00:04:22 danja has joined #rdfig 00:05:54 ya'all in SF now then? 00:06:27 Nope, not yet. :-( 00:06:35 I leave tomorrow morning. 00:21:26 you aren't *that far though* 00:21:44 synched with danc's Palm? 00:22:03 Well, I'm not as far as you, I suppose. 00:22:12 Why would I be messing with dan's Palm? 00:22:31 hurr hurr 00:22:57 (dirty inglis person talk) 00:23:43 sorry though Aa 00:24:11 what fresh ideas have you had in the last coulpa months? 00:25:04 umm, well, not too many ideas... 00:25:16 developments for SWAG? 00:25:24 I wrote an HTTP archiver yesterday... http://logicerror.com/archiverProxy-code 00:25:38 python? 00:25:42 yep 00:25:42 AaronSw's been busy trying to get 5000th post on my site :-) 00:25:52 ;-) 00:26:00 SWAG's been pretty quiet, Sean registered a URN for it. 00:26:32 wmf - do I know you? 00:26:38 probably not 00:26:42 ok 00:27:00 I'm danny, I do metadata... 00:27:03 googler has joined #rdfig 00:27:08 googler, Wesley Felter 00:27:09 Wesley Felter: http://felter.org/wesley/ 00:27:09 bot? 00:27:22 the bot is good 00:27:37 Yes, danja, I am a bot. 00:27:57 from Austin, texas? 00:28:01 yep 00:28:38 ok wmf, I love you already because : 00:28:42 "quality of life" 00:28:59 So what's with everyone linking to "The Right to Read" from their homepages? 00:29:09 am I? hmmm 00:29:13 I'm not - where's that? 00:29:37 wmf is, and a saw a bunch of others in the past few days 00:29:47 url? 00:29:47 it's because it's coming true 00:29:58 http://www.gnu.org/philosophy/right-to-read.html 00:29:59 A: http://www.gnu.org/philosophy/right-to-read.html from wmf 00:30:26 A:|The Right To Read 00:30:26 titled item A 00:30:45 Aaron - any chance of a precis? 00:30:57 it looks like a long essay 00:30:59 use a debugger, go to jail 00:31:07 :-) 00:31:21 what about the russo guy? 00:31:31 also in jail 00:31:33 ;-) 00:31:45 I love you yanks ;-) 00:32:18 wmf - what you do? 00:32:45 starting tomorrow, systems software research 00:32:54 00:33:01 Starts tomorrow, does it? Have fun... 00:33:07 researche qui? 00:33:20 English, please! 00:33:27 * AaronSw wonders if this means HTP will be updated earlier or later... 00:33:45 only between 7-11 PM 00:33:48 researching what (and what is HTP Aaron, please) 00:33:50 Wesley will have to stop sleeping in! 00:33:55 googler, Hack the Planet 00:33:56 Hack the Planet: http://wmf.editthispage.com/ 00:33:59 indeed 00:34:19 excellent url 00:34:27 operating systems, clusters, network protocols, hardware-software interaction... 00:35:07 errm, but what will you be having for breakfast? 00:35:35 a good question 00:35:40 has this been blogged before Aa 00:35:42 ? 00:35:51 has what, danja? 00:36:02 http://wmf.editthispage.com/ 00:36:02 B: http://wmf.editthispage.com/ from danja 00:36:12 B:|Hack the Planet 00:36:12 titled item B 00:36:19 B:| Hack the planet 00:36:19 titled item B 00:36:25 B:Featuring Wesley Felter, David McCusker, and other notables 00:36:26 commented item B 00:36:27 B:|Hack the Planet 00:36:27 titled item B 00:36:42 B: if nothing more, a cool title ;-) 00:36:42 commented item B 00:37:25 Wesley, that's a m ethd 00:37:29 oops 00:37:35 a methodist name 00:37:46 so unlike the old testament 00:37:55 you usually see here 00:38:01 but I'm Episcopal/Baptist 00:38:05 Hmmm... Publications: Coming soon. (No, really!) 00:38:12 You've been saying that for a while now... ;-) 00:38:14 no, really! 00:38:28 ? 00:38:33 Are you writing a book on web architecture? ;-) 00:38:42 I am not in control of the journal 00:38:50 you write wes? 00:39:06 well, writing papers is part of my job 00:39:30 * danja not actually figured what Wesley does... 00:39:43 I'm not sure Wesley has either. ;-) 00:39:53 AaronSw: I thought YOU were writing the Web arch book 00:40:07 Aaron? 00:40:15 Right. I was just making sure you were staying off my turf. ;-) 00:40:22 LOL 00:40:24 cool 00:40:41 question wes 00:40:44 ok 00:40:55 are you writing something on SW then? 00:40:59 no 00:41:11 then what's Aa on about? 00:41:20 he's just kidding 00:41:28 doh 00:41:51 not fair - I used his ontology in a book... 00:42:04 Who's? 00:42:13 you'res 00:42:19 which one? swag? 00:42:27 moment.. 00:43:42 * wmf updates his home page 00:43:43 ok 00:43:44 / start of Admin.java 00:43:45 package com.wrapper.vocabulary; 00:43:45 import com.hp.hpl.mesa.rdf.jena.model.Model; 00:43:45 import com.hp.hpl.mesa.rdf.jena.model.Resource; 00:43:45 import com.hp.hpl.mesa.rdf.jena.model.Property; 00:43:46 import com.hp.hpl.mesa.rdf.jena.model.RDFException; 00:43:48 import com.hp.hpl.mesa.rdf.jena.common.ErrorHelper; 00:43:50 import com.hp.hpl.mesa.rdf.jena.common.PropertyImpl; 00:43:52 import com.hp.hpl.mesa.rdf.jena.common.ResourceImpl; 00:43:54 public class Admin { 00:43:56 public static Property errorReportsTo = null; 00:43:58 public static Property generatorAgent = null; 00:44:00 public static Property prefix = null; 00:44:02 // namespace of this vocabulary 00:44:04 protected final static String uri = 00:44:06 "http://webns.net/mvcb#"; 00:44:08 00:44:10 // Properties 00:44:12 final static String nerrorReportsTo = "errorReportsTo"; 00:44:14 final static String ngeneratorAgent = "generatorAgent"; 00:44:16 final static String nprefix = "prefix"; 00:44:18 public static String getURI() { 00:44:20 return uri; 00:44:22 } 00:44:24 static { 00:44:26 try { 00:44:28 errorReportsTo = new PropertyImpl(uri, nerrorReportsTo); 00:44:30 generatorAgent = new PropertyImpl(uri, ngeneratorAgent); 00:44:32 prefix = new PropertyImpl(uri, nprefix); 00:44:34 } catch (RDFException e) { 00:44:36 e.printStackTrace(); 00:44:38 } 00:44:39 Cool, althought that's more Ken's work than mine... 00:44:40 } 00:44:42 }// end of Admin.java 00:44:44 still not impressed with A LOT OF jENA 00:44:53 * AaronSw wonders what Wesley updated on his homepage. 00:44:56 I want kindaLike ;-) 00:45:10 the school section 00:45:21 Ooh, a degree. 00:45:47 kindaLike: we're working out a deal with the DAML folks... we're still negotiating over royalties ;-) 00:45:48 you can buy a Kalashnikov in your your cornershop, yes Aaron 00:45:50 ? 00:46:20 Cycosis 00:46:26 ??? 00:46:43 ok, Cyc 00:46:48 + 00:47:18 osis (greek or latin - doesn't matter - just machine-gun a member of staff) 00:47:52 SUO, innit 00:48:20 deltab has joined #rdfig 00:48:42 more espians... 00:49:00 what happened to http://rdfig.xmlhack.com/ just? 00:49:31 ? 00:49:50 errors about not being able to find index.html 00:49:54 So what happened to espians.com or whatever... you guys get bought by Microsoft? 00:49:58 did the overnight logging bug ever get sorted? 00:50:17 A:a sci-fi exploration of ubiquitous digital rights management 00:50:18 commented item A 00:50:33 rdig.xmlhack.com works for me 00:50:44 it's working again now 00:51:17 has been for over ten minutes 00:51:36 long time in danc's memo... 00:51:59 flash? 00:52:38 !!! rock and roll !!! dudE! 00:54:05 the words look nice wmf - description of techno?? 00:54:37 I don't understand the question 00:54:54 erm - is XML involved? 00:55:05 in what? 00:55:24 whatever the fuck that site is doing 00:55:41 I don't know much about ESP; I'm not involved in it 00:55:59 * danja sorry for brutto 00:56:28 * danja a bit naive 00:56:37 danja: which site? 00:57:06 10 lines back 00:57:09 I think he means http://espnow.com/ 00:57:14 yep 00:57:31 indexical 00:57:33 we use XML in a few things we're doing 00:57:52 e.g. http://beta.espsetup.com/ 00:58:11 * AaronSw wonders how many esp domain names you guys have... 00:58:40 a server for the freenet network protocol which allows
anyone to freely publish or view information of all kinds 00:58:52 http://beta.espsetup.com/ 00:58:53 C: http://beta.espsetup.com/ from danja 00:59:06 it's worth blogging! 00:59:40 AaronSw: 24 00:59:41 C:| ESP 00:59:42 titled item C 01:00:00 c: a server for the freenet network protocol which allows
anyone to freely publish or view information of all kinds 01:00:18 C: a server for the freenet network protocol which allows
anyone to freely publish or view information of all kinds 01:00:18 commented item C 01:00:41 case sensitive. but it don't mind the <> 01:01:10 chumps are good 01:02:46 DanC_ has quit 01:02:46 AaronSw has quit 01:02:46 danja has quit 01:02:46 deltab has quit 01:02:53 deltab has joined #rdfig 01:03:02 googler has quit 01:03:04 Disconnected from irc.openprojects.net (Connection reset by peer) 01:10:02 logger has joined #rdfig 01:10:02 topic is: Hold on to your hats, it's going to be a bumpy ride. 01:10:02 Users on #rdfig: logger deltab @AaronSw 01:10:07 lasDesk has joined #rdfig 01:10:13 AaronSw: hi 01:10:48 dc_rdfig has joined #rdfig 01:10:53 wmf has joined #rdfig 01:11:03 Hey, the gang's all here. ;-) 01:11:11 http://andyflinn.com/components.htm 01:11:13 D: http://andyflinn.com/components.htm from wmf 01:11:18 tav has joined #rdfig 01:11:33 D:|OpenCOLA Clerver (Mark I) Interface Specification and Reference 01:11:35 titled item D 01:12:47 D: Imagine 01:12:47 that every single computer in the universe had an HTTP url assigned to it. 01:12:47 This means, that we could identify any resource on any computer 01:12:47 (applications, printers, memory, disk etcÉ) by adding a few more legs to 01:12:49 commented item D 01:12:50 tav has quit 01:12:50 wmf has quit 01:12:53 Disconnected from irc.openprojects.net (Connection reset by peer) 01:13:08 logger has joined #rdfig 01:13:08 topic is: Hold on to your hats, it's going to be a bumpy ride. 01:13:08 Users on #rdfig: logger tav 01:13:20 AaronSw has joined #rdfig 01:13:24 Disconnected from irc.openprojects.net (Connection reset by peer) 01:14:02 logger has joined #rdfig 01:14:02 Users on #rdfig: logger deltab_ 01:16:48 deltab_ is now known as deltab 01:18:44 googler has joined #rdfig 01:18:44 AaronSw has joined #rdfig 01:18:44 wmf has joined #rdfig 01:18:44 dc_rdfig has joined #rdfig 01:18:44 lasDesk has joined #rdfig 01:20:12 tav has joined #rdfig 01:30:46 [GlobalNotice] Welcome to the new codebase. You'll notice several differences. Services will be back in a few minutes, and when it returns, channel owners will need to rebuild their access lists and correct the options for their channels. Similarly, all users will need to update any information they would like updated with NickServ. If you have problems, please be patient and take them up with us in #openprojects when it settles down a little bit. T 01:37:44 sbp has joined #rdfig 01:39:14 sbp has quit 01:41:59 danja has joined #rdfig 01:52:08 [GlobalMessage] Please fix your channel access lists. If you have any issues there, please come and talk to us in #openprojects 02:00:55 DanC_ has joined #rdfig 02:13:44 * danja thinks a lot of current SW based on fallacies 02:14:23 Jordan is available for australians 02:16:28 while no-one is listening 02:16:36 I'll give my reasons 02:16:54 number 1 02:17:24 consider the web as a huge library 02:17:48 one wants to find a book 02:18:26 currently, google a query specific 02:18:42 0 responces 02:18:54 vague, 10,000 responses 02:19:32 * danja library analogy not necessary 02:19:46 FOL 02:20:06 how can it be other than the same, 02:20:56 the query has to be logically precise : does A & B log implie C? 02:21:26 if anything more B & W than current search technologies 02:22:26 ok, so my schedule can be combined with that of bwm 02:22:58 but it's naive to think 02:23:18 that Brian would ever want to play me at pool 02:24:26 Web of Trust? 02:24:34 of Folly! 02:25:38 can I rely on even my Bank and Amazon to get everything 100%? 02:25:48 I think not 02:26:40 I do however think 02:26:48 that the SW will develop 02:27:14 but perhaps not as Tim might imagine 02:27:27 * DanC_ wanders by... wonders: "my reasons" ... reasons for what? 02:27:52 mapping the interconnectivity of *precise* machines 02:28:06 to the loose world that humans live in 02:28:36 hi danc - caught me mid-ramble 02:28:59 hi. so I see. don't let me interrupt... 02:29:30 reasons for thinking there are fundamental flaws 02:29:48 * danja sounds like he should be on rdf-logic instead 02:30:48 if you're there, danc, counter these two points for me : 02:32:06 if you use Fol for resolving stuff, your stuff stuff is either T or F 02:32:28 a query will be answered on those terms 02:33:08 so either something will fit exactly 02:33:12 or not at all 02:33:58 ... but I want something that will help me grow healthy cabbages... 02:34:08 yea or nay? 02:34:13 * DanC_ wonders what's to counter, except to quibble about "either T or F"... some things may be unprovable, paradoxical, etc. 02:34:54 so how do you find them using metadata? 