.0.0

23:09:21 <SeanP> 1.0.1

23:09:24 <SeanP> 1.1.0

23:09:25 <SeanP> etc.

23:09:31 <dajobe> 1.1 vs 1.0 - more likely 1.1 adds new terms to 1.0; both can be used together. sticking this in the URL isn't too clever. How about major.minor.version....

23:09:50 <AaronSw> I'm just trying to keep backwards compatible.

23:09:56 <dajobe> anything that means parsing URLs or expecting to do it is naughty

23:10:03 <SeanP> agreed

23:10:10 <SeanP> But naming is useful not just for machines

23:10:11 <AaronSw> I don't expect parsing URLs -- I'm just talking humans here.

23:10:13 <dajobe> that came out slightly wrong, but you get what I mean

23:10:15 <SeanP> Yes

23:10:51 <AaronSw> URLs are strings for the purposes of this -- I'm just asking what strings are best for the people typing them in.

23:11:02 <dajobe> humans - which is why non-version numbers can help in some cases to indicate the persistance. 2001-02-xx is not going away, but more stuff may arrive in 2002-10-yy etc.

23:11:02 <SeanP> But what qualities are we looking for in the namespaces? Persistence, maintainability, versioning

23:11:21 <AaronSw> What makes you think /1.0 will go away?

23:11:28 <SeanP> Unfortunately, these requirements all conflict on the data that you'd have in the URI

23:11:30 <dajobe> if a /1.1 apperas

23:11:40 <dajobe> conflicts - indeed

23:11:40 <AaronSw> [ when DanC gets back, remind me to ask him about RDF mime types ]

23:11:52 <SeanP> Aaron: Yes, that's impotant

23:11:56 <AaronSw> Why would 1.1 eliminate 1.0?

23:12:00 <SeanP> important...

23:12:02 <dajobe> humans might think that

23:12:10 <SeanP> sometmies it might do

23:12:20 <AaronSw> Yeah, but those are the same humans who'd expect 2001 to overwrite 2000!

23:12:29 <SeanP> Maybe it will!

23:12:42 <dajobe> in which case write a document or schema that says what is 'current' or 'overridden'

23:13:00 <AaronSw> Exactly -- that's in the proposal, and I don't argue with it.

23:13:04 <dajobe> I have to take a 2nd reading of that post; came in a busy time for me to digest

23:13:05 <DanC> DanC wonders who rang...

23:13:15 <AaronSw> Hello there.

23:13:26 <SeanP> Bonding metadata at the end of a URL to the URL definition itself is a good idea

23:13:31 <SeanP> Hi Dan

23:13:39 <AaronSw> DanC: I heard that you were writing something on RDF mime types

23:13:40 <SeanP> Aaron: Ask him about the MIME type :-)

23:13:47 <DanC> on RDF mime types, see http://www.w3.org/2001/03mr/rdf_mt

23:14:20 <SeanP> model/?

23:14:59 <SeanP> Did you see Aaron and myself discussing this the other day in the logs?

23:15:06 <dajobe> I'd prefer a text/ something (also?) for rdf intended to be read

23:15:09 <DanC> I thought model/rdf was the way to go, but read the model/* RFC and realized that model/* is constrained to models of the 3-dimensional physical world.

23:15:33 <AaronSw> Isn't it +xml?

23:16:01 <DanC> yes, it's +xml; dunno where I got confused into thinking it's xml+

23:16:02 <SeanP> I think that application/* doesn't stop it from being human-readable

23:16:14 <AaronSw> i think they changed it late in the game to +xml

23:16:16 <SeanP> Aaron suggested *+rdf+xml

23:16:25 <SeanP> or *+rdf

23:16:29 <dajobe> text/* indicates to most browsers that it can be viewed, rather than throwing up 'Save As'

23:16:42 <AaronSw> dajobe: rdf can't really be viewed

23:16:51 <dajobe> only by rdf-geeks

23:16:53 <AaronSw> I wasl also thinking about n3+rdf, etc.

