00:13:33 <danbri> * danbri catches up, grins at the logs
00:14:14 <danbri> Yes 1 year+... we started migrating over from #sw on irc.w3.org to #rdfig on openprojects in early Feb last year, before the last tech plenary w3c meeting in boston
00:18:27 <AaronSw> oh, is that so?
00:18:46 <AaronSw> i recall you making it sound like it wasn't being used
00:18:52 <AaronSw> at that plenary
00:22:38 <danbri> We weren't chumping and logging, and #sw was being used... but I wasn't publicising it 'cos it seemed like camping out on the front steps of W3C... and we'd not really figured out if that's what #sw was for...
00:43:06 <tansaku2> tansaku2 has joined #rdfig
00:44:15 <jonb> * jonb wanders back
00:45:35 <jonb> catching up...
00:46:11 <jonb> i've been talking with some folk who are interested in integrating RDF and XQuery to do distributed queries
00:46:25 <tim-not> tim-not has joined #rdfig
00:46:59 <jonb> wonders if there is anything like db.n3 in the works for XQuery?
00:49:54 <tav> tav has joined #rdfig
00:50:26 <danbri> <jonb> i've been talking with some folk who are interested in integrating RDF and XQuery to do distributed queries
00:50:41 <danbri> what do they want the XQuery bit for? 
00:51:00 <danbri> xpath, I could understand...
00:51:59 <tim-not> Of the various killer applets, dbview is IMHO one, and a xml to rdf representation of the infoset is another.
00:52:20 <tim-not> With that, XQuery becomes just an RDF query on the infoset.
00:53:01 <tim-not> query:document is  like   log:semantics, returning an infoset tree rather than an RDF formula.
00:53:35 <tim-not> Anyone thought about XQuery FLWR clauses?
00:53:57 <tim-not> I couldn't work out why the F, L and W were different. All antecedant.
00:54:36 <tim-not> * tim-not wonders what hte equivalent of d.n3 for XQuery would mean
00:54:50 <danbri> tim-not, I was just re-reading http://www.w3.org/History/1989/proposal.html
00:55:12 <tim-not> tim-not is now known as tim-onNoff
00:55:18 <danbri> you're certainly sticking to the same story :)
00:55:24 <danbri> http://www.w3.org/History/1989/proposal.html
00:55:25 <dc_rdfig> A: http://www.w3.org/History/1989/proposal.html from danbri
00:55:29 <tim-onNoff> Just a one-trick pony
00:55:42 <tim-onNoff>   that's all that horse can do
00:55:50 <danbri> A:|TimBL going on about RDF and databases, as ever... ;-)
00:55:50 <dc_rdfig> titled item A
00:55:56 <danbri> * danbri feigns sympathy
00:56:14 <danbri> So does it feel like things are all coming together yet? 
00:56:43 <danbri> A:"Accessing Existing Data", "The system must achieve a critical usefulness early on. Existing hypertext
00:56:43 <danbri> systems have had to justify themselves solely on new data."
00:56:43 <dc_rdfig> added comment A1
00:56:59 <danbri> A:...systems have had to justify themselves solely on new data. If however, there
00:57:00 <dc_rdfig> added comment A2
00:57:00 <danbri> was an existing base of data of personnel, for example, to which new data
00:57:00 <danbri> could be linked, the
00:57:11 <tim-onNoff> Actually,today was a good day -- or was yesterday?  I have been looking for time to code a DB->RDF server for ages as low hanging fruit, then yesterday turns out DanC did it
00:57:38 <danbri> A:was an existing base of data of personnel, for example, to which new data could be linked, the value of each new piece of data would be greater."
00:57:38 <dc_rdfig> added comment A3
00:58:02 <danbri> Yes, mapping from traditionally stored tables down into RDF is going to get us truckloads of data :)
00:58:15 <tim-onNoff> Yes, exactly.  Except fro the WWW initially, every docuentation  system (help system, etc) was different - needed a separte hack.
00:58:46 <tim-onNoff> With databases now, Todd already harmonized the models, so basically one hack does them all.
00:58:55 <danbri> Did you ever look at the XSB stuff too? The rewrite logic expressions (a certain SQL-happy subset) into complex SQL, eg. using '<', '>' etc within the SQL database engine, instead of pulling all the data out of the RDBMS
00:59:07 <danbri> s/Todd/Codd/
00:59:15 <danbri> * danbri nods
00:59:52 <tim-onNoff> Excpet that hte network protocols from sql to sql are vendor-dependent.  But that makes the RDF versiona killer app, as they all surafce as HTTP/RDF identical!
00:59:54 <danbri> http://www.geocities.com/SiliconValley/Bit/1116/PrologSQL.html
00:59:54 <dc_rdfig> B: http://www.geocities.com/SiliconValley/Bit/1116/PrologSQL.html from danbri
01:00:04 <danbri> B:|XSB prolog to ODBC/SQL bridge
01:00:05 <dc_rdfig> titled item B
01:00:17 <danbri> B:I think the same or similar code is available for SWI prolog too
01:00:18 <dc_rdfig> added comment B1
01:00:53 <danbri> B:There's a [http://www.geocities.com/SiliconValley/Bit/1116/PrologSQLex01.html|simple example] up there...
01:00:53 <dc_rdfig> added comment B2
01:01:47 <JibberJim> JibberJim has quit
01:02:00 <danbri> yes. Things like ODBC, JDBC and DBI hide some of those differences (as will SOAP), but if there's a big public network in the middle, going through a more documented (and useful) format makes sense.
01:02:48 <danbri> http://www.w3.org/2001/12/rubyrdf/swdoc/RAAsoap2rdf.rb
01:02:49 <dc_rdfig> C: http://www.w3.org/2001/12/rubyrdf/swdoc/RAAsoap2rdf.rb from danbri
01:03:41 <danbri> C:|RAAsoap2rdf.rb,v 1.1 -- dump the Ruby Application Archive (RAA) software package metadata from a SOAP service, map to RDF
01:03:41 <dc_rdfig> titled item C
01:04:04 <danbri> C:This was interesting; looking at the Ruby class and property model the data mapped to, I pretty much just read off the arcs and nodes.
01:04:04 <dc_rdfig> added comment C1
01:04:52 <tim-onNoff> One of the things which came up is that when you retreive the database schema, it should tell you how big the database is.  In a hwole lot of cases, it is so small that you do best to just suck it all over and deal with it locally. Then you akso get http proxying for free etc.
01:05:04 <danbri> C:One point that relates to dbview: I had the choice to _either_ use the same namespace for each property, or do a bit of tweaking to use better known namespaces (eg. DC, RPMFind, FOAF etc.). It isn't clear at what stage to map obscure RDF into well known RDF...
01:05:04 <dc_rdfig> added comment C2
01:06:18 <danbri> Interesting point. I found my queries timing out when I was querying that SOAP server, and it was frustrating that the SOAP messages they were sending me weren't being treated as 'proper' web documents that could live online as static docs like HTML, SVG etc.
01:06:47 <danbri> SOAP folk seem to think that docs written using the SOAP encoding vocab will only ever appear in a transient, protocol context
01:07:02 <tim-onNoff> There is a question as to whether that returned from an SQL query is worth caching though it is in principle.
01:07:18 <tim-onNoff> Just as most proxies don't bother caching anything with a "?" now.
01:07:40 <danbri> * danbri nods
01:08:17 <danbri> So tim, your Python talk where you were encouraging them to RDFize... did anything come out of that?
01:08:21 <tim-onNoff> * tim-onNoff wonders whether he should try to understand why Danbri is so keen on ruby
01:08:40 <tim-onNoff> Not that I know of ... but you don't generally know.
01:09:09 <danbri> Especially re software packaging. I'm talking to some Ruby people about combining two apps: a database of software packages and their dependencies on the one hand, and on the other, a software documentation tool that extracts (now to rdf/xml) metadata about classes and properties.
01:09:49 <AaronSw> tim-onNoff, it was cool to see some of your proposals in guido's talk
01:10:12 <danbri> I remembered you talking about similar apps to Python people... if anyone follows up on it, could you let me know. I don't see why different programming languages need different tools for much of this stuff.
01:10:18 <danbri> guido's talk?
01:10:55 <AaronSw> yeah, guido gave a state of the python address
01:11:27 <danbri> URIs for packages etc? That came up when we dumped Ruby class/method descriptions to RDF: http://www.w3.org/2001/12/rubyrdf/swdoc/squish-rdoc.rdf
01:11:49 <danbri> eg (excuse the irc spam)...
01:11:50 <danbri> [[
01:11:51 <danbri>   <Class rd:name="SquishQuery" rd:id="SquishQuery">
01:11:52 <danbri>     <classmod-info>
01:11:52 <danbri>       <InFiles>
01:11:52 <danbri>         <infile>
01:11:53 <danbri>           <File rd:name="../squish/squish.rb"
01:11:54 <danbri>                 rdf:about="#..%2fsquish%2fsquish.rb"
01:11:56 <danbri>            />
01:11:58 <danbri>          </infile>
01:12:01 <danbri>       </InFiles>
01:12:02 <danbri>      <superclass>Object</superclass>
01:12:04 <danbri>     </classmod-info>
01:12:05 <tim-onNoff> I must say I ended up thinking about, and basically stuck on, the question of how a triplestore (formula) could be best represented in a language like python (etc)s integreation of dictionaries.
01:12:06 <danbri> ]]
01:12:09 <tim-onNoff> oops
01:12:25 <sbp> it's odd - I seem to be coming at RDF more from a "what functions can we embed?" view, with the convergance of N3/Python as my proof/goal. But on the other hand we have the MySQL-RDF tie in
01:12:46 <AaronSw> Well, that's another function, right?
01:12:48 <danbri> ...whereas in Java, at least I could write java:org.desire.rudolf.query.etc/SquishQuery or whatever
01:12:53 <sbp> indeed
01:13:13 <sbp> it shows the flexibility of RDF...
01:13:13 <tim-onNoff> Hmmm. Dennis Quan at LCS though a bunch about python-n3 convergence, made some suggestions ... no pointers :-(
01:13:13 <AaronSw> we've got sources of triples, and consumers of them.
01:13:23 <sbp> Quan: I've seen his stuff...
01:13:33 <danbri> pointers are in w3c-semweb-ad where google can't find 'em :( 
01:13:34 <sbp> http://www.ai.mit.edu/people/dquan/n3script.html
01:13:34 <dc_rdfig> D: http://www.ai.mit.edu/people/dquan/n3script.html from sbp
01:13:35 <sandro> Dennis hasn't bought into publish-early, publish-often, alas.
01:13:40 <danbri> and sean's head
01:13:41 <sbp> D:|N3Script
01:13:41 <dc_rdfig> titled item D
01:13:55 <AaronSw> Heh. I asked Dennis for permission to share off of semweb-ad and he said ok.
01:14:39 <sbp> D:A [http://www.python.org/|Python]/[http://www.w3.org/DesignIssues/Notation3|N3] convergence experiment by Dennis Quan
01:14:39 <dc_rdfig> added comment D1
01:14:48 <danbri> integrating triplestore apis into Python, Ruby etc just as a 21st century accompaniment to classic hash/dict seems very plausible
01:15:04 <danbri> forgetting the web and all that, it remains a rather handy data structure
01:16:04 <danbri> * danbri searches for an Edd Dumbill quote on Perl and hashes and RDF, finds danbri quoting Edd instead:
01:16:09 <danbri> http://lists.w3.org/Archives/Public/www-archive/2001Sep/0014.html
01:16:10 <dc_rdfig> E: http://lists.w3.org/Archives/Public/www-archive/2001Sep/0014.html from danbri
01:16:28 <sandro> Is there any difference between (1) a hash of sets and (2) a triplestore?    I guess if the triplestore is an object, it simplifies handling multiple triplestores, which is ... odd in the first model.
01:16:39 <danbri> E:|Quoting Edd dumbill on RDF / hashtable analogy
01:16:39 <dc_rdfig> titled item E
01:17:07 <AaronSw> sandro, what's the key of the hash?
01:17:22 <danbri> E:"In programming language terms, RDF is a Perl hash or a Java Hashtable. The processing application is responsible for figuring out which bits it wants and which it doesn't"   
01:17:22 <dc_rdfig> added comment E1
01:17:26 <sandro> the name of the predicate.
01:17:53 <AaronSw> I see. subject = { predicate: object} ?
01:18:14 <sandro> sorry, what language is that, Aaron?
01:18:15 <sbp> then it's quite different from a triple store. You can't merge them, for a start :-)
01:18:16 <danbri> The analogy doesn't need to be done so literally; it's an infobag full of goodies that you can ask certain very constrained questions of 
01:18:27 <AaronSw> sandro, Python...