02:35:34 point 2 I rambled a bit already 02:35:48 that basically nothing is secure 02:36:16 so A trusts B trusts C is mythology-ville 02:36:29 er... it's possible that documents would be indexed so that a search for "guides to growing vegetables" would match "how to grow healthy cabbages" in a black-and-white way, but I doubt that's the main value of the Semantic Web... 02:36:49 fulltext/heuristic stuff is gonna be ahead of logical classification for years now 02:37:20 * danja thinks the AI guys have been doing it for years 02:37:43 AI guys have been doing it very slowly, in very limited fields of application (drug interactions, etc.) 02:37:50 ok, guides to growing vegetables 02:37:57 at great cost. 02:38:02 I have a site 02:38:38 with exact terminology for how to grow courgettes 02:39:04 but with cwm-style analysis 02:39:24 if someone was looking to grow zucchini 02:39:28 no joy 02:41:14 the mapping between the human and the fol isn't anywhere near enough 02:41:32 to support a fully-functional SW 02:41:36 IMHO 02:41:52 googler has quit 02:41:52 lasDesk has quit 02:41:52 dc_rdfig has quit 02:41:52 wmf has quit 02:41:52 AaronSw has quit 02:41:55 wmf has joined #rdfig 02:43:21 * danja wonders what he was on about in the first place.. 02:43:21 dc_rdfig has joined #rdfig 02:43:26 tav has quit 02:43:37 lasDesk has joined #rdfig 02:43:48 danc - what do expect for the main value of the SW? 02:45:16 I know your quote on the bane of your life 02:45:18 automating drudgery. administrative crud. 02:45:39 yup. fixing the stupid computers... 02:45:46 perhaps I'm a bit cuckoo-land 02:45:55 I don't expect the computer to do magic... I just expect it not to forget stuff I already told it. 02:46:06 agan 02:46:08 again 02:46:38 I have a Palm, that Syncs 02:46:44 googler has joined #rdfig 02:46:50 I have a whiteboard 02:47:04 I try 20-minute manager 02:47:58 the machine allows me to get me english right 02:48:10 so I can can sell BS 02:48:32 tav has joined #rdfig 02:48:40 20 minutes after 20 minutes - all being well I'm back in bed 02:52:58 techies have a very distorted view of what the machine is for - true? 02:53:13 * DanC_ is lost 02:53:42 I read philosophy 02:53:48 I go online 02:54:00 I find goethe.ps 02:54:20 I do informatica 02:54:26 I go online 02:54:38 I find a new Python trick 02:55:26 * danja probably not clear 02:55:49 certainly not clear enough for me to get what the heck you're driving at. 02:56:00 * danja not clear ;-) 02:56:14 point one - 02:56:33 * DanC_ fails to find a flight on Southwest that'll allow me to stay for the whole 1st day of the RDF WG ftf. 02:56:56 A implies B won't help too many people 02:57:02 * danja ouch 02:57:56 suppose what I'm driving at is that 02:58:12 for the information to work 02:58:42 ok, it has to be specified 02:59:16 but a b & w fol search will get you nowhere 02:59:58 needs to be closer to the concept map 03:00:06 folks have in there head 03:00:22 * danja in cloud-cuckoo 03:00:42 their head 03:01:20 such modelling is within reach 03:01:32 e.g. 03:01:52 I'm fairly sure... 03:02:03 wmf has quit 03:03:12 must sleep 03:03:18 good idea 03:03:22 cheers dan 03:03:34 danja has quit 03:34:38 AaronSw has joined #rdfig 03:41:26 AaronSw has changed the topic to: Semantic Web Chat 03:42:08 AaronSw has changed the topic to: Semantic Web Chat: There is no RDF conspiracy. (http://www.w3.org/RDF/Interest#irc) 03:51:41 [globalnotice] In a few minutes I will be shutting down services in order to reenter the time nicks where registered into the database. 04:00:45 AaronSw has quit 04:00:45 tav has quit 04:00:45 dc_rdfig has quit 04:00:50 Disconnected from irc.openprojects.net (Connection reset by peer) 04:02:02 logger has joined #rdfig 04:02:02 Users on #rdfig: logger @AaronSw 04:03:17 AaronSw has changed the topic to: Semantic Web Chat: There is no RDF conspiracy. (http://www.w3.org/RDF/Interest#irc) || New servers: Are we having fun yet? we aren't anywhere close to a stable state yet 04:04:07 dc_rdfig has joined #rdfig 04:04:07 lasDesk has joined #rdfig 04:04:07 DanC_ has joined #rdfig 04:05:01 tav has joined #rdfig 04:07:55 Disconnected from irc.openprojects.net (Connection reset by peer) 04:14:10 logger has joined #rdfig 04:14:10 topic is: Semantic Web Chat: There is no RDF conspiracy. (http://www.w3.org/RDF/Interest#irc) 04:14:10 Users on #rdfig: logger tav googler dc_rdfig AaronSw lasDesk 04:14:36 deltab has joined #rdfig 04:14:49 hmz 04:14:53 hi dan 04:15:01 Are we having fun yet? 04:15:05 dc_rdfig has quit 04:15:27 give it to me AaronSw baby! 04:15:49 :-) 04:15:57 chumpy has gone! 04:16:01 * tav posts urls with glee 04:16:13 heh 04:16:19 * AaronSw notes that googler works fine with chumpy 04:16:25 googler, chumpy? 04:16:28 chumpy: http://www.viewaskew.com/newboard/messages275/1680.html 04:16:29 googler, hello? 04:16:30 hello: http://www.sanrio.com/ 04:16:37 googler, espra? 04:16:38 espra: http://espra.net/ 04:16:44 nice 04:17:01 do we really need a whole bot for that? 04:17:09 Hmm, only half the network went down this time... 04:17:21 whole bot: i suppose no, but it's cute. ;-) 04:19:09 hiroko has joined #rdfig 04:19:29 g chumpy 04:19:33 http://www.viewaskew.com/newboard/messages275/1680.html - Yo, Chumpy... 04:19:33 http://pub52.ezboard.com/fchumpyboyscommcentrefrm1 - Public Forum 04:19:33 For more results go to: http://www.google.com/search?q=chumpy 04:20:29 slightly more verbose... but the same bot can do multiple tasks ;p 04:23:13 tdxdave has joined #rdfig 04:23:59 * AaronSw really needs to learn Zope sometime... 04:24:46 DanC_ has joined #rdfig 04:25:57 zope is going to be one of the most useful tools for us semantic webbers 04:26:25 so, you, oh leading one, should most definitely spare one evening sometime to do so 04:27:06 * DanC_ wonders what he missed. 04:27:21 thanks, tav 04:27:34 i definitely hope to 04:27:40 yeah, zope looks pretty cool; I'm using it for my personal web site (http://dm93.org/). I have to say I'm annoyed by some basic things Zope messes up. 04:28:04 Zope adds a element to every HTML document it serves up. Gaak! 04:28:06 Yeah, I've heard some of that from Wesley 04:28:22 plus, it doesn't seem to send last-modified nor grok if-modified-since. 04:28:54 DanC_: you can add headers... 04:29:13 I've seen documentation about advanced caching stuff, but if it doesn't do if-modified-since out-of-the-box, I have little confidence. 04:29:32 04:29:46 that's last-modified 04:29:56 where would I stick that setHeader thingy? why isn't it on by default? 04:30:10 you can extend the accelerated http cache manager to use if-modified-since 04:30:26 extend? why isn't it the default? 04:30:38 hmz 04:31:02 because zope is about being a tool that you can use to do what you want 04:31:24 tdxdave has quit 04:31:32 That seems like a pretty lame excuse... this is just a basic HTTP feature. 04:31:32 e.g. most of my documents are dynamically created 04:31:34 re the setHeader thingy: if you're suggesting sticking that in my documents, I'm not using DTML for my documents. I'm just using plain XHTML, editing it with HTTP 1.1 (in particular, Amaya). 04:32:06 dynamically created: that makes more sense... 04:32:22 * DanC_ is tempted to go on a jihad about good defaults. customizability is only a burden if the defaults aren't good. 04:32:24 DanC_: you just have a bunch of xhtml docs in a folder? 04:32:44 yup. a bunch of XHTML docs, jpeg images... you know: a family web site. 04:32:58 DanC_: we (the zope community) always welcome suggestions, patches, etc ;p 04:33:10 hmz, do you use a header / footer file? 04:33:38 or some methods to provide 'templates' for your site? 04:33:52 er... I don't seem to be making my point. if-modified-since is a very, very basic feature. If the Zope folks don't support it out-of-the-box, that gives me little confidence that they got other things right. 04:34:12 no, no header/footer. no templates. just GET/PUT. 04:34:34 (I use Zope rather than apache because apache doesn't grok PUT out-of-the-box) 04:35:14 (and because Zope's authentication is more flexible and dynamic. and WEBDAV support, and a few other things. but not DTML) 04:36:10 I've been playing a bit with Zope / Mac OS X WebDAV -- it's awesome! 04:36:22 the munging was shocking too. Data integrity is job one. If I PUT some bytes in a file, I expect to GET the same bytes back unless I say otherwise. 04:36:58 Zope/WEBDAV is quite nice... it allows my wife to publish images using Windows WebFolders. 04:37:32 OS X makes it appear like just another hard drive -- very cool. Can edit it with all my local tools. 04:38:08 yes. ah... life as it should be. 04:38:26 exactly. 04:38:28 have you checked to see if Zope is munging your HTML docs with ? 04:38:54 hmz 04:39:08 Only on server output -- not the actual docs, I think. 04:39:08 * DanC_ hacked out the munging; really should send that patch. 04:39:56 er... what's the difference between "server output" and "actual docs"??? OS/X gets the data via HTTP, no? 04:40:30 Hmm, that's true... I suppose I didn't notice. 04:41:06 hmz, never understood the problem ;p 04:41:54 do you agree it's a problem if I PUT something in an HTML document, but what I GET back is not exactly, byte-for-byte, the same thing? 04:42:22 it can be quite shocking, yes 04:42:42 or at least: element-for-element; I could perhaps tolerate a little whitespace munging. But adding would screw up digital signatures. It screws up editing with Amaya (though perhaps it should not) 04:42:44 DanC, do you agree that it's acceptable in cases where the user requests that functionality? i.e. please tidy my pages for me? 04:42:44 especially if you were unaware of it 04:43:32 user request: sure, it's fine to do stuff with the knowledge and consent of the user. 04:43:52 unaware: exactly. 04:44:26 * DanC_ goes to find the code... 04:46:32 there it is... 04:46:34 def insertBase(self, 04:46:44 I just added 'return' as the first statement of that method. 04:46:52 # DWC: why do this? 04:46:52 return 04:47:00 in ZPublisher/HTTPResponse.py 04:48:44 * DanC_ wanders off... 04:48:54 heh 04:49:10 the why is for dynamic webpages 04:49:24 Why do dynamic webpages need it? 04:51:22 hmz 04:51:42 not necessarily for dynamic webpages, but usage is dynamic 04:52:54 usage is dynamic? i'm not following... 04:52:54 you know how index_html works? 04:53:04 no, sorry - i haven't played with zope much 04:53:09 Disconnected from irc.openprojects.net (Connection reset by peer) 04:58:02 logger has joined #rdfig 04:58:02 topic is: Semantic Web Chat: There is no RDF conspiracy. (http://www.w3.org/RDF/Interest#irc) 04:58:02 Users on #rdfig: logger tav AaronSw deltab 04:58:06 zope has this concept of objects... 04:58:13 everything is an object 04:58:23 But i do sorta get the ideas of inheritance and acqusition and that stuff 04:58:27 some objects can contain other objects 04:58:29 ah good 04:59:09 well, w3.org/folder would render w3.org/folder/index_html 04:59:23 yeah, I've never understood that -- why? 04:59:49 because folder is a simple container 04:59:53 it has no content 05:00:17 Right, but can't it know enough to access it's index property or whatever 05:00:19 and when called, it renders the index_html method 05:00:31 ok, that makes more sense 05:00:35 that's basically what it does 05:01:11 i have a product that both acquires from objectManager (is folderish) and has content 05:01:39 but, that can _really_ confuse people 05:01:47 people like having folders and objects within them 05:01:51 ;-) 05:02:01 anyways, we digress... 05:02:19 let's say that index_html has a relative link to 'foo_html' 05:02:29 and you had accessed w3.org/folder 05:02:45 browsers, i am told, create relative links from the last trailing slash 05:02:51 aha! I think I see, 05:02:57 so, it would link to w3.org/foo_html 05:03:06 the addition is a hack because Zope doesn't support folders properly. 05:03:06 instead of w3.org/folder/foo_html 05:03:13 it should redirect when it hits folders. 05:03:17 yea 05:03:25 this is what every other web server does 05:03:41 Plus it is bad since there is no canonical URL, there are many duplicate ones... 05:03:41 but... zope makes no distinction 05:03:53 a folder is merely an object 05:04:07 Bad Zope... why does no one fix this? 05:04:19 what's to fix? 05:04:29 Add a redirect to folders. 05:04:39 Make it so they can't be accessed without the trailing slash. 