23:16:56 <SeanP> Agreed, but then current browsers won't support text/n3+rdf anyway

23:17:15 <SeanP> So you may as well go for application/*

23:17:16 <DanC> text/* is supposed to indicate to user agents to fallback to text/plain rather than "save as?" but in practice, there aren't many user agents that conform to that part of the MIME spec. Sigh.

23:17:40 <SeanP> That's another problem: theory/practise

23:18:16 <DanC> in practice, we might as well chose harpdy/blrbf for the media type name; as long as we all use the same name, it doesn't matter what it is.

23:18:38 <dajobe> indeed

23:18:49 <DanC> in fact, I'm not sure it's essential to have a MIME type for RDF at all, except for the "@@note that publication of RDF in HTTP or sending via SMTP amounts to asserting

23:18:49 <DanC> the content."

23:19:17 <AaronSw> Hehe.

23:19:25 <SeanP> Well, it's getting impossible to sniff out RDF because of the range of formats...

23:19:34 <SeanP> A MIME would at least help, wouldn't it?

23:19:41 <DanC> that was what motivated me to write the thing. putting <doc>I agree to pay the sum of $500</doc> in foo.xml doesn't oblige one to pay anything.

23:19:43 <AaronSw> Yes, how would RDF profiles work? Like RSS or PRISM.

23:20:10 <DanC> sniff out RDF: is there a need for that in practice?

23:20:16 <AaronSw> I tend to think it should be application/rss+rdf+xml

23:20:22 <SeanP> Well, for SW trawlers maybe

23:20:39 <DanC> hypothetical SW trawlers, or something real?

23:20:47 <SeanP> Google gets text/html, what would SWGoogle get?

23:20:50 <AaronSw> I thought that that was what rdf:RDF was for -- to indicate RDF.

23:20:55 <DanC> what's SWGoogle?

23:20:58 <SeanP> But you don't have RDF:rdf in N3

23:21:11 <SeanP> oops, rdf:RDF

23:21:12 <AaronSw> You also don't have much N3. ;-)

23:21:14 <jhendler> left channel

23:21:30 <SeanP> SWGoogle is anything that trawls the SW looking for informations

23:21:37 <SeanP> no "s"

23:21:55 <SeanP> Aaron: But people shouldn't be limited to XML RDF

23:22:03 <SeanP> Or even N3

23:22:11 <DanC> the question of a 'magic number' for N3 is an interesting one; a candidate is: <> <#writtenIn> <http://www.w3.org/DesignIssues/Notation3>.

23:22:13 <AaronSw> I agree, but N3 should find a way to embed it self in other stuff.

23:22:31 <SeanP> Isn't there already an identifer for N3 in TimBL's logic schema?

23:22:46 <AaronSw> problem with that is that there's no way to start and end blocks of N3...

23:22:59 <SeanP> THere are lots of probs with N3

23:23:06 <SeanP> But it's still very useful

23:23:10 <DanC> start/end: sure there is: MIME, tar, zip, you name it. Not to mention N3's own {}.

23:23:22 <SeanP> <> a <http://www.w3.org/2000/10/swap/logic.n3#N3Document> .

23:23:33 <SeanP> - http://www.w3.org/DesignIssues/Notation3

23:23:37 <AaronSw> yeah, but none of those are easy for humans to do, which I thought was the point of N3.

23:24:00 <DanC> humans: right; to date, nobody has found an N3 magic number to be cost-effective.

23:24:18 <AaronSw> true.

23:24:28 <DanC> on the other hand, I always have to look up the URI for @prefix log: <...>, so one more URI at the top wouldn't hurt too much.

23:24:45 <SeanP> Yeah, I can never remember that either

23:24:54 <SeanP> Why is the schema separate from the namespace???