01:18:54 <AaronSw> it sets a variable to a hash/dict with key predicate and value object
01:18:54 <sandro> Sorry, my python is rusty.
01:19:03 <sbp> easier for OOP people to manage, though
01:19:14 <danbri> my python book is for sale
01:19:23 <sandro> The value in the hash needs to be a set, since not all predicates are uniqueProperties.
01:19:35 <AaronSw> yeah...
01:19:39 <AaronSw> you could nest them... TripleStore = { subject: { predicate: [object]}}
01:19:56 <sandro> I'm also a fan of only using uniqueProperties, ....     I dunno.   :-)
01:19:59 <sbp> Hmm... a bit messy
01:20:11 <AaronSw> indeed
01:20:25 <danbri> so we've moved from talking about how triplestores could live alongside hashes as core part of the language, to how to implement the former in the latter?
01:20:25 <AaronSw> but an interesting possibility for the plex
01:20:52 <sbp> I'd like to see you do object searches on that
01:21:03 <AaronSw> object searches?
01:21:13 <sbp> :Aaron :name ?x .
01:21:29 <AaronSw> TripleStore[Aaron][name]
01:21:32 <danbri> So, some years ago, Guha told me how they did in in 'Netscape 5'
01:21:36 <jonb> isn't it {t|f} = predicate(subject,object)
01:22:01 <sbp> yeah, but you're organizing by subject, so it's necessariliy going to be faster for subject searches
01:22:14 <AaronSw> Umm, no...
01:22:19 <AaronSw> Guido has a good old paper on how to make graphs out of lists in Python.
01:22:26 <sbp> reference?
01:22:29 <danbri> He said: "It's simple. You keep two hashtables, fwd-ptr and bakward-ptr. They takes objid/property (fwd) or rdfvalue/propery (back) as key, and has a set as the values."
01:22:50 <sbp> Hmm...
01:22:54 <AaronSw> Isn't that how rdfdb works?
01:23:03 <AaronSw> * AaronSw tries to parse that
01:23:14 <jonb> * jonb thinks back to how he used to represent graphs as lists ...
01:23:15 <danbri> So that's where all the ILRT RDF triplestores-in-RDF stuff came from, playing around w/ that model. RDFdb's similar but has provenance in there, I think.
01:23:17 <sbp> * sbp was just wondering about DanBri's Ruby approach :-)
01:23:21 <AaronSw> guido on graphs: http://www.python.org/doc/essays/graphs.html
01:23:33 <sandro> Sometimes you want to be object-oriented, sometimes you want to be triple-oriented.
01:23:51 <jonb> * jonb 's very first software job involved an RDB written on top of a graph representation language
01:24:02 <danbri> Same; the in memory version. So was my perl, though I used stringified, text-separated lists instead of native Perl arrays (I think cos I wanted to store them in BerkeleyDB without reading up on perl serialisation stuff).
01:24:11 <danbri> interesting :)
01:24:45 <danbri> http://www.rubygarden.org/ruby?RaaSuccRequirements
01:24:45 <dc_rdfig> F: http://www.rubygarden.org/ruby?RaaSuccRequirements from danbri
01:25:09 <danbri> F:|Notes on Ruby Application Archive package metadata, use cases, requirements etc.
01:25:10 <dc_rdfig> titled item F
01:27:11 <AaronSw> What's Succ mean?
01:27:35 <danbri> successor, I assume
01:27:53 <danbri> Like a baby CPAN, but I think maintained centrally
01:29:44 <jonb> its coming back to me ... my application was optimized for graph traversal
01:30:23 <bijan> Hmm. I feel that I should state, for the record, though completely insensitive to current discussion, that, naturally, if the query itself is order insensitive, then it doesn't matter that prolog is order sensitive.
01:30:23 <jonb> each subject node had a hashtable of predicates
01:30:54 <jonb> and the value of the predicate was a tree of objects
01:32:24 <danbri> So the way you name your relations (worksFor vs employee) affects performance? if lots of 'Person' instances and not many 'Org' instances, one would chose (hmm, my head hurts, I dunno...)
01:32:54 <jonb> exactly! the physical layout needs to be optimized for the charqcteristics of the graph
01:33:51 <jonb> but if you want to inference on "is-a" predicates i.e. rdfs:subClassOf
01:34:16 <jonb> then this is a really fast way to do it
01:45:26 <tansaku2> tansaku2 has quit
01:47:24 <sbp> jon, still there?
01:47:33 <sbp> this XQuery/RDF stuff is very interesting
01:47:44 <jonb> yes
01:48:35 <sbp> because they're two stunning technologies. But RDF is quite limited in the query/function department, whereas XQuery has alot of that. Shouldn't it be possible to parse XML RDF with XQuery?
01:49:01 <sbp> XQuery seems like an extension of XSLT and XPath to me...
01:49:18 <sandro> In what query sense is RDF lacking?   Aggregate functions?
01:49:29 <sandro> (assuming any of the serious RDF-query contenders)
01:49:34 <jonb> yes exactly, the XPath 2.0 and XQuery models are merged.
01:49:57 <sbp> Sandro: yeah. When you consider some of the stuff that XQ lets you do, to the stuff that's possible today in CWM... there's a gap
01:50:15 <jonb> Jonathan Robie has given a terrific demo of implementing RDF queries on top of XQuery
01:50:26 <sbp> ooh! do you have a reference for that?
01:50:48 <jonb> he showed it to me on his laptop :-)
01:51:12 <sbp> argh. Heh, oh well
01:51:28 <sandro> I don't really know XQ, but.....     but I know RDF query is simple because it can be.   You can do anything non-aggregative with just a conjunction-pattern.  Aggregate stuff requires quads/reification, which I don't think anyone has explored much.
01:51:35 <jonb> he gave a talk at Extreme Markup 2001
01:51:58 <jonb> XQuery is getting tremendous industry support
01:52:01 <sandro> the Syntactic Web one, yes?
01:52:19 <jonb> great title :-)
01:52:44 <sbp> I presume that the demonstration was of XQ on RDF... using XQ like Squish/RDFQL etc., right?
01:52:49 <jonb> Jonathan (Robie) can hold his own with semantic formalisms
01:52:59 <jonb> sbp: yes
01:53:51 <sbp> because I wonder if CWM has enough power to go the other way: converting XQ espressions into logic, and then applying them to an RDF representation of an infoset instance. TimBL chatting about it earlier sparked me off
01:54:00 <sandro> I'm not knocking XQ for dealing with XML, but it seems like a lot of machinery we dont need/want for RDF.
01:54:46 <jonb> err... its alot of machinery that you are going to get with your MS/IBM/Oracle/SAG database
01:55:10 <sandro> sbp, I'm sure "log:" it has the expressive power -- I doubt cwm would have satisfactory performance.
01:55:19 <jonb> and if you don't understand it you are going to get left on your rump
01:55:35 <sandro> But are you going to have it on your cell phone, jonb?
01:55:41 <sbp> lol
01:55:54 <jonb> sandro: think "datagrid"
01:56:27 <sandro> the cell phone is just a dumb terminal?
01:56:37 <sbp> it's funny how "converging" necessarily follows "emerging" when talking about technology
01:56:46 <jonb> well a terminal
01:56:55 <sandro> (Of course, a decent 1960's hacker could fit XQ in the wasted space on a cell phone, but still.)
01:57:56 <sandro> I want all the clients/servers/peers/agents to speak the same protocol, personally.
01:58:13 <jonb> go for it :-)
01:58:22 <sandro> (or protocols which are instances of the same abstract protocol, at least, for legacy and performance reasons.)
01:58:27 <jonb> you mean XML over HTTP :-)
01:58:54 <sbp> XMLi over HTTP
01:59:08 <sandro> I prefer S-Triples over UDP, myself, but XML over HTTP is okay, too, if you've got the horses.
01:59:33 <jonb> just so we are clear about who is building the clients/servers/peers/agents
02:00:03 <sbp> not too difficult to map between S-Triples and an infoset representation. Hey, then you might have an RDF description of the S-Triples format
02:00:12 <sandro> Hmmm?  Is that a "Do what you want, Sandro, but the industry will be using XQ" ...?
02:00:28 <jonb> and the cell phones speak WAP today
02:00:37 <bijan> No, this is: Do what you want, Sandro, but the industry will be using XQ!
02:00:56 <bijan> :)
02:01:26 <jonb> sandro: no its: i have other battles to fight
02:02:02 <jonb> i would have been perfectly happy sending s-expressions over TCP if that were the case
02:02:18 <sandro> Yeah, me too -- XQ+XML is just some extra layers of complexity; they can be kept inside the box.
02:02:19 <bijan> ooo, paradise!
02:02:43 <bijan> let me note that there is an active thread on comp.lang.lisp entitled, "The Horror that is XML" :)
02:03:14 <sandro> I tend to think UDP makes more sense, since TCP wont give you real agent-agent reliability anyway -- but I don't really have the data.   Just a guess.
02:03:14 <jonb> that said, since everyone has gotten excited about XML I try to familiarize myself with the details
02:04:14 <sandro> And XML is just LISP with the some redundant syntax:   (a b c) ->  <a><b/><c/></a>       (mostly)
02:04:28 <jonb> my goal is to get people to put reliable data on the web
02:04:38 <bijan> That's easy.
02:04:42 <sandro> That's a good goal, Jon.
02:04:44 <bijan> Oh, you meant *only* reliable data? Too bad.
02:04:46 <bijan> :)
02:04:50 <jonb> sandro: yes, but people have good reasons for the redundancy
02:05:06 <bijan> Reliable data clearly distinguished from crap? Damn!
02:05:10 <sandro> Agreed, re: good reasons for redunancy.    
02:05:24 <jonb> first: the start and end tags facilitate regular expression processing
02:05:27 <bijan> * bijan doesn't agree :)
02:05:34 <sandro> cat /dev/random | PUT http://www.w3.org
02:05:59 <sbp> argh, where's CygBot when I need it?
02:06:07 <AaronSw> heh!
02:06:09 <sandro> The attributes confuse things a bit.
02:06:29 <jonb> second: the end tag -apparently- makes human proofreading, and error correction, easier
02:06:40 <bijan> Hmm. I don't think you need to argue it jonb. 1) I don't think you can correctly, 2) even if you can, you won't convince most of the doubters. Ok, me :)
02:06:51 <jonb> yes attributes are confusing
02:06:58 <bijan> 3) Ubiquity is prolly sufficient for endorsing XML :)
02:07:09 <sandro> Do you have a clean mapping of attributes to s-exprs?
02:07:14 <tim-onNoff> I think you need different kinds of hashing in different circumstances.
02:07:24 <jonb> bijan: 3) yes, and 1) and 2) are the reasons given to me my some XML hands
02:07:31 <bijan> XSML
02:07:44 <bijan> The best (current) mapping from XML to sexpr.
02:07:46 <jonb> tim: yes, very much so
02:07:50 <bijan> Indeed, to executable scheme code.
02:07:59 <bijan> SXSLT and SXPath as well.
02:08:03 <bijan> Truely lovely.
02:08:19 <jonb> bijan: DSSSL is XSLT in Scheme eh?
02:08:26 <bijan> jonb, I believe those are reasons given, I just don't find them *remotely* plausible.
02:08:42 <bijan> Eh.
02:08:52 <bijan> but lacking literal element syntax.
02:08:58 <sandro> gotta go.   g'night all.
02:08:59 <bijan> What's nice about XSML is you preserve that.
02:09:11 <jonb> bye.
02:09:12 <bijan> I.e., you don't have to make-element all the time.
02:09:20 <jonb> literal element syntax?
02:09:23 <bijan> Which admittedly is less painful in DSSSL than in XSLT, but stil.
02:09:47 <bijan> <b>hi there</b> vs. <xsl:element...(whatever)
02:10:07 <jonb> XSLT allows both?
02:10:13 <sbp> huh? you can have that in XSLT
02:10:16 <bijan> Er..yes.
02:10:25 <bijan> that's what I'm saying.
02:10:41 <bijan> XSLT beats DSSSL becasue DSSSL (afaik) doesn't have literal element syntax.
02:10:54 <jonb> did you see my XML representation for KIF and N3 (as an extension to RDF)?
02:10:55 <bijan> DSSSL's non-literal constructors are less painful than XSLTs
02:10:59 <sbp> ah, you meant "more painful" then, I guess
02:11:10 <bijan> SXSLT has both.