05:05:17 well, it's not just the base tag... 05:05:52 some products need it for their renderings 05:06:04 so, if we go forcing a redirect, they will get screwed ;p 05:06:16 But I thought you said there was no distinction between the two... 05:06:44 but... zope makes no distinction 05:06:44 a folder is merely an object 05:06:58 yup... 05:07:48 hmz 05:08:02 so you want someone to modify the folder object to redirect 05:08:10 yes. 05:09:06 i guess that can be done, just for the default 'folder' object 05:09:42 and let product developers add something similar to for their own folderish products if need be 05:10:36 e.g. it would be a _real_ headache for my product ;p 05:11:02 cos a link to 'bar' on site.com/foo where bar is an object in / 05:11:13 would now go to site.com/foo/bar 05:11:40 hmm 05:11:46 which _would_ work with acquisition, unless i had another object called bar in the foo folderish object 05:12:24 in which case, not having screws things up for me 05:12:42 Well, you can always use ../foo 05:12:48 That would be the correct way to do it. ;-) 05:13:36 what happens to relative links that are at the root of a site? 05:14:14 hmm, they stay at the root, i suppose 05:14:24 * AaronSw asks python 05:15:12 yeah, it stays at the root 05:15:44 python? 05:16:05 >>> import urlparse 05:16:08 heh 05:16:34 Python is your friend. ;-) 05:19:19 hiroko has joined #rdfig 05:19:31 exec import urlparse 05:20:09 exec urlparse.urljoin('../../foo','http://www.foo.org/') 05:20:11 'http://www.foo.org/' 05:20:21 ;p 05:20:33 That's pretty cool. 05:20:39 very 05:20:47 one of the other espians did it 05:20:55 Although >>> is easier to type than exec ;-) 05:21:05 heh 05:22:57 Disconnected from irc.openprojects.net (Connection reset by peer) 05:23:03 logger has joined #rdfig 05:23:03 topic is: Semantic Web Chat: There is no RDF conspiracy. (http://www.w3.org/RDF/Interest#irc) 05:23:03 Users on #rdfig: logger googler lasDesk DanC_ 05:23:05 AaronSw has joined #rdfig 05:23:33 It has got to stop doing that... 05:23:39 Hmm, well for DanC, we discovered why Zope has that silly hack: 05:23:45 It doesn't do redirects on folders 05:23:51 so relative links won't work without a base tag. 05:23:59 tav tells me that the issue is fixable. ;-) 05:26:05 deltab has joined #rdfig 05:26:40 tav has joined #rdfig 05:27:16 go OPN! 05:28:41 heh heh 05:32:29 it'd be better if they just said, we are closing down for an hour 05:35:10 that wouldn't be as fun 05:38:35 btw, has someone created a summary of the FoRK list? 05:38:49 i'd love to read throught those thousands of emails, but i just don't have the time 05:39:07 so, if someone has gone through and created monthly summaries, that would be very welcome 05:43:49 don't count on it! 05:43:51 ;-) 05:44:11 it's actually easier than it seems: just delete the emails on politics/children/etc. 05:44:13 ;-) 05:44:17 lol 05:44:28 Disconnected from irc.openprojects.net (Connection reset by peer) 05:45:02 logger has joined #rdfig 05:45:02 topic is: Semantic Web Chat: There is no RDF conspiracy. (http://www.w3.org/RDF/Interest#irc) 05:45:02 Users on #rdfig: logger tav googler lasDesk DanC_ AaronSw deltab 05:45:05 woo! 05:45:13 gah 05:45:17 damned you be logger! 05:45:47 hmz 05:45:55 how large is your inbox AaronSw? 05:46:07 rather, how large is your mailing list mail? 05:46:51 i'm looking for some large source of email from mailing lists to put into this zope application i just finished 05:46:53 umm, do you really want to know? ;-) 05:46:57 yea 05:47:11 hmm... 800 unread as of yet 05:47:13 tell me in terms of number of email as opposed to size 05:47:41 800 messages, although it usually gets up to a couple thousand before I get a chance to read them 05:49:11 heh 05:49:29 hope to make some dent on the plane tomorrow 05:49:31 and total mail? 05:49:43 actually, i've been pretty good about my inbox 05:49:55 i got it down to 60 messages after some intense focus 05:50:51 i think i have a record of unread mail.... 81,124 in the inbox of tavlists (there are a few thousand spread across the other folders which have been filtered into) 05:51:03 oooh, plane, where you flying? 05:51:15 wow, that is a record! ;-) 05:51:35 plane: to SWWS at Stanford 05:51:47 Semantic Web Working Symposium 05:51:53 hmz 05:51:59 sounds like somewhere i would like to be 05:52:07 urls? 05:52:13 and with that i need to get some sleep ;-) 05:52:19 see http://www.semanticweb.org/SWWS/ 05:52:35 nice 05:52:45 well, say hello to people for me 05:52:49 will do ;-) 05:53:13 and i bid you good nite 05:54:29 good nite! 05:54:43 been great chatting with you -- hope to talk again sometime. 06:17:05 dc_rdfig has joined #rdfig 06:26:43 likewise 06:55:58 kumo has joined #rdfig 07:15:27 Disconnected from irc.openprojects.net (Connection reset by peer) 07:16:02 logger has joined #rdfig 07:16:02 topic is: Semantic Web Chat: There is no RDF conspiracy. (http://www.w3.org/RDF/Interest#irc) 07:16:02 Users on #rdfig: logger kumo googler lasDesk DanC_ AaronSw 07:16:27 dc_rdfig has joined #rdfig 07:23:39 kumo has quit 07:55:31 Disconnected from irc.openprojects.net (Connection reset by peer) 07:58:04 logger has joined #rdfig 07:58:04 topic is: Semantic Web Chat: There is no RDF conspiracy. (http://www.w3.org/RDF/Interest#irc) 07:58:04 Users on #rdfig: logger dc_rdfig 07:58:46 lasDesk has joined #rdfig 09:38:45 danja has joined #rdfig 09:46:30 danja has quit 11:15:06 deltab has joined #rdfig 12:19:33 deltab has quit 12:54:19 Disconnected from irc.openprojects.net (Connection reset by peer) 12:55:06 logger has joined #rdfig 12:55:06 Users on #rdfig: logger lasDesk deltab 12:55:06 This channel is logged and blogged: http://logicerror.com/rdfIRCWelcome 12:55:20 dc_rdfig has joined #rdfig 13:02:15 timbl has joined #rdfig 13:11:41 googler has joined #rdfig 13:12:57 tav has joined #rdfig 13:21:31 DanCon has joined #rdfig 14:04:01 barstow has joined #rdfig 14:05:20 shellac has joined #rdfig 14:20:42 shellac has quit 14:34:04 grrrrr has joined #rdfig 14:34:28 grrrrr has quit 14:51:07 rrs has joined #rdfig 15:06:47 tdxdave has joined #rdfig 15:25:46 shellac has joined #rdfig 15:39:36 shellac has quit 15:44:30 sandro has joined #rdfig 15:46:57 sandro has quit 16:21:29 bwm has joined #rdfig 16:32:08 danbri has joined #rdfig 16:32:21 hello dan 16:32:24 * danbri waves from http://www.semanticweb.org/SWWS/ 16:32:27 hi brian! 16:32:29 JHendler has joined #rdfig 16:32:43 I was going to scribe into the channel a bit; unless someone doing that already... 16:32:44 JHendler is now known as JimH-SWWS 16:32:51 danbri is now known as danbri-SWWS 16:32:56 Sorry about last night - got bruise on my forehead from head hitting table as I fell aslepp 16:33:02 s/aslepp/asleep/ 16:33:20 sounds like a good idea 16:33:33 Sorry we were a bit late. Was worth it though to hear the hotel bar staff pronounce "knackered" 16:33:59 * danbri-SWWS will take occasional notes from SWWS talks 16:34:12 swws: Eric Miller giving 1hr talk on SW 16:34:29 em: describes structure of SW Activity 16:35:06 em: SW structure; SW Coord Group members raise hands 16:35:24 em: ...SW part of W3C Technology and Society Domain 16:35:38 em: ... not just tech, but societal issues. 16:35:58 em: ...key technology is RDF 16:36:13 em: RDF core members (raise hand) 16:36:28 em: makes danbri and bwm raise hands 16:36:39 em doesn't realise he's being transcribed into IRC in real time :) 16:36:57 em talks about new WG (WebOnt) 16:37:15 em: W3C Advanced Development, see http://www.w3.org/2000/01/sw/ 16:37:36 em: Education and Outreach... haven't done a lot yet, currently discussing goals and focus. 16:38:01 em: ...will bring together folk to clarify mission etc of SW for public audience 16:38:22 em: ...will clarify what it is we're tryign to do. Will never be one single view of SW but... 16:38:34 em: W3C Approach to Deployment 16:38:51 em: ...trying to graft this into the existing Web. Need to gracefully evolve from what we have now. 16:39:31 em: Focus on short-term deployment. W3C not a research organisation; need immediate wins. 500+ W3C members want things now! 16:39:53 em: short term focus but with an eye towards longterm research issues (hence this group) 16:40:25 em: ...w3c SW activity realises that the SW won't be "just the W3C" 16:40:36 em: so establishing relationships with other organisations etc is important 16:40:58 em: people are working around the clock to make this stuff happen; "a lot of hard work, and a bit of luck" 16:41:14 em: Semantic Web Principles... 16:41:41 em: some of these controversial(ish), some guidelines... basic design issues that we try to weave through the various W3C activities. 16:42:07 em: ...from the SW side, everthing is "viewed to be on the Web"; things in the physical world will have an online representation 16:42:55 em: "...partial information". History of Web 404 error, that the 404 error was the thing that made the (original) Web work. 16:43:29 em: ...by relaxing the global link consistency contstaint, the Web broke from hypertext tradition. And took off... 16:43:55 em: from SW side, fact that people can say anything about anything, then who is saying becomes fundamental 16:44:29 em: ...you'll never know everything that's being said. Recognising up front this is a SW founding principle. 16:44:59 em: ...web of trust issues... evolution... minimalist design -- avoiding over standardisation 16:45:37 em: ...a balancing act. standardising "just enough", focussing on threshold of minimal agreement 16:45:56 em: commmon models facilitate aggregation of networked infromation 16:45:59 em: RDF 16:46:09 em: the focus of common models within W3C 16:46:23 em: RDF is... 16:46:29 em: Resource, things that can be named 16:46:43 * barstow thanks danbri for his notes! 16:46:48 em: Description: statements about the properties of these resources 16:48:13 bwm has quit 16:48:13 danbri-SWWS has quit 16:48:16 danbri-SWWS has joined #rdfig 16:48:19 em: people liked RDF, but they weren't so keen on some aspects of the spec. RDFCore WG aims to tighten up the M+S spec. RDF Core WG also tasked to finishing RDF Schema. 16:48:20 bwm_ has joined #rdfig 16:48:34 * danbri-SWWS notices himself and brian disconnected at same time (?) 16:48:56 em: RDFCore also must build on the work of the Cambridge Communique 16:49:03 (url?) 16:49:21 em: ...integrating XML Schema datatypes 16:49:29 em: makes danbri and brian wave again 16:49:40 Cambridge Communique: http://www.w3.org/TR/1999/NOTE-schema-arch-19991007 16:49:45 em: RDF Core slightly different from a lot of the other W3C WGs 16:49:55 em: taken a very active and open interest in public feedback and discussion 16:50:07 em: RDF Core's decisions are made public; reflected to the Interest Group 16:50:19 em: ...have liaisons and relationships to the larger body of implementors 16:50:32 em: if we make decisions and they say "ow, that hurts", we want to knopw 16:50:42 * danbri-SWWS thinks of rss-dev and namespace-prefix issue 16:50:50 * JimH-SWWS someone should chump pointer to Eric's slides - Icanot read from here 16:51:16 em: want to make sure we're as clear and concise as possible; actively taking input from www-rdf-interest, www-rdf-logic, DAML+OIL 16:51:31 ...these guys providing hugely useful design input and feedback 16:51:36 Can we get a summary of the RSS issue fed back into RDFCore mailing list 16:51:44 em: shows TimBL "standards cake" diagram 16:52:21 RSS issue: yes, i'm trying. basically it breaks XML-based RSS 1.0 apps that don't expect rdf: prefixes on the container resource= 16:52:31 ...a bug in RSS 1.0 spec, really. 16:52:53 em: talks through the hierachy of specs/techs from timbl's ladder leading to logic/proof/trust 16:53:13 em: another diagram 16:53:38 * danbri-SWWS can't read the url for eric's slides; latest version is on his laptop, not checked in 16:53:58 em: talks about lego (legos, as americans call it) 16:54:04 * JimH-SWWS wonders why trust is on top, instead of along the side as something increasing as we go up the layer cake... 16:54:16 * danbri-SWWS never really liked that slide 16:54:28 * JimH-SWWS UK doesn't put an extra "u" in Legou? 16:54:44 em: RDF provides useful basis, RDF Schema plugs in and offers more functionality; DAML+OIL more still... 16:54:52 lego: americans pluralise 16:55:06 em: presents 2 more lego components, rdf rules, rdf query 16:55:37 em: rdf-rules, rdf-query, connections between different traditions, eg. results from queries input into rules. database tradition meets kr trad 16:55:42 :) 16:56:03 em: SW -- Building on proven ideas 16:56:33 Trust is on top because it requires proof which is quite far up 16:56:43 sandro has joined #rdfig 16:57:10 em: may be natural selection amongst ontologies... 16:57:19 em: Semantic Web Advanced Development 16:57:25 em: (my favourite area...) 16:57:40 em: Advanced Dev: designed to explore pre-competitive prototype ideas 16:57:48 em: venue for liaison w/ research community 16:58:07 timbl - I understand that idea, but just as proofs can increase in complexity (i.e. we can do simple proofs in RDF, better in DAML+OIL, better in eventual WebRules, etc. -- thus I see trust increasing as we move up, as opposed to a top layer - IMHO 16:58:15 em: prototypes feed back into standards; working code feeds back into standards 16:58:38 em: collaborative development environment: W3C team, W3C Members, public... explore ideas together 16:58:41 [Ralph ponders the impact of having the original author of a slide on-line in real-time while someone else presents it] 16:59:07 ... presents it 3000 miles and 3 timezones away ;) ] 16:59:24 em: (presents areas of interest for SWAD, ie. the