23:25:00 <DanC> http://www.w3.org/DesignIssues/Notation3

23:25:10 <DanC> er... where did DailyChump go?

23:25:16 <SeanP> It pinged out

23:25:17 <dajobe> at some point I can see myself writing a n3 parser, which would be triggerd by getting some MIME-typed content via http or other means. A mime type is needed for me to pick the right parser for N3, rather than assume type was rdf/xml or do content-identification guessing games

23:25:24 <AaronSw> Yeah, i noticed that too -- it seems to have disappeared.

23:25:37 <AaronSw> we ought to email edd about it.

23:25:38 <SeanP> It went when Edd's server went down I presume

23:25:52 <AaronSw> but it never came back up

23:26:08 <SeanP> Is XMLHack still down then?

23:26:19 <AaronSw> nope.

23:27:29 <SeanP> right, I see it now

23:27:29 <AaronSw> it came back right after i started complaining. ;-)

23:27:29 <DanC> if you're picking N3 out of MIME, don't forget to pay attention to content-disposition, i.e. the property that relates the message to the attachment, which might be "disclaims" or "refutes" or some such.

23:27:29 <dajobe> n3 mime - I mean I could have HTTP header and need to make a decision

23:27:29 <AaronSw> danc, i think your going to have a hard time getting the semantics of rdf past the IETF folks...

23:28:12 <AaronSw> in this case, semantics being "publication of RDF in HTTP or sending via SMTP amounts to asserting the content"

23:28:38 <AaronSw> i doubt they'll go for that easily, they haven't done that for anything else, have they?

23:29:04 <dajobe> AaronSw: IETF != IANA, who register types and things

23:29:33 <AaronSw> Yes, but it looked like DanC was writing an IETF draft, no? which is the only way to get IANA to register things, IIRC.

23:29:43 <SeanP> http://www.iana.org/

23:29:56 <DanC> past what IETF folks? I don't need anybody's blessing to register application/rdf; I just publish an informational RFC and go.

23:30:05 <SeanP> http://www.iana.org/cgi-bin/mediatypes.pl

23:30:59 <AaronSw> but don't you need some portion of the ietf to approve the rfc?

23:31:06 <AaronSw> AaronSw is a newbie to ietf process

23:31:41 <DanC> yes, but the portion of the IETF I have to get by doesn't have the right to object to technical design of my MIME type; only to validity of my registration document.

23:32:10 <AaronSw> ahh, i see. but they may still claim that the bit about assertion is out-of-scope for the registration.

23:32:21 <SeanP> But surely somewhere along the line someone needs to discuss the technical validity of the RFC?

23:32:44 <DanC> hmm... maybe.

23:33:01 <DanC> anyway, if I have to get thru some IETF hurdles, it won't be the first time.

23:33:25 <DanC> let's consult The Rules: ftp://ftp.isi.edu/in-notes/rfc2048.txt

23:33:55 <AaronSw> i'm reading them now...

23:34:02 <DanC> written by the Honorable Freed, Klensin, Postel, worthy opponents one and all (Postel: God rest his soul.)

23:34:19 <SeanP> snap

23:34:24 <AaronSw> amen.

23:34:30 <AaronSw> (to Postel)

23:34:49 <AaronSw> It says they have the right to veto bogus security considerations

23:35:54 <AaronSw> it says that if ietf-types raises any objectons, the area director can veto it.

23:36:00 <AaronSw> (section 3.2.1)

23:36:03 <DanC> which area director?

23:36:36 <AaronSw> of ietfi-types

23:36:41 <AaronSw> which would be..

23:37:11 <DanC> IETF Applications Area Director (3.2.2.)

23:37:41 <AaronSw> so faltstrom and freed

23:37:45 <DanC> apps area director these days is Freed himself, I think; lemme check... I should know this; I have a telcon scheduled with them in a few weeks.

23:37:53 <DanC> thx.

23:38:12 <SeanP> That's useful...