02:11:12 <bijan> No, I didn't
02:11:17 <bijan> I.e., you don't have to make-element all the time.
02:11:29 <bijan> Which (using make-element all the time) is less painful in DSSSL than in XSLT, but stil.
02:11:31 <sbp> jon: nope?
02:11:36 <jonb> http://www.openhealth.org/RDF/RDFxtAbstractSyntax
02:11:36 <dc_rdfig> G: http://www.openhealth.org/RDF/RDFxtAbstractSyntax from jonb
02:11:40 <bijan> John, no.
02:11:52 <bijan> Hmm. an xml syntax for n3.
02:12:07 <danbri> * danbri catches up 
02:12:07 <jonb> G:| XML representation for N3,KIF ... work in progress
02:12:08 <dc_rdfig> titled item G
02:12:20 <sbp> Bundle == Formula? Context?
02:12:31 <jonb> yeah basically
02:12:43 <tim-onNoff> Sounds intersting... animprovement on rdf:parsetype="log:quote" ?!
02:12:47 <danbri> So the syntactic web thing, libby, me and the RQL guys were contributors... J. Robie's talking about doing a successor/followup paper
02:12:52 <sbp> heh, heh
02:13:11 <jonb> danbri: cool work
02:13:15 <danbri> The querying RDF with XQuery wasn't rocket science, and we didn't all agree with him about it scaling etc.
02:13:16 <tim-onNoff> Formula is the noun, content is the property of a statement which is the formula it i in.
02:13:30 <sbp> nice production: "the world"
02:13:51 <bijan> content? or context?
02:13:53 <danbri> Basically, imagine NTriples in XML, and applying XML query to that. Yeah it works, but it isn't pretty and doesn't really create an environment where optimisation is possible
02:14:03 <bijan> it i[s] in?
02:14:16 <bijan> Oops. gotta run.
02:14:25 <tim-onNoff> bijan, w.r.t. your earlier request,    n3:context  rdfs:domain rdf:statement; rdfs:range  n3:Formula. 
02:14:28 <danbri> Xquery processors won't know about all the things that RDF queries will exploit to do faster queries...
02:14:53 <sbp> I still consider Context as a class to be like Formula, except that a Context with a single triple in it is not a superclass of a statement (it is a set of one statement)
02:15:11 <tim-onNoff> I don't.
02:15:13 <sbp> consider as in "that's what I tell people until someone argues with me" :-)
02:15:26 <danbri> * danbri gives up trying to get Ruby binding for Redland working. I got Swig running OK, but I need all that autoconf/automake stuff to integrate it w/ the redland build, and I don't understand how it works! 
02:15:30 <tim-onNoff> I have been trying to root out all uses of Context as a class.
02:15:50 <sbp> so :Context is a non-class? did it *ever* mean what I thought it to mean?
02:15:51 <danbri> sbp, do you know (you seem to know these things...) where the user manual for timbl's old next www browser can be found? 
02:15:58 <danbri> (was it the doc you translated?)
02:16:19 <sbp> yep: http://infomesh.net/2001/enquire/manual/
02:16:45 <danbri> shellac an me were looking for info on the contents of the menubars, since that's the main hurdle to getting it to compile again under OpenStep, and maybe macosx...
02:16:49 <danbri> cool, thanks!
02:17:02 <danbri> tim said he'd help but that it was in the manual mostly
02:17:38 <jonb> tim: is a formula simply an unasserted statement?
02:17:40 <sbp> echo ":Context a rdfs:Class ." | /dev/null
02:18:14 <tim-onNoff> sbp, I did use it as a class, but I realized from the flak I got here and elsewhere that that wan'tr good, and I also realized Formula actually was a word i wanet dto use.
02:18:47 <tim-onNoff> No, that's the enquire manual.  There is a WorldWideWeb manual somewhere else..
02:19:32 <sbp> WorldWideWeb? erk... in /History/ somewhere?
02:19:45 <sbp> Jon, try http://lists.w3.org/Archives/Public/www-rdf-logic/2001Dec/0031 that Sandro chumped earlier
02:20:20 <sbp> looks promising: http://www.w3.org/History/1991-WWW-NeXT/WorldWideWeb
02:20:26 <tim-onNoff> Maybe in the 1992 smapshot
02:20:42 <tim-onNoff> http://www.w3.org/History/1991-WWW-NeXT/Menus.html
02:20:42 <dc_rdfig> H: http://www.w3.org/History/1991-WWW-NeXT/Menus.html from tim-onNoff
02:20:46 <tim-onNoff> ooops
02:21:03 <danbri> not oops, the chump's my way of finding RDF stuff again :)
02:21:23 <danbri> H:|Info on the menus in TimBL's NeXT WWW browser
02:21:24 <dc_rdfig> titled item H
02:21:47 <danbri> * danbri wonders if there's a working copy anyway (still got that NeXT box tim?)
02:22:07 <tim-onNoff> H: Note "next" and "previous" effect - not what you might expect but quite useful all the same
02:22:08 <dc_rdfig> added comment H1
02:22:48 <danbri> excellent, just what we were looking for :)
02:23:27 <jonb> danbri: yes the XQuery/RDF stuff was hacked (J.R. showed me the code) but still, it is the type of integration that is very useful
02:23:37 <sbp> some stuff in: http://www.w3.org/History/19921103-hypertext/hypertext/WWW/NeXT/Implementation/
02:24:18 <tim-onNoff> H:This quits the application as is usual on the NeXT, except that currently (Nov90) it DOES NOT CHECK FOR UNSAVED WINDOWS ;-)
02:24:18 <dc_rdfig> added comment H2
02:24:49 <tim-onNoff> I don't have the next box, CERN has it.
02:25:17 <sbp> I thought Aaron saw it in the London science museum...
02:25:25 <sbp> I'm sure he took a photo. hang on a sec
02:25:29 <AaronSw> I did indeed.
02:25:40 <sbp> aha: http://www.aaronsw.com/photos/uk2001/IMG_1119.JPG/view
02:25:45 <danbri> * danbri thouht I'd seen it in your office (but I've possibly never seen a NeXT box, if it wasn't, so couldn't have recognised it...)
02:26:00 <AaronSw> I was going to try and install some of the old code because we just got a next cube from a friend.
02:27:06 <mnot> cool
02:27:30 <tim-onNoff> Might be in the science museum, now you mention it.
02:28:10 <tim-onNoff> It spent some time as a server before it stopped altogether, so the desktop may have deteriorated.  I know the screen did.
02:28:19 <danbri> * danbri tries to remember how to fire up GNUStep
02:28:25 <mnot> I interviewed in '96 with a company in Melbourne that did *all* of their heavy lifting on NeXT boxes.. they had a glass-walled machine room FULL of the things (tens if not hundreds)... was very cool. I wonder if they still have them...
02:28:32 <tim-onNoff> I have a next cube in my office, one donated from someone in LCS.
02:29:35 <tim-onNoff> Rob Blessing (sp?) runs a next shop and refitted the one in my office. He sent me a photo of a customer's cube polished to a shine ... quite a cusom job.
02:30:43 <danbri> could've got ericP to paint it instead, per custom paint job on his car...
02:31:04 <danbri> http://www.rubyide.org/cgi-bin/wiki.pl?FreeRIDE_Projects
02:31:05 <dc_rdfig> I: http://www.rubyide.org/cgi-bin/wiki.pl?FreeRIDE_Projects from danbri
02:31:47 <danbri> I:|FreeRide (Ruby IDE), chumped for its connection to the software/api metadata issue (Rdoc, RAA etc)
02:31:47 <dc_rdfig> titled item I
02:32:03 <mnot> heh
02:32:05 <danbri> I:And because the ruby channel doesn't have cool bots like this one...
02:32:05 <dc_rdfig> added comment I1
02:33:54 <sbp> incedentally, the oldest HTML doc on the WWW that I know of: http://www.w3.org/History/19921103-hypertext/hypertext/WWW/Link.html
02:33:57 <danbri> I:The [ttp://www.rubyide.org/cgi-bin/wiki.pl?DevDoc/DataBus|data bus] doc is probably the point of connection for metadata apps. "A databus is nothing more than  a hierarchical collection slots. Each slot is a container that can hold some data and is accessed (or addressed) by its path. A path should look very familiar as it is a direct analogue of a filesystem path." 
02:33:57 <dc_rdfig> added comment I2
02:37:00 <tim-onNoff> sbp, hmmm ... you mean the last modified date Last-Modified: Tue, 13 Nov 1990 15:17:00 GMT?
02:37:10 <danbri> I:Looks like tuplespaces stuff to me, interesting... Also "A Properties class sits on top of a specific section of a DataBus tree and offers XML persistence of that section of the tree.    
02:37:10 <dc_rdfig> added comment I3
02:37:35 <tim-onNoff> from http://www.w3.org/History/19921103-hypertext/hypertext/WWW/Link.html,headers
02:39:20 <tim-onNoff> Hmmm.  just realized a connection between 2 things discussed today -- the URIs used by dbview to encode state in a databse and a syntax for {python,...} do interesting langauge things with triple stores. 
02:40:07 <sbp> yep. I found out from locally searching through hype.tar
02:41:54 <tim-onNoff> * tim-onNoff gotta go
02:42:00 <tim-onNoff> tim-onNoff is now known as tim-gone
02:48:11 <danbri> BTW Dean Jackson (w3 SVG guru) is going to stop by here tommorrow sometime (UK pm) to chat about SVG and SemWeb stuff, following on some brainstorming from last week (france f2f and from irc)...
02:48:29 <sbp> great
02:48:45 <sbp> s/great/great!/
02:49:14 <sbp> Hmm... where'd those notes go on linerizing SVG with RDF
02:49:39 <danbri> I want to get an HTML imagemap to SVG/RDF image annotation tool working... so we can say to the folk who maintain the 100s of imagemap editors "tweak it thusly and you're SemWeb ready..."...
02:49:46 <danbri> google
02:50:04 <danbri> where everything goes!
02:50:39 <jonb> i don't quite follow -- how would that work?
02:50:48 <sbp> ah. simple hack: http://lists.w3.org/Archives/Public/www-archive/2002Feb/0024
02:51:14 <sbp> imagemap to SVG/RDF: with XSLT also, I guess
02:51:41 <jonb> cool
02:52:42 <sbp> * sbp likes the various "don't reinvent the wheel!" approaches to stuff being taken today
02:53:05 <danbri> Max F. (w3 xslt guru) made one last week. I'm doing one in Ruby too, cos I want it to have a bit more smarts than I can code in XSLT, and take params so I can say to a tool "make an SVG of just the W3C employees in this picture".
02:53:45 <danbri> ie. we parse the html imagemap, but then probably want a variety of options w.r.t. quite what exactly gets pumped out the other end
02:53:50 <danbri> Did you see Max's XLST?
02:54:38 <sbp> nope...
02:54:46 <sbp> d'ya have a ref.?
02:54:50 <danbri> http://www.w3.org/2001/08/rdfweb/foafwho/imagemap/
02:54:51 <dc_rdfig> J: http://www.w3.org/2001/08/rdfweb/foafwho/imagemap/ from danbri
02:55:27 <danbri> J:|Tools for transforming HTML imagemaps into SVG and/or RDF, in progress. XSLT from Max F., Ruby from DanBri.
02:55:27 <dc_rdfig> titled item J
02:56:05 <sbp> so much stuff to keep track of!
02:57:26 <sbp> Hmm... two totally unrelated, and yet both SVG/RDF, projects
02:57:39 <jonb> 'nite
02:57:49 <sbp> c'ya
02:58:05 <jonb> jonb has quit
03:00:01 <danbri> J:|Example: input [http://www.w3.org/2001/08/rdfweb/foafwho/imagemap/gimp-w34.imagemap.txt|imagemap from Gimp tool] (pre-XML HTML fragment); [http://www.w3.org/2001/08/rdfweb/foafwho/imagemap/gimp-w34.imagemap.html|tidied HTML version] (note the mailboxes in the HREF's for each person; neat trick huh?); [http://www.w3.org/2001/08/rdfweb/foafwho/imagemap/imagemap2svg.xslt|the xslt transform], and [http://www.w3.org/2001/08/rdfweb/foafwho/imagemap/
03:00:02 <danbri> _imagemap2svg-output.svg|SVG output] ([http://www.w3.org/2001/08/rdfweb/foafwho/imagemap/_imagemap2svg-output.png|PNG rasterization] using [http://www.w3.org/2001/08/rdfweb/foafwho/|batik and some scripts]). Nearby: [http://www.w3.org/2001/08/rdfweb/foafwho/imagemap/imagemap2meta.rb|draft Ruby script].          