s from SWAD home page) 16:59:46 em: tools underway at W3C include... RDF triple store, Algae, cwm, Blindfold, Annotea 17:00:02 * danbri-SWWS wishes he'd done something complete enough to live on that list :( 17:00:18 em: (makes ericp wave) 17:00:44 * danbri-SWWS notes a resemblance to danbri's German digital libraries forum presentation... 17:00:48 * JimH-SWWS wished some of the DAML aps would start to make that list - lots of growing tools/appls.., 17:00:54 we should have a database of slides :) 17:01:38 (ideally we could auto-generate slides from the w3c swad database of what we did!) 17:02:07 em: cwm provides RDF in XML, RDF in N3 17:02:22 * danbri-SWWS notes others are saying stuff in [off] that he's responding to in logged msgs 17:02:43 actually, we are working on generating sldies from DAML/RDF -- maybe an amaya plug-in (or however one adds to amaya) would be a winner. 17:02:53 em: RDF/DAML capable aspects of cwm. In mem database. Simple horn rules. Rules can generate rules, queries can generate queries. 17:03:00 I hope EM uses the notation "RDF/XML" and "RDF/N3" -- people should get used to not thinking of RDF as [only one] syntax 17:03:27 Jim: that would be very cool! My german presentation nicked stuff off older presentations of erics and ralphs; and we all nick timbl's diagrams... 17:03:35 s/Jim:/jim,/ 17:03:44 slide he is showing now says RDF/XML, RDF/N3,XML ... 17:04:03 em: describes Sandros work with Blindfold, XSB etc 17:04:04 em: Annotea 17:04:19 em: Ralph, Jose, Marja not here; Eric wave 17:04:21 (eric waves) 17:04:25 em asks Ralph Swick to raise his hand... 17:04:40 em: Amaya has annotations support compiled into it. Latest version is quite excellent. 17:05:04 em: this is generic enough that folk working on plugin and other suport in additional browsers, eg annozilla in mozilla 17:05:26 em: when other people visit a page, browser can query an rdf database for relevant annotations 17:05:46 em: ...query capability based on Algae; realtes to web-of-trust, collaborative filtering etc too 17:05:58 em: will highlight a couple of rdfig task forces 17:06:06 * rrs raises both hands 17:06:13 em: RDF Calendar task force 17:06:22 em: libby miller pls wave 17:06:45 em: RDF calendar, "an informal grouping of individuals interested in calendaring and scheduling, primarily in rdf" 17:06:57 Generating slides from RDF - definitely interetsing espe for PSUM reports. 17:07:19 em: ...this topic a huge problem in groups like the Web Consortium, eg arranging meetings; eg finding 3 people on W3C management team and getting them on phone at same time 17:07:28 * rrs hears Tim's slap, goes back to work :( 17:07:33 * timbl wonders how many realannotation servers are running 17:07:50 em: ..."everyone has different tools to manage their time; rdf-calendar trying to merge data across these 17:08:17 * danbri-SWWS wonders if timbl is writing up designs of cwm for PSUM reports? 17:08:20 * timbl had not meant slap at all, proposed great distraction ... probably not such a goof idea though 17:08:36 em: (talks about rdf calendar demo) 17:08:56 stefan: there is a prize of a bottle of champagne for 1st person to import this data into Outlook or My Yahoo 17:09:04 (danbri, PSUM reports are two-pagers, summaries only. No internal doicumentation) 17:09:12 * JimH-SWWS suggests that would be good for next year, not this one :-) [Psum] 17:09:13 see http://www.semanticweb.org/SWWS/ for champagne challenge! 17:09:20 * timbl goes to get a sandwich for lunch 17:09:28 timbl is now known as tim-brb 17:09:35 * rrs knows Tim didn't mean to rap knuckles ;) 17:09:43 em: libhy and michael's 'hybrid rdf calendar' schema providing experimental basis for this 17:10:09 * sandro spent numerous, numerous hours hacking SW-ish perl accounting code for IRS filings (as a distractions, perhaps) in years past. 17:10:11 em: DanC and I playing with data in this format in cwm. Only when we merged all this data did we realise he wouldn't be here in time for this talk ;-) 17:10:17 * rrs also knows that logger doesn't ignore /me comments 17:10:33 * danbri-SWWS would prefer logger to log everything 17:10:45 (public is public...) 17:11:15 * sandro would like this log, as the congressional one, to have a "revise and extend" system. 17:11:16 em: SW tools used for coodinating a pool game with DanC 17:11:28 I don't mind making public remarks in public, I just want to be able to make in-line comments that don't distract from the persistent record 17:11:45 ... persistent record of the primary thread, that is 17:11:58 em: our apps didn't have unique ids for events; had to merge the data based on properties 17:12:14 em: we experimented with some stuff from the Cyc upper ontology 17:12:18 revise and extend: that's what Annotea is all about ! 17:12:49 em: Another highlight: W3C Workflow 17:12:56 Annotea -- absolutely :) 17:13:15 Does Annotea have an eye towards giving readers a post-revision view of things? 17:13:58 em: example: we have a spec reaching a certain status, we need 3 of for people (in certain roles; eg. chair) to have signed off on a document or workflow event... 17:13:59 Re:threading -- that's what a threading interface is for (instead of this.... irc thing.) 17:14:11 em: ...state transistion eg. from Working Draft to Last Call WD 17:14:38 em: ...we currently do this in a rather ad-hoc manner. We could do better... 17:14:52 jim, do you have a url for the daml/ppt stuff? 17:15:11 em: ... the W3C Team want automation of the W3C Tech Reports page, ie http://www.w3.org/TR/ 17:15:34 em: ...right now chair gets notified, comms team get notified, web team then change site. We want to automatie this. 17:16:06 em: what happens in our scenario: the chair asserts a (structured) msg to a special mailing list, asserting a doc reaching a certain status. 17:16:39 em: ...eg doc goes to status with a given editor; fullname, mailbox etc 17:16:56 em: "for all those who don't read rdf..." (showing some tanglebrackets on screen) 17:17:12 * danbri-SWWS wonders what % of the audience can read RDF without machine assistance 17:17:39 * rrs wonders if tanglebrackets was an intentional spelling -- finds it very attractive 17:18:15 tanglebrackets: word just popped into my head :) 17:18:29 :) 17:18:36 em: shows workflow example (document announce) 17:18:38 we heard it here first, folk 17:19:06 credit: tanglebracket - DanBri 2001-07-30T17:15Z 17:19:13 http://mr.teknowledge.com/daml/ 17:19:13 E: http://mr.teknowledge.com/daml/ from JimH-SWWS 17:19:25 E:| DAML meets (hate to say it) Powerpoint 17:19:26 titled item E 17:19:26 * danbri-SWWS bows gracefully, files trademark app 17:19:47 E: Bob Balzer project links DAML to Powerpoint 17:19:48 commented item E 17:19:57 em: (talks about document metadata meets workflow metadata) 17:20:06 E: would be interesting to see if any of his work could be coupled with a better (open) format slide producer 17:20:07 commented item E 17:20:24 em: (shows purty autogen'd picture, with colorized edge-label graph of data from mixed sources; WG, docs, etc) 17:21:04 SWWS-attendees, anybody volunteer to scribe from other sessions into #rdfig? 17:21:16 * danbri-SWWS wonders what to do for parallel sessions 17:21:24 em: these visual graphs done in RDF 17:21:39 em: we didn't agree unique IDs up front, we did a graph merge based on uniquely identifying properties 17:21:59 "tanglebrackets" for people who thought it couldn't get any worse than Lots of Irritating Silly Parentheses. 17:22:02 * JimH-SWWS is impressed by how SWAD is using uniqueproperty and equivalentTo - nice to see DAML making it into the world... 17:22:31 * danbri-SWWS has a Strong Opinion on daml:UniqueProperty 17:22:58 ...it doesn't quite do what I want for http://xmlns.com/foaf/0.1/ -- it can't make guarantees about _uniqueness over time_. 17:23:10 (something for RDF Schema 3.0 ;-) 17:23:40 em: this technology (workflow modelling) set us up to answer lots of different questions 17:23:41 * JimH-SWWS is nervous to ask... 17:23:58 em: we can very easily repurpose the app by adding a few more rules 17:24:31 ask? RDFS 3.0? (just teasing!) or daml:UniqueProperty? (serious and interesting issue imho) 17:24:40 em: what's hot? what's hard? 17:24:49 em: small rdf files define new apps; 17:25:00 em: vocab mapping between these things, for small apps, is easier than we thought 17:25:15 em: from library community context, mapping seemed much harder 17:25:20 em: What's Hard? 17:25:26 em: teaching others to fish... 17:25:47 em: can almost hand over the code to allow others to re-use these swad tools 17:25:57 * danbri-SWWS laptop battery fading; looks for power points... 17:26:08 em: Open Issues 17:26:18 oop 17:26:46 em: what's hard: social grounding of the tech/data structures; making mass understandable; recovering semantics from legacy apps 17:27:00 JimH-SWWS has changed the topic to: DanBri's running commentary from SWWS 17:27:16 (feel free to co-scribe if i miss stuff) 17:27:35 you're doing great! 17:27:36 tim-brb is now known as tim 17:27:39 em: Open Issues> Are agreeing to URIs enough? Is there some basre ontology that's required 17:27:51 * tim keeps forgetting to reset /nick 17:28:01 em: daml:UniqueProperty a useful middleground 17:28:19 JimH-SWWS has changed the topic to: DanBri's (and others) running commentary from SWWS 17:28:21 em: scaling an issue. Small w3c workflow datasets we can do our datamerge; for 40 trillion nodes, different matter 17:28:31 em: Where are we going? 17:28:37 em: Lots of work still to do... 17:28:45 em: relationship to other groups, eg topicmaps 17:28:53 em:... clearer ties to xml family of specs 17:29:02 em: huge marketplaces understand 'syntax' to mean XML 17:29:02 sbp has joined #rdfig 17:29:15 em: we need to integrate these concepts with the larger industry picture of XML 17:29:21 (Re time, danbri, I think you have to bite off a large time management chunk the moment you start. Others have done this, you can too, but you can't just add a new form of "unique over time" and hope to have solved it. IMHO) 17:29:24 em: better udnerstanding of Semantic Web and Web Services 17:29:35 em: two communities attaching similar problems in fundamentally differnt ways 17:29:41 (did he say *different*???) 17:29:53 em: EU SW initiative, call looming 17:30:07 em: W3C plan to support,guide and submit a bit 17:30:13 em: WebOnt WG looming 17:30:28 * danbri-SWWS remembers that he owes Jim his review of WebOnt charter 17:30:58 em: upcoming: query, rules etc 17:31:17 em: We need to clarify mission/goals of W3C SW Activity, structure of taskforces, Interest Group(s) etc 17:31:26 em: want more taskforces... 17:31:36 em: ...better ingtegration with collaborative work environments 17:31:48 em: eg. opensource .NET / gnu.net folk 17:32:02 em: moving ahead w/ publications, outreach etc important too 17:32:06 em: More Information: 17:32:15 em: RDF Interest Group / SW Interest Group 17:32:23 em: Encourages for to join the #rdfig channel 17:32:45 em: "this is still pretty early in the game; this won't work without public input 17:32:58 "there are conduits forpublic to get involved. Please do! 17:33:03 em: "Thank you" 17:33:08 (mucho clapping) 17:33:11 --- 17:33:23 * danbri-SWWS seeks electrons for laptop 17:34:47 Cool talk. Slides online yet/soon? 17:34:48 JimH-SWWS has quit 17:35:24 soon! 17:36:22 Great! 17:36:35 Ugh, such a shame I'm missing this in person 17:36:43 bwm_ has quit 17:36:44 But at least I get the running commentary/logs :-) 17:39:11 * danbri-SWWS and william l. chat about talkin signs :) 17:39:17 "persistent record of the primary thread" is a really good idea: could we have a channel just for the talks, and then keep #rdfig for incedental discussion? 17:39:25 Say hi to WL for me 17:39:36 Er... and anyone else that's there! 17:39:38 he reading this 17:39:46 Oh; hi William! 17:44:37 If we prefixed everything with A:, B:, then chump would separate the threads. (hack hack) 17:45:33 Yeah, but then you'd end up with people forgetting, or putting on the wrong prefix 17:47:01 [separate channels for primary thread and for commentary about primary thread make it hard to do the interconnects] 17:47:47 I agree 17:48:21 I'm sure we could come up with something, but it's hard to funnel it through the IRC command line. 17:48:56 nick == xx-trackid 17:49:20 * sbp looks at the SWWS program list 17:49:33 eg danbri-t3 17:49:59 what workshops are y'all heading off to? 17:51:17 danbri-SWWS has quit 17:52:52 Seems like Dan couldn't find those electrons... 