23:38:45 <DanC> God Help Us All if the RDF media types gets into that telcon. That telcon is for the worst nightmares only.

23:39:05 <DanC> stuff like XML/URI stuff ;-)

23:39:19 <SeanP> Ah :-)

23:39:20 <AaronSw> Hehe. IETF/W3C let's get along better one?

23:39:24 <DanC> yes.

23:39:28 <SeanP> Heh heh

23:39:43 <SeanP> Dan: Why on earth did you choose MIME for the WWW anyway :-)

23:40:14 <DanC> get along better: it's more like "tell us what the hell is going on over there; we've got so much stuff going on over here that we don't spot your horrendous architectural issues until you've already standardized them" (in both directions)

23:40:58 <AaronSw> ahh, i see.

23:41:02 <AaronSw> that's useful.

23:41:13 <SeanP> That's the inherent problem with multiple standardization bodies

23:41:32 <DanC> MIME for WWW: have you read the archives on that?

23:41:38 <SeanP> Yes, many times

23:41:46 <DanC> mixed blessing, that.

23:41:59 <SeanP> Well, we have a Web don't we?

23:42:09 <AaronSw> Any plans to move to URIs? Or do we just use CC/PP?

23:43:00 <DanC> plans to move to URIs: in fact, the SMIL chair just asked me that the other day; in SMIL, they have some sort of attribute where you give the URI of a feature that the host implementation might or might not have...

23:43:26 <DanC> ... and they asked what URI to use to test for support for MIME type foo/bar; I told them ftp://isi.edu.../foo/bar, of course, per the RFC.

23:43:52 <DanC> which reminds me... gotta double-check that RFC...

23:44:09 <AaronSw> can you send me the url, too?

23:44:16 <SeanP> SeanP is looking for it now

23:44:53 <DanC> 2.5. Location of Registered Media Type List

23:44:54 <DanC> Media type registrations will be posted in the anonymous FTP

23:44:54 <DanC> directory "ftp://ftp.isi.edu/in-notes/iana/assignments/media-types/"

23:44:54 <DanC> and all registered media types will be listed in the periodically

23:44:54 <DanC> issued "Assigned Numbers" RFC [currently STD 2, RFC 1700].

23:44:59 <DanC> -- ftp://ftp.isi.edu/in-notes/rfc2048.txt

23:45:16 <SeanP> thanks

23:46:09 <DanC> anybody know implementation detials of Daily Chump? php hack?

23:46:15 <SeanP> BTW, you may want to update http://www.w3.org/Addressing/schemes.html with this new "Protozilla" stuff that Aaron was pointing out to me the other day

23:46:45 <DanC> yes, schemes.html doesn't get enough of my attention. 1000 brownie-point bonus for an N3 version of schemes.html

23:46:52 <AaronSw> You may want to update that page with a lot of stuff. Most of it I've sent to uri@w3.org. ;-)

23:47:06 <SeanP> An N3 version? I'll give it a go

23:47:15 <AaronSw> I knew he'd go for that! ;-)

23:47:24 <SeanP> he he he

23:47:41 <DanC> hmm...

23:47:50 <SeanP> As for daily chump...

23:47:52 <DanC> "http" a URIScheme.

23:48:00 <DanC> doesn't work. can't use strings as subjects.

23:48:13 <SeanP> data:,http a :URIScheme

23:48:13 <AaronSw> Re media types: http://logicerror.com/mediaType-uri

23:48:24 <DanC> <#http> a URIScheme; :schemeName "http".

23:48:36 <AaronSw> Ahh, much better.

23:48:43 <SeanP> Good idea

23:49:00 <DanC> DanC wonders about the relationship of :schemeName and rdfs:label

23:49:05 <DanC> and dc:title, while we're at it.