03:00:02 <dc_rdfig> titled item J
03:01:10 <danbri> J:...[http://www.w3.org/2001/08/rdfweb/foafwho/imagemap/_imagemap2svg-output.svg|SVG output] ([http://www.w3.org/2001/08/rdfweb/foafwho/imagemap/_imagemap2svg-output.png|PNG rasterization] using [http://www.w3.org/2001/08/rdfweb/foafwho/|batik and some scripts]). Nearby: [http://www.w3.org/2001/08/rdfweb/foafwho/imagemap/imagemap2meta.rb|draft Ruby script].          
03:01:11 <dc_rdfig> added comment J1
03:01:19 <danbri> two unrelated SVG/RDF projects?
03:03:34 <sbp> well, the other one is for lineraizing SVG, and giving the relationships between the various parts
03:03:53 <sbp> just allowing people to tab through would be quite a success. Waypoints
03:06:34 <danbri> they're not so unconnected. All my SVG meets RDF experiments are due to Chaals' good influence, the SVG Accessibility Note etc., and the FOAF/RDFWeb/etc discussions we've had. The linearizer stuff I heard about was via WAI and DanielD, so not too distant...
03:07:54 <danbri> cf http://www.w3.org/WAI/ER/IG/minutes/20000511.html "RDF tools discussion led by Charles":
03:07:55 <sbp> yep
03:08:08 <danbri> .... /* charles commences handwaving session on RDF.  invisible partner Dan
03:08:08 <danbri> Brickley is not here */
03:08:13 <danbri> * danbri grins
03:08:20 <sbp> heh, heh
03:08:39 <danbri> we've been geeking on about WAI meets SemWeb since we met, is really nice seeing some of it coming together...
03:10:46 <danbri> * danbri -> zzz
03:11:04 <sbp> c'ya, danbri
03:13:53 <danbri> ftp://ftp.musicbrainz.org/pub/musicbrainz/musicbrainz-1.1.0.tar.gz   
03:14:13 <danbri> BLURB:MusicBrainz client library 1.1.0 released  
03:14:13 <dc_rdfig> K: MusicBrainz client library 1.1.0 released   from danbri
03:15:19 <danbri> K:[ftp://ftp.musicbrainz.org/pub/musicbrainz/musicbrainz-1.1.0.tar.gz|musicbrainz 1.1.0 download], includes trmgen code for identifying content of MP3s, somewhat independent of encoding details. MB has a corresponding metadata lookup service, so you can go from  file to it's metadata. 
03:15:19 <dc_rdfig> added comment K1
03:15:36 <danbri> danbri is now known as danbri_really_nothere
03:50:31 <DanCon> DanCon has joined #rdfig
04:05:55 <KyleC> KyleC has quit
04:26:43 <connolly> connolly has joined #rdfig
04:43:16 <DanCon> DanCon has quit
04:53:24 <xena> xena has quit
04:54:14 <xena> xena has joined #rdfig
05:08:16 <xena> xena has quit
05:09:28 <xena> xena has joined #rdfig
05:54:02 <dmiles> is there anyone here that has time to help me with a CVS import?
05:54:23 <dmiles> i have a src tree i am trying to add 
05:54:47 <dmiles> i cant beleive i am having such a duifficult time
05:55:05 <dmiles> know i have done this before ;P
05:55:36 <dmiles> all the docs i find haver to do with 3rd party syncing
05:56:17 <dmiles> they could have a doc that says.. "you have a source tree you wish to put into CVS.. this is howto.."
05:59:13 <tav> tav has quit
05:59:52 <tav> tav has joined #rdfig
05:59:55 <dmiles> cvs import -m "Import of Logic CVS Tree (Incomplete)" logicmoo dmiles logicmoo_0_6
05:59:55 <tav> tav has quit
06:00:14 <tav> tav has joined #rdfig
06:00:18 <dmiles> but i am not sure if itstarts in the pwd
06:00:30 <dmiles> i plann to be one directory aove
06:48:20 <dmiles> ok i have it in all in there .. 
06:48:23 <dmiles> :)
07:06:34 <xena> xena has quit
07:06:34 <MarkB> MarkB has quit
07:06:34 <dc_rdfig> dc_rdfig has quit
07:06:34 <sandro> sandro has quit
07:06:34 <mnot> mnot has quit
07:06:34 <lasDesk> lasDesk has quit
07:06:34 <dmiles> dmiles has quit
07:06:34 <grove> grove has quit
07:06:34 <sbp> sbp has quit
07:06:34 <deltab> deltab has quit
07:07:38 <xena> xena has joined #rdfig
07:07:38 <grove> grove has joined #rdfig
07:07:38 <dmiles> dmiles has joined #rdfig
07:07:38 <sbp> sbp has joined #rdfig
07:07:38 <lasDesk> lasDesk has joined #rdfig
07:07:38 <mnot> mnot has joined #rdfig
07:07:38 <deltab> deltab has joined #rdfig
07:07:38 <sandro> sandro has joined #rdfig
07:07:38 <dc_rdfig> dc_rdfig has joined #rdfig
07:07:38 <MarkB> MarkB has joined #rdfig
07:07:50 <tav> tav has quit
07:07:50 <tav> tav has joined #rdfig
07:08:19 <tim-gone> tim-gone has quit
07:08:36 <tim-gone> tim-gone has joined #rdfig
07:09:04 <connolly> connolly has quit
07:09:04 <danbri_really_nothere> danbri_really_nothere has quit
07:09:04 <larsbot> larsbot has quit
07:09:04 <jrn> jrn has quit
07:09:04 <Skyline__> Skyline__ has quit
07:09:29 <connolly> connolly has joined #rdfig
07:09:29 <danbri_really_nothere> danbri_really_nothere has joined #rdfig
07:09:29 <larsbot> larsbot has joined #rdfig
07:09:29 <jrn> jrn has joined #rdfig
07:09:29 <Skyline__> Skyline__ has joined #rdfig
07:15:40 <Skyline__> Skyline__ has quit
07:15:40 <jrn> jrn has quit
07:15:40 <danbri_really_nothere> danbri_really_nothere has quit
07:15:40 <larsbot> larsbot has quit
07:15:40 <connolly> connolly has quit
07:15:40 <tim-gone> tim-gone has quit
07:15:40 <MarkB> MarkB has quit
07:15:40 <dc_rdfig> dc_rdfig has quit
07:15:40 <sandro> sandro has quit
07:15:40 <mnot> mnot has quit
07:15:40 <lasDesk> lasDesk has quit
07:15:40 <dmiles> dmiles has quit
07:15:40 <deltab> deltab has quit
07:15:40 <grove> grove has quit
07:15:40 <sbp> sbp has quit
07:15:40 <xena> xena has quit
07:15:40 <tav> tav has quit
07:15:40 <em> em has quit
07:15:40 <bijan> bijan has quit
07:15:40 <deus_x> deus_x has quit
07:15:40 <jang> jang has quit
07:15:40 <AaronSw> AaronSw has quit
07:15:40 <tbl> tbl has quit
07:17:06 <Skyline__> Skyline__ has joined #rdfig
07:17:06 <jrn> jrn has joined #rdfig
07:17:06 <larsbot> larsbot has joined #rdfig
07:17:06 <danbri_really_nothere> danbri_really_nothere has joined #rdfig
07:17:06 <connolly> connolly has joined #rdfig
07:17:06 <tim-gone> tim-gone has joined #rdfig
07:17:06 <tav> tav has joined #rdfig
07:17:06 <MarkB> MarkB has joined #rdfig
07:17:06 <dc_rdfig> dc_rdfig has joined #rdfig
07:17:06 <sandro> sandro has joined #rdfig
07:17:06 <deltab> deltab has joined #rdfig
07:17:06 <mnot> mnot has joined #rdfig
07:17:06 <lasDesk> lasDesk has joined #rdfig
07:17:06 <sbp> sbp has joined #rdfig
07:17:06 <dmiles> dmiles has joined #rdfig
07:17:06 <grove> grove has joined #rdfig
07:17:06 <xena> xena has joined #rdfig
07:17:06 <em> em has joined #rdfig
07:17:06 <bijan> bijan has joined #rdfig
07:17:06 <deus_x> deus_x has joined #rdfig
07:17:06 <AaronSw> AaronSw has joined #rdfig
07:17:06 <jang> jang has joined #rdfig
07:17:06 <tbl> tbl has joined #rdfig
07:17:16 <ChanServ> [#rdfig] This channel is logged and blogged: http://logicerror.com/rdfIRCWelcome
07:17:16 <larsbot> larsbot has quit
07:17:16 <Skyline__> Skyline__ has quit
07:17:16 <danbri_really_nothere> danbri_really_nothere has quit
07:17:16 <jrn> jrn has quit
07:17:16 <connolly> connolly has quit
07:17:16 <tim-gone> tim-gone has quit
07:17:16 <lasDesk> lasDesk has quit
07:17:16 <mnot> mnot has quit
07:17:16 <sandro> sandro has quit
07:17:16 <dc_rdfig> dc_rdfig has quit
07:17:16 <MarkB> MarkB has quit
07:17:16 <grove> grove has quit
07:17:16 <dmiles> dmiles has quit
07:17:16 <xena> xena has quit
07:17:16 <sbp> sbp has quit
07:17:16 <deltab> deltab has quit
07:17:16 <bijan> bijan has quit
07:17:16 <em> em has quit
07:17:16 <jang> jang has quit
07:17:16 <deus_x> deus_x has quit
07:17:16 <tav> tav has quit
07:17:16 <AaronSw> AaronSw has quit
07:17:16 <tbl> tbl has quit
07:17:47 <Skyline__> Skyline__ has joined #rdfig
07:17:47 <jrn> jrn has joined #rdfig
07:17:47 <larsbot> larsbot has joined #rdfig
07:17:47 <danbri_really_nothere> danbri_really_nothere has joined #rdfig
07:17:47 <connolly> connolly has joined #rdfig
07:17:47 <tim-gone> tim-gone has joined #rdfig
07:17:47 <tav> tav has joined #rdfig
07:17:47 <MarkB> MarkB has joined #rdfig
07:17:47 <dc_rdfig> dc_rdfig has joined #rdfig
07:17:47 <sandro> sandro has joined #rdfig
07:17:47 <deltab> deltab has joined #rdfig
07:17:47 <mnot> mnot has joined #rdfig
07:17:47 <lasDesk> lasDesk has joined #rdfig
07:17:47 <sbp> sbp has joined #rdfig
07:17:47 <dmiles> dmiles has joined #rdfig
07:17:47 <grove> grove has joined #rdfig
07:17:47 <xena> xena has joined #rdfig
07:17:47 <em> em has joined #rdfig
07:17:47 <bijan> bijan has joined #rdfig
07:17:47 <deus_x> deus_x has joined #rdfig
07:17:47 <AaronSw> AaronSw has joined #rdfig
07:17:47 <jang> jang has joined #rdfig
07:17:47 <tbl> tbl has joined #rdfig
07:17:49 <ChanServ> [#rdfig] This channel is logged and blogged: http://logicerror.com/rdfIRCWelcome
07:20:50 <tansaku2> tansaku2 has joined #rdfig
08:16:12 <tansaku2> tansaku2 has quit
08:17:45 <larsbot> larsbot has quit
09:37:53 <libby> libby has joined #rdfig
09:38:24 <tbl> tbl has quit
09:38:30 <libby> anyone know what's up with the logs?
09:38:56 <libby> The requested URL /discovery/chatlogs/rdfig/2002-03-08.html was not found on this server.
09:43:02 <tansaku2> tansaku2 has joined #rdfig
09:51:35 <jang> looking now libby
09:52:07 <jang> * jang has finally figured out a coherent explanation of Dan's worries about Superman.
09:52:37 <dajobe> dajobe has joined #rdfig
09:53:16 <libby> hey jang
09:53:43 <libby> will someone with ops change the title for me too?
09:53:56 <dajobe> eh?
09:54:42 <dajobe> libby: re logs; today is the 7th
09:54:47 <libby> dean from w3c is comign along later to chat about possible semantic web projects he and a team could do
09:54:56 <libby> is it?
09:55:00 <dajobe> all day :)
09:55:20 <libby> my watch says 8th...
09:55:24 <jang> heh heh, and there was I diving in and looking for a technical solution
09:55:28 <libby> ok, guardian says 7th.
09:55:37 <libby> I've been out of sync for 7 days!
09:55:48 <libby> thanks jang, sorry about that!
09:56:18 <dajobe> * dajobe surprised libby trusts guardian to get it right :)
09:57:04 <libby> more trustworthy than a casio
09:57:57 <libby> so, will you chage the title dajobe?