17:57:08 sbp has quit 18:10:42 danbri-SWWS has joined #rdfig 18:10:55 * danbri-SWWS reappears from WebServices talk 18:11:01 channel free for scribing? 18:11:11 WebServices: DAML-S presentation 18:11:17 missed 5 mins 18:11:26 quotes indicate speaker 18:11:33 "upper ontology of services 18:11:44 "imagine some resource, providing some services 18:12:17 "3 perspectives; the service profile it presents (what it does); the Service Grounding it suports; ..." 18:12:25 "starting w/ service profile 18:12:46 "advertises through a yellowpages like system; 18:12:54 "represents wht it does 18:13:12 * rrs suggests DanBri change his nick to 'danbri-t3'; i.e. xx-trackid 18:13:20 "3 main components of service profile: human readable desc; spec of functionalities service provides; functional attributes 18:13:26 "'jellyfish' diagram 18:13:43 "looks at provonance of description of some service 18:14:00 "details about 3rd party actors involved in the service offered 18:14:03 dlm has joined #rdfig 18:14:21 "functionality description: defines preconditions, inputs, effects of outputs 18:14:47 "characterisations of geographic scope (don't want pizza in wrong city!) 18:15:01 "quality descriptions and guarnatees 18:15:23 "linking into ontologies for service types, categories; commercial problem solving etc 18:15:27 "service params 18:15:30 -- 18:15:35 "Service Model 18:15:47 "flow of info, process; example amazon.com 18:15:54 "you need that description of the workflow 18:16:08 "facilities automatic web service composition/syndication 18:16:18 "you can compose high-level services (into sub-tasks) 18:16:25 "eg. what you need to do to buy a book 18:16:41 "monitoring: how far has this service progressed during its invocation 18:16:47 "service model may be used to 18:17:18 "analyse whether service meets need; compose service descriptions from multiple services to perform a task; during course of service enactment ... 18:17:41 (slides changing fast and speaker talks fast; please ACK if my notes are in any way useful!) 18:17:50 ""how does it work?"" 18:18:01 "think of a service as either a simple process or composite process 18:18:20 "associated with each service is a set of inputs, of outputs, preconditions and effects 18:18:25 "function and action metaphor 18:18:45 "composite processes are compositions of simple or other composite processes in terms of constructs such as 18:18:52 "sequence, if-then-else; fork 18:18:58 logger, help? 18:19:01 I'm logging. I found 308 answers for 'help' (showing 0...5) 18:19:02 0) 2001-07-30 18:18:58 logger, help? 18:19:03 1) 2001-07-30 02:56:56 A implies B won't help too many people 18:19:04 2) 2001-07-30 02:33:58 ... but I want something that will help me grow healthy cabbages... 18:19:05 3) 2001-07-29 00:41:32 no, but if we don't get better consistency and quality in small scale rdf applications, scaling it up wont' help. 18:19:06 4) 2001-07-29 00:35:27 i wrote the query library to learn what rdf was useful for, expecting that querying the samples would give me a good intuition. i can't really say it helped me, because i wasn't able to get particularly brilliant insights by querying rdf that i coudl find. 18:19:07 5) 2001-07-29 00:32:33 how do you help users know which mona lisa they are talking about? 18:19:19 * danbri-SWWS just checking still online 18:19:27 "function/dataflow metaphor 18:19:31 yes - please do keep taking notes - i am in the ontology working track so some highlights from yours are interesting. might not need quite so much detail though if you want to take a break 18:19:43 "input: customer name; origin; destiniation; pickup date 18:19:53 "preconditions: knowlege of the input 18:20:01 "-> AcmeBook Truck Shipment 18:20:07 "-> outout: confirmaition number 18:20:16 * danbri-SWWS gives up scribing diagram into ascii in real time 18:20:37 (nice slides though! dunno if online yet) 18:20:54 "simpler example of the way we'd expand a simple process to a composite process 18:21:10 "expand into: confirm shipping region, get quote, get shipping dates, ...etc" 18:21:44 metacomment: this is really interesting, very "AI meets the real world" 18:21:48 -- 18:21:51 "ServiceGrounding 18:22:02 * rrs suggests dlm change his nick to 'dlm-t1'; i.e. xx-trackid 18:22:04 "ServiceProcess: a spec of service access info 18:22:16 "specifics: comms protocols, transports etc 18:22:27 * danbri-SWWS wonders who dlm is? 18:22:35 Deborah McGuinness 18:22:44 * danbri-SWWS waves! 18:22:55 co-facilitating the ontology track 18:23:09 * rrs apologies to Deb for gender error 18:23:12 "related work: UDDI, WSDL, XLang, ebXML, (dot).NET, BizTalk, e-speak etc 18:23:40 is anyone scribing the other tracks (into other channels?) Do we need a logger running elsewhere? 18:23:48 "Questions!" 18:24:07 "EricQ: in terms of the abstract process description... 18:24:18 "...do you have any fun use cases for playing around?" 18:24:31 "speaker: go to web site, yes, we have congo.com example scenario 18:24:46 "...alpha air, yahoo stuff etc; will be building this up over next few months 18:25:02 "ericP: you have demos showing 'this is where the user wins' etc 18:25:29 "speaker: yes; these services are out there, so you need to be able to decompose desires into functional 18:25:44 Speaker=Terry Payne, CMU 18:26:11 question2: ...most services will come from web servers will use WSDL etc 18:26:25 "...how do you think this is going to all work out?" 18:26:32 (question transcribed badly) 18:26:55 answer from Daniel ??? CMU: "we're attempting to build something compat'ble with uddi, wsdl etc... 18:27:25 "...wsdl is a message essentially. what we have in the process model is a someone bigger picture which will include a msg specification in next release (currently this part fairly shallow in daml-S) 18:27:27 dlm has quit 18:27:41 "...you'll see a msg spec from us that's interconvertable w/ wsdl 18:28:01 "...w.r.t. UDDI, UDDI is very flexible, allows other descriptive practices to be plugged in, should be ok there 18:28:10 "(another DAML_S speaker...) 18:28:23 "the fact that there's a formal model udner this can be hidden from user 18:28:43 "...we're trying to develop more expressive languages that can describe complex relationships between services that can be manipulated 18:28:56 "...want to be able to automate things that aren't be automated right now 18:29:12 "...not just more sophisticated discovery, but automated execution, monitoring, composition 18:29:28 "...hopefully our language is in harmony with what's happening in industry right now, but more expressive 18:29:39 "...we don't want the language to be marginalised; invite input 18:30:09 questoin 3: "...lots of standards work in Business Objects community, they're trying to come up with common model for describing processes etc 18:30:17 ..."how much does that inform work of this spec? 18:30:43 "(answer) s/daniel/david martin/: We have backgrounds in many contexts... 18:31:09 "...hopefully enough context to avoid wheel re-invention 18:31:28 reminder for logs: http://www.w3.org/RDF/Interest/#irc -> ... http://rdfig.xmlhack.com/ 18:32:00 another q: people won't move to new markup languges 18:32:45 sheila McIllraith (another DAML-S team person) 18:32:47 dlm has joined #rdfig 18:32:59 "...tried not to impose too much on users of this stuff 18:33:40 another q: "can you talk more about conditions stuff, transactions support etc 18:33:52 "...for serious business apps we need this 18:34:13 David Martin: "a subgroup of our coallition is investigating these issues; more details will be in next version of the spec 18:34:59 Q: "danger of saying 'you can only use daml-s if you have a transaction model... 18:34:59 Label Q not found. 18:35:20 next speaker: Searching forServices on the Semantic Web Using Process Ontologies 18:35:37 Abraham Berstein (NYU) and Mark Klein (MIT centre for coodination science) 18:35:56 ab speaking; of NYU-Stern School of Business; Klein is of MIT Sloan. 18:36:08 ab: "preaching to the choir... Imagine you want to build a house 18:36:14 "...lots of services needed to achieve this 18:36:34 "design building; financing; contracting; carpeting; painting; landscaping..." 18:36:48 "imagine you need the last 1000$ of the building 18:36:59 "you type 'load house 100,000 into google 18:37:23 "you find blah... loan calculator, a mortgage house, and (!) campaign against arms trade 18:37:54 "Or you try a better search using TFIDF (using vector representation), you'll do better 18:38:10 "...but none of the top hits provide a loan service." 18:38:28 "We need to find _Services_ rather than documents 18:38:42 "...we want an agent or software component that'll do something for us 18:39:00 bwm has joined #rdfig 18:39:04 "...or software to download; or an organisation who'll do it; or a description 18:39:07 JHendler has joined #rdfig 18:39:14 hi brian, jim! 18:39:24 JHendler is now known as JimH-SWWS 18:39:25 * danbri-SWWS continues scribing from webservices 18:39:38 * JimH-SWWS wishes Stanford rooms had electrical outlets 18:39:42 "currentstate of the art / limitations of current tech 18:39:59 "recall/precision tradeoffs 18:40:29 "keywords -> keywords + TFIDF analysis -> frames 18:40:48 "... enumerated vocab; semantic nets; imprecise matching 18:40:56 "...we want to get to deductive retrieval 18:41:07 ...But, of course, precision and recall are useless for evaluating interactive browsing 18:41:08 "mostcommon techniques proposed: frames/slots based and deductive approach 18:41:17 (my comment, not his) 18:41:33 (yes, info retrieval folk often ignore browsing/hypertext) 18:41:46 "frame-based: matching slots, eg. Jini, eSpeak etc 18:41:55 "prob is that they require perfect matching of the vocab 18:42:16 (yes! big prob with Javaspaces/JINI, it's very crude. Are these 2 java beans the same?) 18:42:38 "prob also: can't query on the 'inside' (or how) of the service 18:43:09 "deductive retrieval forservices... 18:43:30 "eg. name: sytnax: input-types, output-types: semantics: (precond/postcond) 18:43:44 "...can be computationally very expensive. Related to AI planning literature. 18:44:00 "indexing is manually very complicated (eg. MS Word) 18:44:05 "The challenge is to... 18:44:36 "...capture enough service/query semantics to _increase_ the precision/recal without making it _too difficult to use_" 18:44:54 "...in comparison with /keyword based approaches/ and /deductive based/ 18:45:01 "...we write these two off, focus on frame-based 18:45:07 "...take a Process Ontology 18:45:15 danja has joined #rdfig 18:45:31 "define a process ontology; then use this to index services and announce them on the web 18:45:37 "users define queries that match on these 18:45:41 "Process Ontology: 18:45:58 "...builds on an MIT project since '93, the Process Handbook Projec 18:45:59 t 18:46:06 "...a handbook of how processes work 18:46:17 "over 5000 descriptons at various levels of abstraction 18:46:46 "specialisation hierachy of nouns and of verbs 18:46:59 "...eg. _sell_ and _loan_ combine 18:47:04 "noun/verb pairs(?) 18:47:36 (shows sell loan process model) 18:47:51 "you might be interested in a loan that has a specific behaviour when you default the loan 18:48:07 * danbri-SWWS actually likes the idea of taking a crapload of verbs from (eg) wordnet and generating event ontology from that 18:48:30 dlm is now known as DeborahMcGuinness-SWWS 18:48:35 "ok... we know how to define services. So how to we index them? 18:48:46 "...you take an existing item in your ontology 18:49:07 "..eg Grant Loan; and you take the Grant Mortage 18:49:16 (is he telling us how to subclass? i don't get it?) 18:49:31 "When you're granting a loan, its w.r.t. real estate value 18:49:45 "So: analyse real estate; establish mortgage conditions; ... 18:49:50 "So how do you query? 18:50:06 "query is combination of a clauses; entities and relationships 18:50:18 "we want: a loan that accepts as collateral a piece of real estate 18:50:20 * JimH-SWWS wonders about the scaling of this work -- and how well it will work on the web - also assume (i) keyword base (why do that on the SW?) and (ii) that it works when you do matches among small homegeneous sets -- but won't work when the services can include thousands of things unrelated to the main request... 18:50:30 "eg: ENTITY ?proc isa sell-loan 18:51:00 * danbri-SWWS assuems youd filter by provenance of service descriptions first 18:51:17 "probably 50% of room here wil l say 'you could do that with a frame-based simple match 18:51:18 "but.. 18:51:29 "say you wanted a loan that won't wipe me out if i default on the loan 18:51:42 "...you're looking in effect for a mortgage that hhas insurance 18:52:00 * danbri-SWWS thinks this sounds like olde AI stuff 18:52:12 "..you can't do this with frame based stuff." 