23:49:22 <SeanP> I'll work on it and send it over... or just post the URI here

23:49:44 <DanC> in most KR systems, rdf:Resource is called Thing, and there's a 'name' property that would be a superproperty of things like rdfs:label, dc:title, etc.

23:49:45 <SeanP> as for Daily Chump... Edd was talking to me about how it was implemented, but I can't access the logs!

23:50:11 <SeanP> Like dam:thing

23:50:13 <SeanP> daml

23:50:41 <DanC> whoohoo! RSS really is driving the deployment of generic RDF tools, isn't it? (cf RDFedt)

23:51:05 <SeanP> Yes, "rdfs:label" and "dc:title" have a very close relationship... but it isn't piinted out in wither schema

23:51:21 <DanC> I sometimes use rdfs:label to mean traditional-kr:thing, but I wonder if that's abuse; e.g. is it reasonable to say that dc:title rdfs:subPropertyOf rdfs:label?

23:51:38 <SeanP> I think so

23:52:12 <DanC> also... is it reasonable to give a name to html:title, or is dc:title all we need?

23:52:29 <dajobe> best to keep rdfs:label separate as KR concept of name of thing

23:52:33 <SeanP> Give a name to html:title? What do you mean?

23:52:41 <DanC> there's no "view source" link on rdfig.xmlhack.com

23:53:32 <SeanP> dc:title is defined as "A name given to the resource", rdfs:label "Provides a human-readable version of a resource name".

23:54:17 <SeanP> html:title is an XHTML document node title

23:54:27 <AaronSw> see http://rdfig.xmlhack.com/index.xml -- anyone want to make it rss?

23:55:00 <dajobe> rss the items you mean, not the comments?

23:55:15 <AaronSw> the comments could go to, but the items first.

23:55:34 <DanC> html:title: I mean: { :r a http:Reply; :from :r; :content :bytes. :bytes :xmlParseAs :dom; :dom [ :xpath "html/head/title"] :t } log:imples :r html:title "t. (where html/ should be namespace-qualified)

23:56:00 <DanC> s/"t/:t/

23:56:48 <SeanP> That's an eloquesnt way of representing the transfer

23:56:49 <DanC> sean presents a compelling argument that dc:title rdfs:subPropertyOf rdf:label, if not daml:samePropertyAs.

23:57:15 <DanC> oops... I used :r for both the reply and the resource. didn't mean to.

23:57:28 <dajobe> hmm, dc:title does not necessarily have to be human readable. picking hairs.

23:57:31 <AaronSw> hey, my email to edd worked -- the chump is back!

23:57:32 <DanC> welcome back, dc_rdfig; where have you been?

23:57:32 <SeanP> Maybe even the other way around

23:57:46 <SeanP> So if dc:title doesn not have to be human readable...

23:57:51 <AaronSw> e:

23:57:52 <AaronSw> E:

23:57:55 <SeanP> then rdfs:label is a subproperty of that

23:58:26 <SeanP> So, rdfs:label daml:subPropertyOf dc:title .

23:58:36 <dajobe> SeanP: rdfs is self consistent, it doesn't subclass any other external concepts. bad idea

23:58:40 <AaronSw> E:I think this is just awful -- not more date-based namespaces, ugh!

23:58:44 <DanC> I think dc:title is fairly clearly intended to have a domain smaller than all resources: fairly traditional documents.

23:58:57 <dajobe> yup

23:59:09 <SeanP> But it doesn't even ask for human readability, maybe that is a DCMI prose mistake

23:59:26 <AaronSw> BLURB: the workings of the chump

23:59:38 <AaronSw> F:DanC wanted to know what the chump was written in

23:59:39 <dajobe> no; all dc elements can take URIs pointing at anything, external vocabs etc. so readability isn't required

23:59:49 <DanC> AaronSw, you don't think that using dates in URIs is the greatest thing since sliced bread?

23:59:51 <SeanP> But for rdfs:label it is

23:59:59 <AaronSw> hehe, sorry danc, but no.