09:58:32 <libby> something like: come and talk to Dean at 12 GMT about ideas for SWeb projects?
09:59:16 <dajobe> dajobe has changed the topic to: Web, Semantic Web and RDF chat - http://rdfig.xmlhack.com/ | Meeting with Dean at 12:00 UTC about ideas for Semantic Web projects
09:59:22 <dajobe> is that midnight or midday
09:59:24 <dajobe> * dajobe never can remember
09:59:34 <jang> midday
09:59:42 <jang> midnight = 00:00 or 24:00 (occasionally)
09:59:46 <dajobe> I meant; 12 GMT
09:59:49 <dajobe> sure
09:59:55 <libby> can you put - all ideas welcome or somesuch?
10:00:19 <dajobe> go ahead
10:00:24 <jang> can never remember if UTC or UCT is the DST equivalent f the other
10:00:51 <JibberJim> JibberJim has joined #rdfig
10:00:52 <dajobe> you can set the topic with /topic now
10:01:01 <libby> ok, here goes
10:03:29 <libby> libby has changed the topic to: Ideas for semantic web projects: meeting with Dean at 12:00 UTC - come and contribute | Web, Semantic Web and RDF chat - http://rdfig.xmlhack.com/
10:03:37 <libby> that sound ok?
10:03:51 <dajobe> sure, but who is Dean?
10:04:03 <libby> dean might be starting a SW project in Australia, and is looking for fun things to do
10:05:53 <libby> Dean Jackson: Dean is a W3C Fellow employed by CSIRO Australia. He is interested in Web Graphics and is a member of the  SVG  working group.
10:06:23 <libby> ...and a really cool guy
10:07:57 <dajobe> dajobe has changed the topic to: Cool Ideas for semantic web projects: meeting with Dean Jackson (http://www.w3.org/People/Dean/) at 12:00 UTC - come and contribute | Web, Semantic Web and RDF chat - http://rdfig.xmlhack.com/
10:08:14 <libby> heh, title getting longer and longer
10:08:16 <dajobe> I hope he doesn't mind his name in lights
10:08:17 <JibberJim> SVG?  If Dean wants an editor he can have mine.
10:08:26 <JibberJim> and make it work.
10:08:44 <libby> heh, he'll tell you, but I'm not sure he wants to concentrate on svg in this context
10:09:00 <libby> but you shoudl show him, cos its a very cool idea
10:09:02 <JibberJim> but it's RDF too...
10:09:14 <JibberJim> I'll be adding querying soon.
10:09:29 <libby> I wish I could see it on my mac :(
10:10:06 <libby> so I thought we could chump some ideas and discuss them, give him some pointers. 
10:10:27 <JibberJim> Mozilla 0.9.1 might be able to see some of it, but Mozilla have been so bad at changing their plugin interface there's no way Adobe can keep up.
10:10:55 <libby> I met him in France, he did some work with danbri and Max converting imagemaps to svg outlies, I think it was chumped by danbri
10:11:04 <libby> jim: bummer 
10:11:41 <JibberJim> Those outlines were quite good...
10:13:20 <libby> yeah, pretty neat stuff
10:16:01 <JibberJim> thankfully none of me though around.
10:17:12 <libby> no...? dont you fancy being in the codepiction database? everyone's doing it ;)
10:17:14 <dajobe> * dajobe just works out Jim is in UK
10:17:37 <JibberJim> Just doesn't fancy people having to look at pictures of me...
10:18:21 <libby> heh, well me neither, that's why I take all the pics :)
10:18:34 <dajobe> hehe
10:18:51 <JibberJim> * JibberJim is in London.
10:18:57 <JibberJim> * JibberJim is in London, England.
10:19:29 <libby> * libby better go to work...
10:19:51 <libby> see you later, thanks for clearning up my confusion  about the logs dajobe
10:19:56 <dajobe> lol
10:20:01 <libby> * libby back on everyone elses' time now
10:20:11 <libby> libby has quit
10:45:50 <JibberJim> Are there any RDFAuthor knowledgeable folk about?  I was wanting some info on how the querying works with more complicated queries than is in the tutorial.
10:52:34 <danbri_really_nothere> * danbri_really_nothere waves to all
10:54:18 <danbri_really_nothere> danbri_really_nothere is now known as danbri
10:55:00 <maxf> maxf has joined #rdfig
10:55:16 <danbri> Dave? A redland cfg question. How exactly does redland/librdf/tcl/Makefile.in come into the world?
10:55:26 <danbri> Hi Max!
10:55:29 <maxf> morning
10:55:36 <maxf> Libby just invited me
10:56:59 <danbri> * danbri realises he missed some changes to top level Makefile.am; fixing
10:57:01 <danbri> :)
10:57:55 <dajobe> * dajobe back
10:58:17 <dajobe> all Makefile.in come from Makefile.am via running automake
10:58:36 <maxf> danbri, I'm getting bounces from rdfweb-dev on yahoogroups. Didn't you say you'd subscribe me?
10:58:38 <dajobe> if you add a directory; you need to edit the end of configure.in
10:58:39 <danbri> I just grepped for tcl in top dir and put 'ruby' everywhere I found 'tcl'; missed a line though.
10:59:02 <danbri> YahooGroups was down the other day when I was going to do some list chores...
10:59:12 <dajobe> yeah, their raid5 blew up
10:59:17 <dajobe> and backup CPU was dead
11:00:31 <danbri> if it wasn't for spam, I'd be happy to have it open to non-listmembers posting. Time for whitelist based filtering maybe? like anyone "in the Web community" (a la w3c's list filtering) could do it
11:01:15 <maxf> I'm happy to subscribe. (though I'm not keen on giving out all the details they ask for an account)
11:10:02 <danbri> you don't have to give accurate details...
11:10:26 <jang> ...but the semantic web is happy to conclude all sorts of things if you give it contradictions :-)
11:10:41 <jang> we should be a double-act.
11:10:47 <danbri> :)
11:11:04 <maxf> maxf is now known as max-lunch
11:26:29 <neilJ> neilJ has joined #rdfig
11:54:44 <dean_> dean_ has joined #rdfig
11:56:16 <max-lunch> max-lunch is now known as maxf
11:57:48 <dean_> dean_ is now known as dino
12:06:25 <JibberJim> Hi Dean,
12:06:32 <dino> hi
12:06:48 <JibberJim> * JibberJim doesn't have any cool ideas for you, just thought he'd say hello.
12:07:01 <dino> * dino appreciates it :)
12:07:18 <dino> max is here also
12:07:32 <maxf> but doesn't have cool ideas either...
12:07:35 <dino> another recent semantic web convert
12:07:45 <JibberJim> I'm told you know SVG, you could take my RDF Editor and make it work... 
12:08:13 <dajobe> an SVG/RDF image annotaiton tool?
12:08:15 <dino> yeah max, good idea. :)
12:08:43 <dino> jokes aside, yes jim I've been meaning to take a look at it
12:08:54 <dino> I tried quickly once, but it didn't work in my browser.
12:09:29 <dino> or have I got it confused with something else? danbri gave me a url
12:09:33 <JibberJim> Nah, unfortunately I'm IE5.5 win specific in effect, 'cos of Mozillas weaknesses and Mac IE5's weaknesses.
12:09:54 <dino> right, ecmascript to plugins is a very dark art
12:10:04 <danbri> * danbri waves at Dean
12:10:08 <dino> hi dan
12:10:30 <dino> actually, the adobe svg plugin has a nice scripting engine built-in
12:10:39 <dino> but then you lose the html forms :(
12:11:02 <JibberJim> True, it has SeaMonket, but it doesn't give html forms, or more importantly the xmlhttprequest object which lets me go get chunks of RDF.
12:11:55 <dino> the svg viewer has a getURL and postURL method, that returns a DOMString. You should be able to use that to get the RDF
12:12:24 <JibberJim> I have an XML RDF parser not a String XML parser!
12:12:24 <dino> in fact, it also has a parseXML method which would give you a Document object, making RDF navigation a little easier
12:12:39 <JibberJim> It does?
12:13:07 <dino> yes, all Adobe specific, but so useful that every other viewer will implement them, regardless of what W3C says.
12:13:42 <dino> so does your XML parser take a Document?
12:13:55 <dino> I'll try to dig up some documentation for you.....
12:14:00 <maxf> (Adobe drives SVG anyway ;-)
12:14:16 <dino> s/drives/owns/
12:14:39 <JibberJim> which is annoying as it lessens the chance of MS building it native into IE.
12:15:02 <maxf> what difference does it make?
12:15:23 <JibberJim> Then it will be actually usable, people don't install plugins IMO.
12:15:55 <dino> it would be nice if MS preinstalled the adobe plugin
12:16:00 <dino> a la flash
12:16:10 <maxf> Hmm. Well they should. Death to monolithic browsers.
12:16:18 <maxf> they == people
12:16:23 <JibberJim> That would be okay too, although using the svg namespace like you can with their vml one would be nice too.
12:16:44 <dino> * dino agrees - mixed XHTML/SVG/SMIL would be cool
12:16:51 <maxf> MathML have solved that.
12:17:11 <dino> http://www.protocol7.com/svg-wiki/default.asp?AdobeSVGViewerWindow
12:17:12 <dc_rdfig> L: http://www.protocol7.com/svg-wiki/default.asp?AdobeSVGViewerWindow from dino
12:17:18 <dino> whoa?
12:17:26 <dino> hi chump
12:17:37 <maxf> MML use XSLT to send in-line MAthML to a plug-in.
12:17:40 <tansaku2> tansaku2 has quit
12:17:53 <JibberJim> it's not mentioned in http://www.adobe.com/svg/indepth/pdfs/CurrentSupport.pdf
12:18:03 <JibberJim> which is what I was using...
12:18:21 <dino> it is hidden deep within the window interface - the link I posted has better documentation
12:18:50 <dino> in fact it was written by adobe, which makes me wonder why they didn't put it into their own documentation ?!?!
12:19:06 <JibberJim> it doesn't allow cross domain :-(
12:19:14 <dino> so, how do I chump that link?
12:19:26 <dino> ahh, no.. bit of a pain.
12:19:48 <JibberJim> clientside semantic web stuff is made so difficult by that. IE at least gives it as a configuration option with a "prompt".
12:21:04 <dino> L| documentation on Adobe specific SVG interfaces
12:21:10 <dino> nope?
12:21:14 <JibberJim> : I think.
12:21:23 <dino> L: documentation on Adobe specific SVG interfaces
12:21:24 <dc_rdfig> added comment L1
12:21:27 <dino> thanks
12:21:47 <JibberJim> I think it's too aggressive about grabbing urls.
12:22:12 <dino> having said that, what is the URL of your RDF/SVG editor? :)
12:22:23 <JibberJim> - http://jibbering.com/rdf/editor.html
12:22:34 <dajobe> you need L:|title L:foo adds foo as a comment
12:23:10 <dino> L:|Adobe SVG Viewer ECMA script documentation
12:23:10 <dc_rdfig> titled item L
12:23:25 <dajobe> see the results at http://rdfig.xmlhack.com/
12:26:13 <dino> jim, it works for me now. Nice!
12:26:24 <libby> libby has joined #rdfig
12:26:32 <dino> hi libby
12:26:42 <maxf> hi
12:26:47 <libby> hia!
12:27:16 <libby> I was in a meeting....have people been chatting to you dean?
12:27:22 <dino> yes
12:27:31 <libby> good :)
12:27:35 <JibberJim> I started talking to him, 'cos I felt guilty no-one else was...
12:27:54 <libby> cheers jim :)
12:28:10 <libby> damn, maybe the SW community is out of ideas ;)
12:28:17 <libby> is aaron about?
12:28:18 <danbri> Dean, where exactly are you based? Is there a group of people doing Webby stuff?
12:28:35 <dino> i'm based in canberra, australia
12:28:37 <danbri> I didn't want to start on my "Dean how do I do X in SVG" wishlist _just_ yet... ;-)
12:29:03 <libby> heh
12:29:08 <maxf> oh yes. please...
12:29:12 <danbri> that's not too exact a reference...
12:29:20 <dino> I probably will have a bunch (1 < x < 5) of people to do webby stuff
12:29:29 <dino> + max of course :)
12:29:35 <libby> dean, we did talk about image annotation tools, like extending RDFPic, for example
12:29:48 <danbri> CSIRO... 