18:52:24 Question: "why dcan't you do it w/ a frame based approach?" 18:52:53 "because frame based approach doesn't let you define whats going on inside the box 18:53:31 "slot based approaches typically do not announce what's 'inside the box' (to my knowledge)" 18:53:34 ... 18:53:40 "Finding Matches 18:53:43 I think he means you cannot do it in simple UDDI match approach - which is not real deep 18:54:04 "1st part: matching of clauses w/ ontology + service declarations 18:54:12 * JimH-SWWS I agree with DanBri, sounds to me like repackaging of a lot of old ideas 18:54:19 "we can use Mutation Operators on the query 18:54:37 "We can put sub-task to be any number of levels down to make it more general 18:54:53 "...conclusions... 18:54:53 "Need a set of service retrieval approaches 18:55:14 "...that capture enough service/query semantics to increase precision/recall without making it too difficult to use 18:55:24 "Ontology-based approach is more expressive 18:55:39 "...simplifies classification 18:55:40 "...better ontology -> better matches 18:55:42 --- 18:55:46 Questions 18:55:58 Q1: "you characterise deductive retrieval as wonderful but expensive... 18:56:04 "...but your examples also look fairly copmlicated 18:56:18 "...what can you say about scalbility of this? eg. googl.e indeex 1billion+ pages 18:56:20 "reply: 18:56:35 "2 parts... announmement of service... and complexity of matchup 18:57:03 "latter easier to answer... the query language i show here may look complex, but doesn't have combinational explosiion we see with PLanning / deductive retrieval appprach. 18:57:14 "talking about the 1st part, announcement + indexing... more interesting 18:57:26 "...you're trying to describe how your service performs in a logic-type language 18:57:45 "...nice thing of our approach is you're not starting from a blank slate 18:57:58 "...you can describe them in terms of pre-existing services and ontologies that are already defined 18:58:10 "...you may choose to make it complex; but you needn't. 18:58:22 "..also: those descriptions often exist in your company 18:58:32 "...whereas logic-based descriptions often don't 18:58:35 * JimH-SWWS notices that the gap between AI and Business types in sem web web services is even worse than in ontology/rule/query types *sigh* 18:58:44 "carol Groble: i'mconfused about your critique of frames 18:58:54 "you seem to be proposing a subsumption based approach 18:59:00 * JimH-SWWS lots of people using same words to mean drastically different things 18:59:17 "you quote Lasse(sp?), yet s/he does just what you're dismissing 18:59:24 "reply: 18:59:33 "Let me say what i think we're trying to say in paper... 18:59:47 "most of the frame based approaches... they draw this box (slide!) 19:00:01 "...they describe inputs/outputs, they don't describe what happens inside the process/box 19:00:14 "...so queries about what's going on inside won't get answers 19:00:15 danjay has joined #rdfig 19:00:23 * JimH-SWWS notes that the Web Services room has over 100 seats in it, and is completely full (standing room only - about 15-20 extra folks -- over half the workshop. Definitely an important area 19:00:28 "...prob with 'frame' word is that its overloaded 19:00:34 danja has quit 19:00:44 "carol: people are pretty clear about what a frame is!" 19:01:05 another q: "how does this differ from disambiguation techniques already being used in industry 19:01:28 "..also how can this be used in the Web if this stuff is behind firewalls; like deep web crawling; questions of privacy" 19:01:32 * danbri-SWWS misses guts of the question 19:01:36 reply: 19:02:00 "...using unermated vocab; sem nets; stats... On top of this we give a process structure... 19:02:16 "...w.r.t. firewalls: we are not concerned with this 19:02:30 "...we are concerned with doing efficient retrieval; next step is to get people to buy into it... 19:02:37 * JimH-SWWS speaker claims danbri's idea (taking a crapload of verbs from (eg) wordnet and generating event ontology from that) won't yield good results - but unclear he has any real alternative 19:03:15 (discussion of whewther people will open up their processs descriptions to wide view) 19:03:31 shellac has joined #rdfig 19:03:34 david martin: "market forces will encourage companies to describe their processes at some level 19:03:44 * danbri-SWWS waves at shellac; is taking notes from SWWS session 19:04:03 last questoin: " 19:04:13 * shellac was perceptive enough to realise 19:04:22 ...the focus on describing what goes on inside the box is key 19:04:27 * danbri-SWWS didn't realise shellac saw the log :) 19:04:52 I saw one line - that and your nick was enough 19:05:02 ..."huge database plus a thin layer of execution engine; build on understanding of schema above" 19:05:49 speaker, "In some industries you are forced to prove you know what you're doing, eg. ISO 9000 auditing 19:06:47 JimH: (chairing) We have tommorrow afternoon w/ time for discussion on this key issue... 19:07:22 nextspeaker... 19:07:27 David Trastour, HP Labs 19:07:54 dc:title "a semantic web approach to service description for matchmaking of services" 19:08:12 ("we don't consider behaviour here, but rather characteristics of services") 19:08:17 outline: 19:08:17 - intro 19:08:21 - rdf approach 19:08:25 - daml and description logic 19:08:29 - conclusions 19:08:30 .... 19:08:31 intro: 19:09:00 "b2b marketplaces; strong b/g in negotiation; automation of differnt phases of ecommerce process..." 19:09:07 * danbri-SWWS wonders if speaker if from bristol labs 19:09:17 "interesting contrast between SW approach and UDDI etc 19:09:30 "...we're trying to automate more; UDDI provide only a basic respository 19:09:41 "...we're tryign to do much more, put some intelligence, reasoning on top 19:10:11 diagram: matchmaking, negotiation, contact formation, fulfillment/execution 19:10:24 we're not trying to do matchmaking for one individual to one service instance; 19:10:30 instead, doing matches between classes 19:10:37 eg group several buyers together to form an auction 19:11:10 matchmaking: proceess by which parties that are intersted in having exchange of economic value are put in contact with potential counterparts 19:11:13 ... 19:11:40 ..."electronic interactions allow increased reach across a global market: proliferation of offers and requests 19:11:56 ..."automate the matchmaking prcess: electronic service description, matchmaking algo" 19:12:11 * danbri-SWWS realises he spent last 2 mins transcribing slides instead of spoken words! 19:12:30 "3 functional aspects of matchmaking: advertising, querying, browsing... 19:12:45 "advertising and querying similar; adverts are more persistent, that's all 19:13:03 ("...like stored queries) 19:13:20 "browsing: to form and create service descriptions from various ontologies 19:13:29 "...augment likelihood of matching 19:13:37 "...eg bulletin board 19:13:57 "...one typically browses to see how people are describing their offerings, and then describe own in simialr terms 19:14:05 "Service Description: 19:14:11 "...whole space of possibilties. 19:14:39 "...typcially i don't have all the details; often you'll go into a shop and come out with a different purchase than one originally expected! 19:14:40 * danjay found old post : Speaker David Trastour - Hewlett-Packard Laboratories - Bristol - UK 19:14:54 hehay, bristol rules :) 19:15:19 requirements: flex/exprssve metadata model; semistructured data; support for types/subsumption; expreesion through constraints 19:15:55 ..."we want to be able to symmettrically describe offers and requests 19:15:58 * danjay danbri's transcript rules! 19:16:17 thanks for the encouragement! my fingers getting tired... 19:17:03 "car selling example: one might want to add in more data 19:17:20 "...and we conversely don't want to force folk to add in stuff they may not... know, or want to specifiy 19:17:32 sbp has joined #rdfig 19:17:40 "...contrast w/ UDDI, where you're sometimes forced to fit data into rigid descriptiove templates 19:18:36 "matchmaking algo: two ads are compatible if some constraints are satisfited. 19:18:51 "...in both ads and there is no contradiction between them 19:19:03 "the more the 2 ads have in common (ie. the more the constraints are compatible) the better the match" 19:19:19 * danbri-SWWS intrigued by talk of 'better' matches, measures/metrics etc 19:19:20 bwm has quit 19:19:32 (often this logic-based stuff doesn't admit degrees...) 19:19:43 ..."moving on to describe RDF based work 19:19:47 "rdf approach 19:19:57 "using Brian McBride (bwm)'s Jena stuff 19:20:05 "good as it gives us an API to the rdf model not the syntax 19:20:13 "each service description... 19:20:21 * danbri-SWWS misses slides detail 19:20:56 * danbri-SWWS reminded of his old MSc projects from '97 19:20:57 http://ilrt.org/~ecdb/msc-projects/ilrt-projects.g.html 19:20:57 F: http://ilrt.org/~ecdb/msc-projects/ilrt-projects.g.html from danbri-SWWS 19:21:09 F:|Danbri's 1997 bristol msc projects writeup 19:21:09 titled item F 19:21:37 "example... myAdvert service=myPurchaseOfRoses, myshop... 19:21:56 (i can't transcribe this uml-ish chart; wait for url of slides; sorry guys!) 19:22:09 * danbri-SWWS recalls the little service description example in rdfs spec 19:22:19 "difficulties: in plain RDF, you don't have very rich metadata 19:22:25 "...not having datatypes is a pain! 19:22:29 "...string literals only 19:22:43 use DAML for your datatypes :-> 19:22:46 "...so we came up with idea of annotating our onotlogy/rdfs with matcihn rules 19:22:51 "visitor pattern 19:22:56 "default matching rule 19:23:01 s/DAML/DAMLw\XSD 19:23:13 "(we put the java class in the ontology 19:23:27 "we have matching algos for 2 rdf graphs 19:23:41 ericp: 19:23:47 ...could be 50 adverts/ 19:23:57 "Yes. Not going itno detail here" 19:24:10 "talking about our DAML version 19:24:17 paper for current presentation : http://www.semanticweb.org/SWWS/program/full/paper52.pdf 19:24:31 "...defintiely richer, solves our problem w/ datatypes, and w/ constraints 19:24:42 url: thanks! can you get the earlier ones for us too? 19:25:07 if someone could chump them and annotate w/ url into these irc logs would be great 19:25:08 will do 19:25:11 ta 19:25:19 "example in daml+oil.... 19:25:39 They're all linked to from: http://www.semanticweb.org/SWWS/program/index.html 19:25:57 http://www.semanticweb.org/SWWS/program/index.html 19:25:57 G: http://www.semanticweb.org/SWWS/program/index.html from sbp 19:26:00 "example uses a class defined as an intersection of other classes 19:26:04 G:|SWWS Full Program 19:26:05 titled item G 19:26:09 "...constraints naturally expressed in daml easily" 19:26:14 G:Includes links to the papers in PDF 19:26:14 commented item G 19:26:32 * danbri-SWWS wonders whether this is +ve feedback on the RDFS design 19:26:38 "description logics... 19:26:47 "use of a desc logic reasonaer to classify the service descriptions 19:27:04 DeborahMcGuinness-SWWS has quit 19:27:07 "daml+oil is too rich for current dl reasoners (SHIQ dl): we want a subset of daml+oil 19:27:27 "because of lack of support for individuals and datasets, problem 19:27:35 (asks for input from dl experts) 19:27:42 "matchmaking algo: 19:27:49 " matches for service desc S: 19:28:05 " - equiv concepts; sub-concepts; super-concepts of S that are subsumed by the service desc concept 19:28:11 * danjay thanks sbp 19:28:17 "Using the FaCT reasoner for that 19:28:49 np 19:29:41 conclusions: "firstexperience of sw tech and tools (jena, protege, fact, racer) 19:29:56 "interest in rdf stres; 19:30:06 "future work: (see slides)". 19:30:08 questions 19:30:09 --- 19:30:18 "any reference implemetnation of algo? 19:30:22 "yes! we're done it. 19:30:27 "is it available? 19:30:28 "maybe" 19:30:28 . 19:30:49 JimH-SWWS has quit 19:30:53 tony payne: in DAML-S community, we do have a demo that does matchmaking w/ profiles; see daml-s website for demo 19:31:12 question: in your model, how flexible are the metrics for a match? can users change params? 19:31:27 ...each user may have a different interp of sufficiency? 19:31:44 reply: a service provider may put a very general or a very specific expression 19:32:02 q: "...seem to be two sides of this; often user-defined 19:32:10 ericp: if you want to weight these 19:32:18 carol groble: yes, its the weighting 19:32:33 ...see enrico F's mid-90s work on this, with weighted similarity metrics 19:32:42 ...its in the literature. we built one too! do-able 19:32:58 questoin: is there something specific about matchmaking that's particularly w.r.t. web services? 19:33:22 ..."reply: " 19:33:38 carol (to me; sat next to each other) tversky 19:36:19 sbp has quit 19:36:38 danbri-SWWS has quit 19:42:48 * danjay thinks enrico F == Enrico Franconi 19:44:23 sbp has joined #rdfig 19:54:59 sbp has quit 20:04:41 shellac has quit 20:11:13 http://www.cs.man.ac.uk/~horrocks/FaCT/ 20:11:14 H: http://www.cs.man.ac.uk/~horrocks/FaCT/ from danjay 20:11:46 H:| FaCT Reasoner 20:11:47 titled item H 20:12:12 H: FaCT (Fast Classification of Terminologies) is a Description Logic classifier 20:12:12 commented item H 20:14:02 H: Lisp based (has CORBA inteface), GNU 20:14:02 commented item H 20:50:39 tdxdave has quit 20:52:57 danjay has quit 20:53:41 shellac has joined #rdfig 21:02:50 em-swws has joined #rdfig 21:05:28 em-swws has quit 21:05:54 em-swws has joined #rdfig 21:06:08 * em-swws returns... 21:09:31 DeborahMcGuinness-SWWS has joined #rdfig 21:10:00 * em-swws waves from across the room to DeborahMcGuinness-SWWS 21:12:28 JimH-SWWS has joined #rdfig 21:13:08 JimH scribing from web-service/applic meeting -- this session is application 21:13:24 first talk is the PPT to DAML work discussed this AM 21:13:54 speaker Marcio Trello (?) - see E: for web pointer 21:14:12 E: presented at SWWS as well 21:14:12 commented item E 21:14:49 ppt is annotated with graphic icons that map to specific semantic contents 21:15:00 ppt can export/import this markup 21:15:12 brifing can be rebuilt, to some degree, from the briefing 21:15:22 oops - mean from the markup 21:15:26 shellac has quit 21:16:21 Architecture slide presented, doesn't make it clear in detail 21:16:36 don't use briefing associate to do their briefings... 21:17:03 shows a satellite/groundstation example that they've been demoing for over two years 21:18:03 * JimH-SWWS anyone know a good network time server- mine is messed up and causing me intermittent problems 21:18:10 * JimH-SWWS ?? 21:18:56 demos system - but not clear how it all works 21:19:02 having demo problems as well 21:19:29 sbp has joined #rdfig 21:19:55 "it is supposed to show this slide which has a semantic ontology in DAML" 21:20:17 "Would create a URI for each object" 21:21:09 shows some examples of analysis it allows - creates a link that doesn't go anywhere, and the system notes the problem 21:21:12 DanC has joined #rdfig 21:21:21 DanC is now known as DanC_swws 21:21:35 speaker is very embarassed, but carries on nonetheless - 21:21:46 * DanC_swws waves from the web services session at SWWS, via cell modem 21:21:49 conclusions: Encourage generation of semantic markups via COTS tools 21:22:10 * DanC_swws would like wireless details... nssid? 21:22:18 ....semantic markup produced at no extra cost 21:22:22 future work 21:22:30 ...new metaphors (example containers) 21:22:39 ...visual rendering of attribute values 21:22:43 ...browsing tools 21:22:57 ...Use similar idea in Word 21:24:24 Scribe asks: but you aren't showing much value to the DAML 21:25:26 I want to read from a DAML file - you just show a particular set of things I need to know. 21:25:36 answer: doesn't really understand the question 21:25:48 answer: his boss knows the correct answer 21:25:57 (sorry, I said that, not him) 21:26:04 heh! 21:26:09 answer: he says we can only do very limited stuff 21:26:46 * JimH-SWWS is embarassed - was meant to be a powerpuff question because they are doing good things - but this student obviously only knows about some of the work... 21:27:30 next question: similar - like the "markup as byproduct" idea, but seems like you require someone in the loop to do a lot of graphics work - doesn't that defeate purpose? 21:28:10 answer: they can use these tools to build content - but someone needs to engineer the ontology (which is a fair answer IMHO) 21:29:36 3rd: asks question a better way "I want to go to a new domain (eng. diagrams) how hard is that to do"? 21:29:48 * sbp notes the URI for this paper is at: http://www.semanticweb.org/SWWS/program/full/paper54.pdf 21:30:29 this strikes me as quite similar, in some ways, to my work on circles-and-arrows diagrams. Somebody has to put in the smarts about which properties get drawn as arcs etc. 21:30:59 * DanC_swws tries to get wireless access using linux... loses... wonders if anybody else is winning. 21:31:29 DanC, it's very similar - but they need some better tools (which they are working on) to make it easier (interactive) -- that is something you folks could use 21:32:24 the difference is here there is a lot of domain specific stuff (based on content), where the circ-arrows stuff are based on syntactic figures of DAML/RDF (mainly at present) 21:33:10 end of speaker 21:33:12 -------- 21:33:37 Next talk ITTALKS - Tim Finin presenting - also a DAML project (ittalks.org) 21:33:51 http://umbc.edu/~finin/papers/swws01/ 21:33:51 I: http://umbc.edu/~finin/papers/swws01/ from JimH-SWWS 21:33:59 * sbp wonders how synonymous "DAML" is beoming with "Semantic Web" :-) 21:34:14 I:|Tim Finin presentation at SWWS 21:34:14 titled item I 21:34:23 I: ITTALKS project 21:34:23 commented item I 21:34:43 I:cf. [http://www.semanticweb.org/SWWS/program/full/paper41.pdf|PDF at SWWS homepage] 21:34:43 commented item I 21:35:36 * em-swws listening to interesting paper on Ontology versioning http://www.semanticweb.org/SWWS/program/full/paper56.pdf 21:35:42 * JimH-SWWS sbp, obviously a positive thing - right? (remember DAML sits of RDF - so if you're bothered, just think of DAML as the killer ap for RDF -- I once told danbri that would happen (and now I don't need the ":->" anymore 21:36:09 this talk is on line and straight forward - em want to scribe his talk? 21:36:53 * DanC_swws gives up on wireless... 21:36:56 JimH-SWWS, signal strength seems odd here... i don't think so 21:37:09 ok, we have good signal here. 21:37:16 Of course, a very positive thing. People really see DAML as a serious application, and as more and more people use it, that benefits the Semantic Web... but I hope that we don't get a "Mosaic/WWW" thing :-) 21:37:20 i'll try and capture some of the salient points 21:37:25 DanC_swws has left channel 21:37:49 s/as a/as useful for... 21:38:46 sbp real difference between DAML and generic RDF in this respect is that I (in my DARPA role) give people over $10M/yr to build DAML apps... 21:39:27 Well, DAML is clearly the first step on the TimBL diagram where a mass of incoherant data becomes data fit for a Semantic Web 21:39:37 And yeah, money is very important... 21:39:58 * sbp thinks he should start building some DAML apps :-) 21:40:12 em, at some later date check out http://www.cs.umd.edu/projects/plus/SHOE/pubs/aaai2000.pdf which is a paper on versioning Jeff Heflin wrote - I see it is cited in that paper you're hearing 21:41:33 * JimH-SWWS I agree with sbp - he should :-> 21:41:48 * sbp notes that EARL uses DAML in the Schema. Impossible not use use it, really 21:42:33 JimH-SWWS, thanks for the reference! 21:42:49 * em-swws wishes these would be XHTML instead of pdf 21:43:05 * JimH-SWWS yes - we (royal we) were pleased to see that - 21:43:27 * sbp wonders if there are any PDF to XHTML convertors around... 21:44:03 ta da: http://www.ra.informatik.uni-stuttgart.de/~gosho/pdftohtml/ 21:44:12 * JimH-SWWS em - problem is that these are published in various conferences which insist on camera ready in pdf, because the XHTML stuff is problematic w/respect to actual layout (a feature usually, but a problem to publishers) - converter sounds nice... 21:45:18 * em-swws understands.... 21:45:26 * JimH-SWWS tim - you still here? When you and I were talking to Mark Lewis (social network guy) you sent a web pointer to some ideas about "fractal" reps on the web -- do you have a pointer to that around? 21:46:32 Presenter discusses why agents haven't been as popular as they might be -- using "message based P2P" rather than publishing on web 21:47:06 UMBC exploring use of agents (includes calendar agent) which publish on web, use DAML for communication - interesting idea 21:47:38 I: Slide 20 shows interesting view of agents on semantic web 21:47:38 commented item I 21:49:07 * JimH-SWWS Tim - found it, never mind. http://www.w3.org/DesignIssues/Fractal 21:49:11 Jim - I: gives a 403 forbidden 21:51:48 scribe apologizes, got caught up as speaker said lots of nice things about DAML... 21:54:08 * JimH-SWWS sbp - trying to check, can't seem to get out to web well, but can get to irc - strange 21:55:14 * JimH-SWWS Finin has protection set wrong, will ask him to fix it 21:55:29 Cheers 21:56:17 * sbp wonders why URIs are being used in strings at http://www.cs.umd.edu/users/hendler/jhendler.n3 21:57:21 future work - extend to "Xtalks" -- arbitrary domain 21:57:34 * JimH-SWWS hmm, sbp - that was done by automatic translation from jhendler.daml -- which doesn't - suspect it was a bug in an earlier version of rdfton3, which I never noticed 21:58:56 question: looked into RSS? 21:59:23 answer: yes, are publishing in RSS - but not integrated 21:59:43 * sbp works on an automatic translation of JimH's DAML thing 21:59:53 Ta da: http://swag.semanticweb.org/n3tordf?uri=http%3A%2F%2Fwww.cs.umd.edu%2Fusers%2Fhendler%2Fjhendler.daml&from=rdf 21:59:58 * JimH-SWWS am trying to get him and Libby talking - he has a lot of scheduling info that could be good for the calendaring apps 22:00:49 * JimH-SWWS notes that sbp's is better than DanC's -- can you guys fight... 22:00:58 shellac has joined #rdfig 22:01:32 We'd have to check our schedules... :-) 22:01:52 Actually, we use DanC's notation3.py 22:02:31 It's using the SWAG web form. We get the latest version of notation3.py automatically through CVS 22:02:49 sbp, yes - I think I must have used an old version before this was fixed - I usually use the swag form 22:03:40 Cool. But the URIs are still in quotes: there's a problem in the original .daml file 22:04:43 * sbp wonders if Jim knows about DanBri's FOAF stuff 22:05:17 * JimH-SWWS yes, but only what I've seen on web - danbri has promised to tell me more about it, but we neevr seem to get there 22:05:48 There's a schema for it at http://xmlns.com/foaf/0.1/ 22:06:03 using DAML :-) 22:06:08 * JimH-SWWS sbp - would welcome input on what is wrong w/DAML file - was my first RDF, done by hand, and drove me crazy figuring out the rules for when to use quotes and when not to. 22:06:32 Ugh, tell me about it. RDF in XML certainly wasn't meant for people scribbling it down... 22:06:33 next speaker 22:06:44 -------- 22:06:56 repositories/metarepositories 22:07:14 Open learning repository - RDF Metadata (look - no DAML!) 22:07:33 The correction in the .daml file: s/http://[...]/ 22:08:16 PDF for this talk: http://www.semanticweb.org/SWWS/program/full/paper38.pdf 22:08:48 "Open Learning Repositories and Metadata Modeling" - Hadhami Dhraief, Wolfgang Nejdl, Boris Wolf, Martin Wolpers 22:09:43 (corrected element needs a "/>" at the ned of it... oops 22:11:39 This paper is huge! 1.61MB 22:16:15 tim has quit 22:16:17 scribe was busy preparing his comments for tomorrow's panel - sorry. 22:16:31 * JimH-SWWS KUTGW? 22:16:46 * JimH-SWWS oops, never mind - just hit me... 22:17:59 sbp has quit 22:22:12 sbp has joined #rdfig 22:25:26 em-swws has quit 22:26:37 AaronSw has joined #rdfig 22:27:00 * JimH-SWWS hi aaron. 22:27:11 Hi Aaron 22:27:25 shellac has quit 22:27:32 Hi there! 22:27:42 Finally got my computer to work... 22:27:45 * JimH-SWWS was impressed to see Aaron is pals with Doug Englebart - way to go! 22:27:49 ;-) 22:27:55 And Ted Nelson! 22:28:01 That crosses two life goals off my list! ;-) 22:28:44 was that Ted Nelson with him -- gezz, I knew he was someone I used to know, but couldn't place him - foo. He and I were friendly a lot of years ago, but I guess we've both changed... 22:28:58 Friends... really.... wow 22:29:46 well, research associates - one of his first employees at Xanadu was a college classmate of mine, so he introduced us and we all went out to dinner and the like a bunch of times. 22:30:11 Really? Quite neat! 22:30:24 DeborahMcGuinness-SWWS has left channel 22:30:42 yes, they were planning a new revolution in which there would be a web of hypertext. Somehow the idea never caught on... 22:30:51 Heh 22:31:12 AaronSw is now known as Aaron-SWWS 22:31:16 problem is they were focused on trying to make money, rather than giving it away -- suspect Ted rues that to this day... 22:31:22 break time. 22:31:26 Hmm, where am I... 22:31:31 JimH-SWWS has quit 22:31:51 Describe the place 22:32:09 OK< so it's a very large conference hall. 22:32:18 There's Chris.Busssler@Oracle.com on the screen. 22:32:42 In the main buidling. Very nice new architecture, pretty glass, etc. 22:33:14 OK, now it's over.... folks are leaving, etc. 22:33:24 Aw, you got there too late :-) 22:33:57 What track are you planning on going to next? 22:34:11 Don't know... I'm going to go see where Libby went. 22:34:15 See you in a bit... 22:34:23 O.K., c'ya 22:34:28 bye 22:34:35 Aaron-SWWS has quit 22:37:46 AaronSw has joined #rdfig 22:37:50 wb 22:37:56 Thanks. 