12:29:50 <dino> * dino wonders why dan wants X windows implemented in SVG
12:29:56 <libby> http://www.daml.org/2001/03/reference.html
12:29:57 <dc_rdfig> M: http://www.daml.org/2001/03/reference.html from libby
12:29:58 <danbri> * danbri visits http://www.cmis.csiro.au/
12:30:06 <dino> ahhh... ignore that.
12:30:16 <libby> M:|RDFPic image annotation tool
12:30:17 <dc_rdfig> titled item M
12:30:21 <dino> that page is written by a bunch of monkeys
12:30:22 <danbri> heheh, X windows... kinda.
12:30:32 <libby> M:embeds RDF in JPG images
12:30:33 <dc_rdfig> added comment M1
12:30:34 <danbri> :)
12:30:53 <danbri> Can I say what I'd love to see built? 
12:31:13 <libby> do it danbri!
12:31:19 <dino> so, we'd be a bunch of geeks, low on pure academic bg, but hopefully ok programmers
12:31:25 <dino> go ahead dan!
12:32:05 <danbri> Tools that show how to extend existing HTML imagemap editors, so that instead of doing trad 'clickable map' stuff, they help make Web descriptiosn of images more accessible, more machine processable, more managable using SVG, and (per original requirements) imagemappy as before
12:32:25 <dino> personally, I really like the photo-metadat stuff, mostly because I take a lot of crappy photos
12:32:38 <dino> I'd like to describe exactly how crappy they are
12:32:56 <danbri> There are *loads* of imagemap editors out there in tools, but they're all a bit crufty, and have tended towarsd making web content less rather than more accessible
12:33:11 <danbri> * danbri nods; mine are crappy too :)
12:33:22 <dino> oh yeah... i forgot to say that it appears everyone writes a photometa data tool
12:33:29 <maxf> so you want an editor that not only does imagemaps but also adds annotation to the page?
12:33:31 <dino> maybe it is hello world for SW?
12:33:50 <dino> that sounds doable
12:33:59 <dino> * dino checks that his irc client is logging
12:34:30 <JibberJim> * JibberJim reminds dino that there are public logs
12:34:32 <libby> that would be very cool
12:34:35 <danbri> So one of the various rdfweb/foaf spin off ideas was to do imagemap2svg overlays for image regions, and evangelise whatever technique comes out of that to tool creators. It potentially shows image metadata, SVG, HTML, XML, XSLT, RDF, RDF query, ontology/schema etc all working very nicely together, which is good for the W3C story that all these things are consistent.
12:34:49 <dino> * dino thanks Jim and has found the logs
12:34:52 <shellac> shellac has joined #rdfig
12:35:01 <bijan> One of the newest, most faddish drugs is "Web Services", which most of the largest corporations in the industry, including IBM, Microsoft, and Sun, are pushing vigorously. 
12:35:06 <bijan> * bijan giggles.
12:35:06 <dajobe> arn
12:35:14 <danbri> We have a bit of time in an upcoming EU semweb project to do related work, thouhg it isn't clear what things to do as Formal Project Deliverables, and what things to just hack on here with friends in evenings/weekends.
12:35:55 <logger> logger has joined #rdfig
12:35:55 <zahn.openprojects.net> topic is: Cool Ideas for semantic web projects: meeting with Dean Jackson (http://www.w3.org/People/Dean/) at 12:00 UTC - come and contribute | Web, Semantic Web and RDF chat - http://rdfig.xmlhack.com/
12:35:55 <zahn.openprojects.net> Users on #rdfig: logger shellac libby dino neilJ maxf JibberJim @dajobe jang AaronSw deus_x bijan em xena grove dmiles sbp lasDesk mnot deltab sandro dc_rdfig MarkB tav tim-gone connolly danbri jrn Skyline__ 
12:35:55 <ChanServ> [#rdfig] This channel is logged and blogged: http://logicerror.com/rdfIRCWelcome
12:36:07 <dino> which won't be far off the truth, even if i do them during my day :)
12:36:22 <JibberJim> It's always evening somewhere.
12:36:25 <danbri> Dean, what kind of projects did you have in mind? Are there things they'd need to achieve for your department? eg. -improving peoples skills, -getting peer reviewed papers published  -creating a software toolkit -fame/fortune -supprting consultancy -getting people PhDs etc. etc...
12:37:07 <JibberJim> You mean you can do that sort of stuff with SemWeb projects?
12:37:19 <danbri> nope, I'm just trying to let him down gently ;-)
12:37:29 <dino> it could be all of those eventually, but it has been described as a "long term, shut people in a room and don't bother them too much" thing
12:37:37 <danbri> best I can think of for fame/fortune is selling rdfweb t-shirts. And libby thouht of that!
12:37:54 <maxf> * maxf wants a t-shirt ;-)
12:37:58 <dino> * dino really wants one of those shirts, and is prepared to pay for it
12:38:07 <libby> * libby still has some spares to sell
12:38:09 <libby> :)
12:38:19 <bijan> * bijan would like one, and is prepared to let dino pay for it. :)
12:38:22 <libby> (a red one? sure)
12:38:24 <libby> heheh
12:38:33 <libby> I got 2 idaes for projects:
12:38:37 <danbri> would be better to make a new one with more people on it, SVG'd mugshots all morphed together with photoshop...
12:38:47 <shellac> tshirts - /the/ killer app of the semantic web 
12:39:02 <dino> danbri, I think the first goal would be to improve skills, and through that the research in our division can be improved.
12:39:06 <danbri> we could make a web service. Incoming messages via HTTP, outgoing via DHL.
12:39:07 <libby> 1) ranting about this in france: conference management tools via irc: querying and creating events, e.g. talks, bofs etc, and annotating them
12:39:21 <libby> ..could use soap to do the queries
12:39:35 <dino> yeah, you mentioned you ahd done some calendar stuff
12:40:10 <libby> 2) just talking with a collegaue: an ontology and tools for project aangement, rather likw what danbri has been doing for the europe project, but with more of a calendar perpective: who is overloaded with work, when are the deadlines etc
12:40:13 <danbri> Dino, what sort of people would likely be working on these projects? SVG Gurus? mathematicians? library/web/ai/etc. You run into such a mix of backgrounds in this field... (or is all that open/unconstrained?)
12:41:18 <dino> danbri, not known at this point, but probably a mix of people.
12:41:49 <danbri> BLURB:SemWeb project ideas from chat with Dean Jackson
12:41:50 <dc_rdfig> N: SemWeb project ideas from chat with Dean Jackson from danbri
12:41:53 <maxf> libby, a EU project managament would impress the EU ;)
12:42:00 <dino> I still appreciate the ideas - max and i will do stuff regardless of what csiro decides :)
12:42:22 <libby> heh, maybbe. but they expect us to use MS project and word I think
12:42:27 <dino> * dino has never had an IRC topic with his name in it... feels kinda invasive :)
12:42:42 <danbri> N:Here's a [http://ilrt.org/~ecdb/msc-projects/ilrt-projects.g.html|list of student projects] I wrote in 1997. Pre-RDF! 1st and 3rd are in SemWebby vein.
12:42:43 <dc_rdfig> added comment N1
12:43:12 <dajobe> dajobe has changed the topic to: Cool Ideas for semantic web projects around the world - come and contribute | Web, Semantic Web and RDF chat - http://rdfig.xmlhack.com/
12:43:47 <JibberJim> A topic in another channel is currently "* Par|wg looks at JibberJim, you always this annoying?"
12:44:02 <JibberJim> so you should feel pleased.
12:44:14 <dino> :)
12:44:44 <danbri> N:Other wishlist stuff: [http://rdfweb.org/people/danbri/2001/02/semanticweb-and-me/|'semantic web and me', passport renewal story]. It tails off towards the end when gets into details. Interesting use case though; dunno if Australian passport system run differently to UK's...
12:44:44 <dc_rdfig> added comment N2
12:46:04 <danbri> So I wrote http://rdfweb.org/people/danbri/2001/02/semanticweb-and-me/ after the UK passport office required me to prove what "Daniel Alexander Brickley" looked like, either by going back from Wales to Bristol to get a photo signed by a 'professioanl person', or else by waiting for them to fax a murky mugshot from their files in Belfast.
12:47:27 <danbri> There Has To Be A Better Way... So imho that's a nice scenario for a project as it connect image stuff (always fun) with the use of digital signatures etc. You could set a project up as "how could you use rdf/xml/etc to prove to a sceptical party that some picture was claimed to depict some named indivual"?
12:47:51 <dino> interesting stuff.
12:47:56 <dino> the project and the story :)
12:48:03 <danbri> There's a short answer (which makes it an achievable project): just PGP-sign a text file of a certain kind. But there are spin-off lines of enquiry everywhere:
12:48:19 <danbri>  - how do you know it's the right persons PGP pubkey (classic web of trust / key signing webs)
12:48:52 <danbri>  - how do you write rdf/xml that uniquely identifiies  people on the Web (via their homepage, mailbox, humanGenomeChecksum etc)
12:49:18 <danbri>  - how do you convince bureacrats that this stuff is deployable? (is it?: semweb maturity sanity  check)
12:49:30 <danbri>  - how do we use XML DSig instead of PGP? What does it buy us?
12:50:25 <danbri>  - how do we webify the passport office's notion of "a respectable figure from the community"? If 1000 XML hackers say that danbri.jpg depicts the person whose PGP pubkey is <mykey> is that enough?
12:50:29 <danbri> etc etc...
12:50:35 <danbri> I could go on... but I'm hogging the channel.
12:50:37 <danbri> * danbri shuts up now
12:50:40 <dino> * dino wonders what danbri actually does... not who he is, as Dean has placed is trust in whatever professional person vouched for dan's identity
12:50:51 <danbri> so does danbri :)
12:51:40 <maxf> dino, would it be students who'd work on that project at CSIRO?
12:51:43 <JibberJim> You seem to be wanting a higher standard than the passport office wants currently though.
12:51:43 <dino> danbri, could you flood the channel with shared calender thoughts?
12:51:57 <JibberJim> and then bookmarks?
12:52:11 <dino> maxf, could be (they are cheap :)
12:52:20 <maxf> what level?
12:52:35 <dino> honours or starting honours
12:52:46 <dino> +maxf of course :)
12:52:52 <danbri> calendars: the oughta be shared! events are such a *really* useful RDF modelling idion. They allow us to describe concerts, tv listings, recommendations/annotations/bookmarks, workflow process logs etc etc.
12:52:52 <maxf> what's that in uni years?
12:53:04 <dino> 4 years of uni
12:53:08 <maxf> (I'm not cheap)
12:53:14 <JibberJim> * JibberJim seems to be concluding that maxf is neither "people" or "student" and is beginning to wonder exactly what he is.
12:54:06 <maxf> * maxf was just at the wrong place at the wrong time in Cannes last week, with the wrong people ;-) 
12:54:22 <maxf> and wrongly knows about XSLT
12:54:26 <danbri> * danbri recently extracted a bio-blurb from a project proposal; what I do: http://rdfweb.org/people/danbri/mini-work-bio.html
12:54:30 <danbri> ish
12:55:44 <tansaku2> tansaku2 has joined #rdfig
12:55:46 <maxf> * maxf should write a bio page
12:55:59 <maxf> JibberJim, I'm W3C staff, doing XSL and MathML
12:56:08 <danbri> DanConnolly had a good rant on the need for shared calendars. Talked about Yahoo and all the other nifty calendar systems, and how one can invest so much time storing data in them, only for it to be locked up and not easily merged with other related data.
12:56:10 <maxf> and is somewhat interested in SW stuff
12:56:21 <danbri> wohoo, SWism spreads :)
12:56:25 <maxf> and I like Australia ;-)
12:56:30 <danbri> me too
12:56:33 <maxf> and Bristol!
12:56:37 <danbri> except for the flies and the hay fever
12:56:52 <JibberJim> Bristol sure has changed since I was last there
12:56:53 <danbri> me too, except for the rain and the hills
12:57:41 <libby> calendar stuff: also trying to get conferences marked up in RDF and shared
12:58:07 <libby> maybe using RSS 1.0 event module; I guy here is doing that (Martin Poulter)
12:58:08 <dino> the closest I've been to bristol is bath. It was really nice, except there was a lack of flies and hay fever.
12:58:18 <danbri> some good calendar tech work, evangelism and wide eyed idealism can be found in the SkiCAL project.
12:58:39 <danbri> dino, Rub honey in your eyes and you can fix both problems at once
12:59:01 <libby> skical, yep, conmbining calendar data about concerts and things with information about age limits, locations, costs etc
12:59:13 <danbri> libby, got a ptr to hand?
12:59:22 <libby> a ptr?