22:38:09 Did you find her? 22:38:19 Yes, she's off chatting with folks. 22:39:03 wb 22:39:17 woo! good ole' lag 22:39:19 Hi Tav! How's it going.. 22:39:24 pretty good 22:39:29 how's the conference? 22:39:45 Well, the conference isn't so great but the comapny is. ;-) 22:40:17 it's going out to dinner / parties in the evenings [with the right people] that makes conferences so great 22:40:38 Yep -- I've got a date for dinner tonight which should be rather exciting. 22:57:18 * AaronSw logs live from the Demos session 22:57:26 presenters play a little game of who's on first 22:58:22 ...sound tests... 22:58:31 Track: Tutorial Track, TCSEQ 200 (http://www.stanford.edu/dept/registrar/tcseq/200.html), Demos; (Facilitator: Natalya F. Noy, SMI, Stanford University) 22:59:13 First off we have VerticalNet: Ontology Builder 22:59:21 Industrial Strength Ontology Management 22:59:28 Verticalnet - Aseem Das, Spirit-Soft - Steve Ross-Talbot, Mondeca - Bernard Vatant, Empolis - Hans Holger Rath, LastMileServices - Raj Bapna, Stanford Medical Informatics - Mark Musen, Monica Crubezy, Natalya F. Noy, DSTC/Griffith University 22:59:28 Peter Eklund 23:02:43 demos get started 23:02:46 15 minutes per demo 23:02:49 ontology builder first 23:03:07 a multiuser generatrion and management tool 23:03:13 built on java2 enterprise platofrm 23:03:16 does import and export 23:03:26 has a verification engine to maintain consistency 23:03:31 sxecuiry for ontology access 23:03:42 has i18n support 23:03:59 a simple java swing app 23:04:04 connects using http to servlets 23:04:10 servlets connect to app server using RMI 23:04:27 business logic is implememnted in session thing(??) security, etc. 23:04:36 also set of db drivers for db acccess 23:04:45 finished ontology stored in rdbms (oracle, sql server, etc.) 23:04:58 http://www.ontobuilder.com 23:04:58 J: http://www.ontobuilder.com from AaronSw 23:05:03 J:|Ontology Builder 23:05:03 titled item J 23:05:12 showing 2 things: browsing and editing 23:05:57 opens an ontology 23:06:09 ontology interface like Windows Explorer 23:06:32 system appears to have crashed... restarting 23:06:40 ...back 23:07:04 Can do cut and paste, regrouping, etc. like in Windows Explorer 23:07:19 Shows simple newspaper ontology. 23:07:37 supports multiple ??? (recursions?) 23:07:48 ahh multiple inheritance 23:08:04 has a column list showing parents in the smae onotlogy 23:08:21 also has metaclass arch to show other ontologies (???) 23:08:35 system ontology contains pre-defined concepts like class, and metaclass 23:08:41 as well as predefined primitive types 23:08:46 and slots. 23:08:53 and predefined classes 23:09:26 all the concepts defined in system ontology are available for new ontology 23:09:51 data model is normalized, ontologies don't need to be copied 23:10:04 changes are propoagated auitpmatically 23:10:11 sbp has quit 23:10:18 all sorts of properties 23:10:27 bwm has joined #rdfig 23:10:39 if you click on the slot name you can retrieve a slot 23:11:00 You can put an ontology in a particular folder... 23:11:07 (showing how to create ontologies) 23:11:13 can do batch adding 23:11:43 tdxdave has joined #rdfig 23:12:02 can do cut-and-paste with ontologies 23:12:39 bwm has left channel 23:12:52 supports batch editing 23:13:00 AaronSw has changed the topic to: AaronSw reports on SWWS Demos 23:13:13 define a root level, children level. 23:13:44 shows how to translate ontologies for i18n 23:13:58 very easy to translate 23:20:15 can load in a translated ontology. 23:20:17 can do translations in different languages, character sets easily 23:20:19 [end of demo] 23:20:21 spiritsoft: the integration company 23:20:23 founder speaking, founded in '97 23:20:25 doing similar work as timbl 23:20:27 had bad fortune of not being where timbl was 23:20:30 ...my focus is on change in the web 23:20:31 this is about semantics and locating 23:20:33 but life is more interesting afterwords when you want to deal with change 23:20:37 i'm involved with ruleml which you'll hear more about tomorrow 23:20:37 ruleml separates out where rules can be useful 23:20:39 can be used to deal with change 23:20:41 AaronSw is now known as Aaron-SWWS 23:20:43 useful for web services 23:20:45 everyone is writing white papers on web services 23:20:49 everyone jumping on the SOAP bandwagon 23:20:53 someone asked "isn't this text over IP" 23:21:00 that's exactly what it is 23:21:09 integration problems about web services, semweb 23:21:20 interesting work is reactive rules -- spiritintellect 23:21:27 way of consolidating events and change 23:21:31 based on event-condition-action 23:21:43 most research comes from active database research 23:21:49 decoupled action from the database 23:21:58 sintellect born 3 yrs ago 23:22:05 created eca framework 23:22:11 moving it to standards bodies 23:22:29 ruleml captures deductive rules, facts. Work to be done: reactive rules, integirtyu constraints 23:22:34 working with ruleml community 23:22:41 reacitve rules based on eca model 23:22:55 where event and condition is combined, looking at changes over tiem 23:23:08 like opening address about cows in texas 23:23:20 well, you want to know about cows in the next ten minutes, not just now 23:23:33 how many cows born in paris, tx 23:23:45 it's about working together, to find how the rules can work together 23:24:00 ruleml provides rules where needed with mechansism existing 23:24:17 towards an active semweb, one about capturing change as well as morning 23:25:02 imagine four processes: pubnlisher, settlement engine, business transaction rule, and ??? 23:25:09 connected using message middleware 23:25:16 using jms interface, could be on top of HTTP 23:25:22 important: change moves around the system 23:25:30 could be an xml document 23:25:31 could be soap 23:25:46 trade publisher sends trades to queue 23:25:50 settlement engine picks them up 23:25:55 durable queue 23:26:01 don't want to lose a trade -- you get sued 23:26:19 trade publisher also publishes to a bus, settlement publishes trade to a bus 23:26:29 settlement settles osme trades, not others, some partially, others fully 23:26:34 we'll change rules on the fly 23:26:38 using an eca paradigm to do the rule 23:26:44 very event, data-centric 23:26:55 community looks at deductive/reactive differently 23:27:09 eca is forward chainging, doesn't need a rules engine 23:27:16 can compile java on the fly and send it 23:27:28 xml form could be sent instead, and agent could launch it inflight 23:27:35 (starts demo) 23:27:57 opened business transaction rule 23:28:07 uses a winnie the pooh icon, no apologies, interesting philosophy 23:28:21 rule sends events about new/settled trade 23:28:28 shows rule, displayed as a graph 23:28:40 eca framework describes interfaces of components 23:28:47 we can change any interfface/component 23:28:49 very open, standatdsx-based approach 23:29:02 rule says you're preteferred or ordinary 23:29:15 very boring way to do it -- just looks it up 23:29:27 use a deductive engine to work along to do more interesting wueries 23:29:37 waits for a settlement engine by going around in a loop, or timeing out 23:30:05 goign to change rule on the fly 23:30:11 rule publishes text which the gui picks up 23:30:41 trades are being settled in full 23:31:09 also reporting partial confirmations 23:31:39 someone's name is in the rule, taking it out... 23:31:39 checking it in 23:31:55 so now generating java, compiling java, serializing rule on message bus 23:32:00 using a jms connection 23:32:33 new rules are modified, but ones running still follow old rule 23:32:48 that's about it... 23:32:57 looking forward to working with everyone, especially ruleml folks 23:33:14 hope to show how deductive and reactive rules work together using ruleml 23:33:17 Questions: 23:33:21 does it compete with ilog rule engine? 23:33:29 A: we don't have a rule engine... 23:33:29 commented item A 23:33:31 but yes 23:33:42 A: oops - ignore that above 23:33:43 commented item A 23:34:08 java has the capabilities to interpret and introspect, so it doesn't make sense to have an interpreter on an interpeter 23:34:14 ours does events, theres does large data sets 23:34:25 [end of demo] 23:35:05 mondeca: making sense of content (TM) 23:35:13 smeantic graphs software for index mangaement 23:35:37 technology is based on topic maps spec 23:36:08 we are doing more "semantic graphs" than topic maps because we are working away and see that topic maps are just one of the semanrtic formats 23:36:15 rdf and other formats may be useful too 23:36:47 french company, 2 yrs old, in paris and one person in san jose 23:37:02 14 people working (believe we'll have 20 by end of year) 23:37:11 objective is to be market leader in index management 23:37:16 integrated solution for index managemtn 23:37:47 active participation in semweb community at large 23:37:53 especially xtm/topic maps community 23:38:18 you have big contnet to manage 23:38:23 serach enging retrievals 23:38:38 mondeca fits in the middle between data management and the publication 23:38:57 provides intermeidaite layer based on semantic graphs / topic maps 23:39:09 architecture: built on oracle db, uses graph manager/topic map engine 23:39:18 server side is oracle conversing with topic map engine 23:39:21 uses java application 23:39:51 client has all sorts of output formats, xml/svg (via xslt), html via jsp 23:39:55 sbp has joined #rdfig 23:39:57 in the middle of xml standards community 23:40:01 working with topicmaps.org 23:40:11 currently reorganizing between oasiss and iso 23:40:21 rely upon mathematicians with graph theory 23:40:35 representing graphically the graph 23:40:41 all technology based upon java app 23:40:56 features: ontologies, directories, encyc, datbase, corp knowledge 23:41:03 these are all potential markets for index-management 23:41:19 with topic maps you can manage more than just directories 23:41:24 you have graph-like directories 23:41:30 you can do databases for libararies 23:41:45 usnign collaborative tool you can build collaborative enviroments 23:41:57 how do you feed the database? 23:42:01 there are three ways to: 23:42:04 handcrafted index 23:42:09 database importation 23:42:12 and realtime indexation (tbd) 23:42:27 handcrafed: collaborators give labels, etc. 23:42:40 ... few projects to incorporate KM 23:42:57 importation: transfer relationships from classical dbs 23:43:04 ... most important demand of customers 23:43:08 ... already have databases 23:43:24 ... already rearranged database and don't want to / can't afford to change db 23:43:43 need some layer to change without breaking 23:43:51 realtime: still in R&D stage 23:44:07 build thru data mining and content extraction 23:44:09 still working on that 23:44:14 ---- 23:44:20 can deliver several client solutions 23:44:58 working on ontolofies buuilt on topic maps 23:45:13 ontolofies built on classing, etc. 23:48:39 loading topic maps in internet explorer 23:48:47 mondeca is in the center, connected to many things 23:49:52 shows stefan decker, and projects he's involved in 23:50:32 [end of demo] 23:50:54 How many more to come? 23:51:03 I don't know. 23:52:10 Out of all the demos so far, how many have really had anything to do with the Semantic Web? 23:52:19 empolis: transforming information into value 23:52:36 sbp, none with the semantic web as we know it... except for maybe the first 23:52:43 ooh, this is a bertelsman company 23:53:13 Yes, I had a feeling it was a "well, this is kind of related..." thing 23:53:17 :-) 23:53:33 shows picture with KM in the middle and access around the outide 23:54:01 KM = Knowledge Management 23:54:10 everything connected together, embedded, reusable 23:54:24 and he has a product! it's called k42 23:54:37 The "Topic Map Knowledge Server" based on topic map stadnard 23:55:15 we're following rdf with interest to see relation to other standards 23:55:38 k42 is similar to Mondeca, but has its own persistent storage, does not requre opracle 23:55:42 everything java, everything xml, xslt, etc. 23:56:16 speaker is coeditor of topic map Query Language standard 23:56:18 does link management 23:56:33 another guy with invited expert to XLink WG, supports XLINk, first prouct to 23:56:52 get a free copy of k42 for non-profit research products (with upgrades) 23:57:02 for folks who need knwoldge servers and don';t want to buidl there own 23:57:22 http://k42.empolis.co.uk/ 23:57:23 K: http://k42.empolis.co.uk/ from Aaron-SWWS 23:57:34 K:k42 knowledge management software 23:57:35 commented item K 23:57:50 K: get a free evaluation version, or free copy for non-profits 23:57:50 commented item K 23:57:58 has a web user interface 23:59:24 you can add scopes for some names