12:59:40 <danbri> pointer
12:59:46 <libby> ah
13:00:16 <maxf> not a Paper Tape Reader...
13:00:21 <libby> heh
13:00:40 <danbri> dean, http://lists.w3.org/Archives/Public/www-rdf-calendar/2001May/0041.html http://lists.w3.org/Archives/Public/www-rdf-calendar/2001May/0042.html and nearby talk about rdf calendar goals
13:00:43 <libby> http://www.skical.org
13:00:44 <dc_rdfig> O: http://www.skical.org from libby
13:01:17 <danbri> http://www.xml.com/pub/a/2001/07/25/rdfcalendar.html
13:01:17 <dc_rdfig> P: http://www.xml.com/pub/a/2001/07/25/rdfcalendar.html from danbri
13:01:29 <danbri> P:|XML.com piece on the RDF calendar taskforce
13:01:30 <dc_rdfig> titled item P
13:01:48 <libby> O: see also [the skical internet draft| http://www.ietf.org/internet-drafts/draft-many-ical-ski-05.txt]
13:01:48 <dc_rdfig> added comment O1
13:02:33 <maxf> skical don't seem to use RDF
13:02:37 <danbri> P:"RDF has been experiencing something of a renaissance lately"... (the author, Leigh Dodds, is at Ingenta nearby in Bath)
13:02:37 <dc_rdfig> added comment P1
13:03:22 <libby> O: and also [the RSS 1.0 events modeul | http://groups.yahoo.com/group/rss-dev/files/Modules/Proposed/mod_event.html]
13:03:23 <dc_rdfig> added comment O2
13:03:31 <danbri> skical didn't, but Greg's very interested. They were very involved in iCalendar, but wanted to do app specific extensions (notably for pub opening times and other critical metadata). So XML, namespaces and RDF as a convention for using XML-with-namespaces was a natural line of enquiry.
13:03:41 <dino> * dino is now on a W3C call but is still listening
13:03:52 <danbri> * danbri dials in too
13:03:58 <maxf> * maxf too
13:04:03 <libby> O: and also the [RDF calendar draft | http://groups.yahoo.com/group/rss-dev/files/Modules/Proposed/mod_event.html]
13:04:03 <dc_rdfig> added comment O3
13:04:11 <libby> * libby feels left out ;)
13:04:36 <maxf> libby, don't worry. This is much more interesting ;)
13:04:45 <libby> heh
13:05:30 <libby> problem with RDF calendar stuff at the moment is defining repeatign events, which are dead useful - TV scheules, opening times, but a bit tricky
13:05:40 <dino> why?
13:06:00 <libby> actually all the calendar stuff is rather hard, though that doesnt mean you cant do cool little demos with it
13:06:47 <libby> well it seems to me at any rate that a repeating event is rather like a rule, so encoding it as a fact in RDF is rather peculiar thing to do
13:07:47 <libby> skical: dont use RDF, although they have an xml format; greg Fitzpatrick and I are working on a DAML+oil version though, to see if that's useful
13:08:17 <danbri> No it isn't peculiar, lots of people are encoding rules explicitly in RDF. Encoding rule-fodder information (not the explicit formal logic of the rules) is quite reasonable to do.
13:08:49 <libby> you reckon?
13:09:08 <libby> well, we did, anyway, but the work Michael Arick and I did on that is untested
13:09:42 <danbri> repeating events are something you could just describe, eg by writing <RepeatingEvent cal:day="tues" cal:time="lunch"/>, but what we'd *ideally* want is for RDF query tools to eventually understand the repeatingness.
13:10:06 <libby> sure
13:10:20 <danbri> So that one could use generic RDF stuff and ask 'is there a meeting this tuesday', and it'd apply the rules it knew about RepeatingEvents to infer that there was a specific event on some specific tuesday.
13:10:38 <dino> is there a mecca of RDF link pages? ie. *the* definitive page
13:10:41 <libby> the issue of repeating events regardless of the rdf aspect is a very complex one: the icalendar group got a bit stuck on it, and are still tweking it
13:12:03 <libby> e.g. you want to say every monday in may unless its the 15th, this turns out to be rather hard to encode
13:12:06 <danbri> Interesting thing is that (i) we can deploy vocab that uses a <RepeatingEvent> construct now, without *needing* any fancy rules/inference stuff in actual tools (though we should use logic when creating our vocab and prose version of the rules in the spec). (ii) we can use good old fashioned procedural code to populate a database with <Event> instances based on a temploate <RepeatingEvent>: we don't need fancy code for this.
13:12:52 <danbri> It's hard to encode in DAML, esp if you have a 'Not', but a custom vocab can "carry that infro thru RDF", such that other tools (hand coded Perl etc) can make use of it.
13:12:58 <libby> ...although some of the fancier code now seems to be very good, e.g. Jos deRoo's Euler project
13:13:10 <libby> sure
13:13:17 <danbri> indeed :)
13:13:34 <libby> * libby likes good old fashioned querying though
13:14:15 <danbri> Another (related to the RSS connection) project idea that I'd love to see explored: an RSS Deployment Howto
13:14:29 <libby> I think the custom code thing is an interesting issue; daml have done it, but I find it very confusing in that case
13:14:48 <dino> danbri, good idea. I just did my first!
13:15:31 <libby> that would be cool: I dont think theree's much that tells you how to implement modules rather than straight RSS
13:15:50 <libby> MartinP ran into that problem with this events in RSS issue
13:16:18 <danbri> An RSS Deployment HOWTO would show (i) the use of XSLT and XHTML (as per w3c home page) to manage HTML in a way that allows RSS to be exposed. (ii) the use of extension namespaces to augment the channel with various kinds of additioanl metadata (iii) a quick howto on how to create your own extension namespace (iii) an example of accessing this extension data from RDF APIs and query languages (iv) an example of pulling two RDF/RSS feeds (extended 
13:16:19 <danbri> with other namespaces) from different places in the Web and querying.
13:16:41 <libby> yeah! do that
13:16:53 <maxf> with a list of software supporting RSS
13:16:58 <dino> * dino assigns danbri the ACTION item
13:17:27 <libby> we had an example here of people wanting to use RSS modules for teaching and learning resources, questionaires, as well as events
13:18:21 <danbri> (i) is done but not documented well in http://www.w3.org/2000/08/w3c-synd/, (ii) there are some articles on this (iii) & (iv) libby and I have a rough writeup and demo
13:18:30 <dino> actually, i tried looking for some software that supports rss today, but didn't find anything for windows that was as nice as Bert's KDE desktop ticker
13:18:37 <danbri> * danbri has to pay attention to W3C teleconf / back later
13:18:59 <dino> danbri is actually being mentioned by name in the teleconf
13:19:12 <dino> so i guess he had to say something :)
13:19:48 <dino> * dino notes that danbri is a lot louder on the phone than in person :)
13:20:10 <maxf> indeed! I hadn't realised it was him talking until I saw the irc log!
13:22:01 <libby> I asked about RSS/windows on the rss-dev channel, and nothing came up
13:22:09 <libby> - a real pity that
13:22:19 <JibberJim> what's wanted?
13:22:21 <dino> hmmm... makes me wish my laptop was unix
13:22:43 <maxf> * maxf hasn't found a good linux RSS client yet
13:22:55 <JibberJim> What does a good RSS client look like?
13:22:58 <bijan> RSS "client"?
13:23:02 <maxf> (except gnus, which I had to hack to make it work actually)
13:23:07 <libby> actually, what I'd really like to do is a windows RDF query thing. h'm maybe that wouldnt be so hard in vbscript now that we have a fairly generic SQuishQL->SQL rewriter
13:23:08 <maxf> RSS feed reader thing.
13:23:44 <libby> a desktop thing?
13:23:50 <dino> yes
13:23:54 <bijan> hmm. Shouldn't the various perl and python based readers work fine?
13:24:11 <dino> kde has a nice thing that sits in the taskbar
13:24:15 <bijan> Ah. Ye olde interface problem :)
13:24:15 <danbri> * danbri wonders if his headset is set too loud
13:24:22 <maxf> there must be plenty but I haven't looked hard enough I guess
13:24:31 <libby> rather than,say, somethign like danbri's http://ilrt.org/discovery/rdf/
13:24:32 <bijan> WHAT DANBRI? YOU CAN'T HEAR CLEARLY?!??!
13:24:39 <bijan> TRY TURNING UP YOUR HEADSET!
13:24:49 <danbri> http://ilrt.org/discovery/2000/11/rss-query/
13:24:50 <dc_rdfig> Q: http://ilrt.org/discovery/2000/11/rss-query/ from danbri
13:24:52 <JibberJim> RDF querying in windows shouldn't be hard at all, I already do it in javascript I just need to understand how a general purpose querying would look like.
13:25:08 <danbri> Q:|RDF: Extending and Querying RSS channels, by DanBri and Libby,
13:25:08 <dc_rdfig> titled item Q
13:25:19 <danbri> Q:My favourite half-finished writeup. Libby has a demo page somewhere too :)
13:25:19 <dc_rdfig> added comment Q1
13:25:27 <bijan> er..indeed, i query in SWI-Prolog just fine.
13:25:33 <bijan> CWM and eep both run on windows.
13:25:59 <JibberJim> There's a difference between runs on windows after you've installed lots of stuff, and runs on windows.
13:26:16 <bijan> Well, they both run after a python install.
13:26:18 <libby> sure, I'm sure a bunch of stuff; I have limited access to some MS access dbs,that's all, and am somewhat asp focussed when it comes to windows
13:26:27 <bijan> SWI-Prolog runs, after a SWI-Prolog install :)
13:26:56 <JibberJim> * JibberJim has his javascript xml/rdf parser in ASP aswell as on the client.
13:27:01 <libby> Q:[an RSS query demo|http://swordfish.rdfweb.org:8085/rdfquery/]
13:27:01 <bijan> none of these are killer installs :)
13:27:01 <dc_rdfig> added comment Q2
13:27:10 <libby> excellent
13:27:26 <bijan> Since windows doesn't come with RDF querying out of the box, you always have to install *something* :)
13:27:30 <libby> now with added working-ness after a query engine rewrite
13:27:58 <JibberJim> Windows comes with JScript and VBScript out of the box, you only need  a jscript or vbscript parser/query engine.
13:28:00 <libby> does any one query access RDF dbs on windows
13:28:06 <libby> ?
13:28:18 <bijan> Yes, you need a jscript or sbscript parser/query engine.
13:28:24 <JibberJim> Yes...
13:28:28 <JibberJim> I have one of those.
13:28:31 <libby> cool
13:28:50 <bijan> This is a worthless distinction, IMHO. If you want a Jscript or vbscript query engine, you should ask for that.
13:29:10 <JibberJim> Although the query engine doesn't accept general queries, you have to hack javascript source at the moment.
13:29:24 <bijan> If you want something that runs on windows, then I think eep, cwm, or cwmclone are perfectly reasonable responses.
13:30:19 <JibberJim> * JibberJim doesn't because they aren't going to be usable by your average windows user.
13:30:25 <libby> I was just thinking of a squish (or some other generic rdf query lanuage) rewriter to sql
13:30:39 <bijan> Why aren't they so usuable?
13:30:59 <libby> I had something like this that nicked and altered from Matt Bidulph; danbri cleaned it up and did it in ruby
13:31:11 <bijan> One could freeze eep, or cwm, and do something similar with swi prolog.
13:31:19 <JibberJim> installing python, perl and cvs etc... aren't things that your average windows user is going to be happy with.
13:31:25 <bijan> So that it boils down to installing a query app.
13:31:40 <bijan> CVS?
13:31:48 <JibberJim> to get cwm.
13:31:49 <libby> Q:related Squish->SQL stuff by danbri|http://rdfweb.org/2002/02/java/squish2sql/intro.html]
13:31:50 <dc_rdfig> added comment Q3
13:32:14 <bijan> There are tarballesque downloads.
13:32:32 <bijan> One could make even better ones.
13:32:47 <bijan> These are mere packaging issues.
13:32:54 <libby> for squish->SQL, all you'd need is a quick parser for the syntax and a thing that creates sha1s, and odbc.
13:33:12 <libby> asp scripter might be able to write their own
13:33:14 <bijan> It's far easier to pack the extant solutions up than to rewrite in jscript or vbscript.
13:33:42 <JibberJim> however jscript requires _no_ install even of the parser.
13:34:03 <dino> * dino notes that the link to RubyRDF on dan's page is broken
13:34:11 <libby> well, hopefully someone would write one
13:34:16 <libby> yeah, I just saw :(
13:34:19 <bijan> Fine, but no install is *hardly* a requirment to meet the needs of the "average" windows user.
13:34:39 <bijan> ...especially the average user of an RDF query engine :)
13:34:56 <dino> s/11/12/ to fix the link
13:35:01 <libby> * libby just thinks the SQl thing would be cool. the average windows user doessnt look at this stuff anyway
13:35:01 <JibberJim> For an RSS ticker on the desktop though, they aren't RDF querying.
13:35:22 <dino> i just want a simple ticker for now
13:35:23 <JibberJim> Libby - can you point me at what I need for squish to sql ?
13:35:29 <libby> sure....they might be if it was an extended module though, that's all
13:35:37 <dmiles> dmiles has quit
13:35:55 <libby> .http://rdfweb.org/2002/02/java/squish2sql/intro.html is danb's writeup
13:35:57 <bijan> * bijan lost what we're talking about, and no longer cares as requirements and descriptions seem to be fluctuating at will.
13:36:01 <xena> Inkling: Squish-to-SQL rewriter Inkling Squish-to-SQL query rewriter This tool takes a Squish query and transforms it into an SQL query targeted at a certain style of RDF database implemented in SQL. It was written by Libby Miller <libby.miller@bristol.ac.uk> based on a PHP version by Matt Biddulph <matt@picdiary.com>. See the brief howto document for internal details. Documentation The following
13:36:02 <xena>  may be of use... Java howto, usage guide etc. Libby's 'how it works' notes Inkling website Ruby-RDF now includes a Ruby rewrite of this code; see Ruby-RDF Squish overview SQL table structure and sample queries danbri@rdfweb.org
13:36:11 <libby> oops
13:36:22 <libby> weird
13:36:58 <libby> a simple ticker yu could write easily with a perl script, because of the way RSS is designed I think
13:37:12 <libby> i.e. RSS 1.0 has a canonnical XML syntax as well as being RDF
13:38:10 <libby> bijan, we're just thinking of things that dino/dean might be able to do - semantic web projects
13:38:28 <JibberJim> * JibberJim adds windows RSS ticker and squish2sql onto his todo list.
13:38:33 <libby> talking about stuff we wanted done but didnt have time to do
13:38:36 <libby> :)
13:38:43 <libby> ...or at least I was ;)
13:38:47 <danbri> http://www.w3.org/2001/12/rubyrdf/dataquery.html
13:38:48 <dc_rdfig> R: http://www.w3.org/2001/12/rubyrdf/dataquery.html from danbri
13:38:53 <dino> yes, send in your wishlists!
13:38:55 <bijan> Thanks for the explanation libby, but I still don't care anymore :)
13:39:18 <libby> aw, go on...dont you have anything you wanna build but dont have time?
13:39:35 <bijan> Sure. Lot's. I'd love to have minions.
13:39:39 <danbri> R:|Squish and data store support in Ruby, a transliteration of Libby's Java reimplementation of Matt's php squish2rdf script
13:39:41 <dc_rdfig> titled item R
13:39:58 <danbri> q+
13:40:04 <danbri> doh
13:40:10 <libby> SWeb needs more minions!
13:40:22 <bijan> Indeed.
13:40:29 <bijan> Well, I, *personally*, need them :)
13:40:32 <libby> I guess we all need to get our own minions
13:40:51 <libby> though dean is noones minion
13:41:45 <libby> ooh, shared bookmarks, there's a project that I got part of the way around to and never finished
13:42:16 <dino> is there a big difference between that and what chump is doing for you?
13:42:41 <libby> not really....and the chump does it much better that I could, but not eeryone wants t use irc
13:43:24 <libby> I did some work colevcting bookmarks files from people at work and turning them into RDF, but I dont have a url, and in any case wasnt v successful because of the privacy issue
13:43:36 <bijan> If they like the cmdlineness of the chump, it's easy enough to make a local version.
13:44:02 <dino> hmmm.. a simple windows interface to the chump could be easy
13:44:02 <bijan> If they want a more GUIy sort of thing, that's a bit different.
13:44:05 <libby> i.e. accidentally putting all your hot pr*n bookmarks into a communal database, indexed by your name 
13:44:15 <libby> esp when the bookmarks interfaces are crufty
13:45:00 <libby> I guess the people who've asked me about this were probably thinking of a webpage; they're not hackers, but librarian-types
13:45:28 <JibberJim> Could you not use an Annotea like method, and instead of using local bookmarks you use Annotea like-clients for them.
13:45:31 <dino> i'm always shaking my head at these weblogs that have a list of links down the side of the page
13:45:52 <dino> it seems only one step ahead of the page of links from 1994
13:45:57 <JibberJim> Modifying any of the Annotea client side code to talk slightly different RDF should be pretty simple.
13:46:02 <libby> heh, yep
13:46:24 <libby> annotea, that would be cool. To be honest I've never tried to install the server
13:46:43 <dino> a shared bookmark list would be nice, with or without the p*rn links.
13:47:25 <dmiles> dmiles has joined #rdfig
13:47:26 <JibberJim> If you make it I'll add support for it to my browser.
13:47:39 <dino> your browser?
13:47:49 <libby> I guess the key thing though is getting them back (not be url, but by other types of queries), and what those would be is as much an vocab design and inpputting care as anything else
13:47:58 <dino> * dino wonders if there is a short danbri-like bio for jibberjim
13:48:22 <JibberJim> jibbering.com/snufkin/ - an "browser" built on top of IE.
13:48:44 <JibberJim> and there's no bio...
13:49:51 <libby>  I did do a little search thing for the chump's rss feed, but my stuff doesnt do very good freetext searching, since it just uses postgres' free text stuff: http://swordfish.rdfweb.org/discovery/2001/04/ 
13:50:55 <libby> course there's all the mozilla bookmarks stuff too
13:51:19 <libby> oops, http://swordfish.rdfweb.org/discovery/2001/04/rdfig/ rather
13:58:17 <dino> thanks for the chat everyone. I'm going to have to pay more attention at my next telcon (starting in 3 minutes). I'll still be listening here if you have any more wishes/suggestions
13:58:41 <libby> thanks for coming along dean - let us know what happens :)
13:59:28 <dino> it was a pleasure
13:59:40 <libby> :)
14:09:18 <danbri> * danbri back, wondering where up to...
14:10:08 <danbri> oh, such timing! yes, thanks for coming. #rdfig is a 24x7 SemWeb meeting; maybe we should pick a timeslot for weekly brainstorm / project ideas session?
14:10:44 <libby> I do like the idea of using specific times fr tings on the channel: I was sligjhtly at a loss as to how to publicise it
14:10:59 <libby> - mayeb rdfig mailing list
14:11:04 <libby> ?
14:11:47 <maxf> good idea re. brainstorm
14:12:11 <libby> jim, are you going to Brighton in may for WAI meeting btw?
14:12:19 <danbri> www-rdf-interest, though there's a slight disconnect between this relatively small group of regualrs and the large list of readers on that list.
14:12:30 <danbri> I've neglected rdfig email in favour of RDFIG IRC lately.
14:12:41 <JibberJim> I intend to yep libby, although I don't normally plan that far ahead!
14:12:43 <danbri> I prefer the granularity...
14:12:44 <libby> * libby has a cunning plan to unmask mysteryJim and codepict him
14:13:09 <libby> maybe we need a mailing list for the irc channel ;)
14:13:29 <danbri> * danbri has a cunning plan to mask people's faces out with SVG clippaths, to preserve their sekret identities
14:13:45 <maxf> danbri, did you subscribe me to rdfweb-dev or should I do it?
14:13:47 <dajobe> we (not me!) should have more feedback, reports from irc -> mailing list
14:14:14 <danbri> Well lots of these projects (the ones I listed, I mean) are things I've wanted to do in RDFWeb/FOAF rather than "day job", so rdfweb-dev a fine place...
14:14:19 <danbri> maxf: not yet. doing now.
14:14:34 <libby> maybe if it was a regular meeting it would be a good thing to send to rdfig mailing list. would be terible if 300 people joined thogh
14:14:37 <libby> sure
14:14:41 <danbri> yes, I'd like to get a weekly report from the chump into email
14:14:49 <libby> did you subscribe shellac?
14:14:58 <maxf> danbri, thanks.
14:15:01 <danbri> If only we knew someone who was nifty with XSLT, we could compile weekly summaries of the RDFIG chump...
14:15:12 <maxf> if only...
14:15:18 <libby> ah, if only...
14:15:30 <danbri> got the tarball url to hand?
14:15:38 <libby> how would that work anyway? need a summarise funstion. macosX has that
14:15:45 <danbri> something like: http://rdfig.xmlhack.com/chump.tar.gz
14:15:53 <danbri> ...but not that. I'm always losing it...
14:16:28 <libby> I didnt know there was one - I harvest it from the site
14:16:30 <danbri> No, it would just take 7 days (and a prettier stylesheet) and make an HTML doc. Then Lynx or something html2text, crontabbed to mail
14:16:44 <danbri> the tar ball gets bigger each week...
14:16:56 <libby> oh, ok, sure, sounds good
14:17:05 <danbri> logger, grep chump tar archive snapshot
14:17:12 <logger> I'm logging.  I found 1 answer for 'chump tar archive snapshot'
14:17:13 <logger> 0) 2002-03-07 14:17:05 <danbri> logger, grep chump tar archive snapshot
14:17:22 <danbri> thanks a bunch, logger.
14:17:29 <libby> shellac was saing that often chump is most useful in conjunction with the logged discussions
14:17:30 <danbri> logger, grep chump archive snapshot
14:17:34 <logger> I'm logging.  I found 1 answer for 'chump archive snapshot'
14:17:35 <logger> 0) 2002-03-07 14:17:30 <danbri> logger, grep chump archive snapshot
14:17:36 <libby> heheh
14:17:36 <JibberJim> That was really helpful :-)
14:17:41 <libby> :))
14:17:41 <dajobe> .xena google chump tar archive snapshot
14:17:49 <dajobe> or something
14:18:02 <danbri> .google chump tar archive snapshot
14:18:03 <xena> chump tar archive snapshot: http://ilrt.org/discovery/chatlogs/rdfig/2001-10-04.txt
14:18:28 <libby> it would be very cool to link discussions back to the logs: I tend to see a chump and then go look at the discussion tat generated it
14:18:34 <danbri> yes re archives, often the chump is (like RSS) a 'site summary' and if you're intersted you think "hey, why were they talking about xyz"
14:18:37 <libby> what is xena anyway
14:18:41 <libby> ?
14:18:53 <dajobe> aaron's infobot
14:19:01 <libby> how does it work?
14:19:05 <danbri> don't think it's aarons. 
14:19:06 <maxf> * maxf has to turn his attention elsewhere for a bit
14:19:18 <libby> see ya max, thanks for coming :)
14:19:42 <danbri> I don't like Xena as currently packaged/deployed. It's intrusive and ill documented, so I complain about it hoping it'll understand and fix itself.
14:19:43 <maxf> back in a few minutes. Please summarise that XSLT thing for me then
14:19:47 <danbri> Yes, cheers maxf!
14:20:06 <danbri> http://rdfig.xmlhack.com/archive.tgz
14:20:07 <dc_rdfig> S: http://rdfig.xmlhack.com/archive.tgz from danbri
14:20:11 <libby> but what does it do!?
14:20:14 <maxf> maxf is now known as maxoff
14:20:41 <danbri> D:|The RDF Interest Group chump, archives of raw XML (snapshotted nightly).
14:20:41 <dc_rdfig> titled item D
14:21:28 <libby> raw, juicy xml. mmm
14:21:30 <danbri> D:It would be good to have XSLT that created a week's view of the discussion instead of a day's. I want to run such a thing in a crontab on w3.org mailing RDF Interest a summary of the chump.
14:21:31 <dc_rdfig> added comment D2
14:21:41 <connolly> connolly is now known as DanC
14:21:43 <danbri> <libby> but what does it do!?
14:21:57 <DanC> D:what I want is a month's worth of titles on a page.
14:21:58 <dc_rdfig> added comment D3
14:22:03 <DanC> D:... with links to details.
14:22:03 <dc_rdfig> added comment D4
14:22:10 <libby> xena I meant
14:22:47 <danbri> exactly! nobody quite sure. Has the usual bot tricks (searches google etc). Can bridge two IRC nets, which is handy (but also spooky, it can log stuff and/or beam to another channel in realtime, but doesn't clearly announce itself as doing so). 
14:22:56 <danbri> P3P statements for IRC bots, that's what we need...
14:23:00 <libby> cool!
14:26:43 <danbri> On the chump logs: find . -name '*.xml' -exec grep title {} \;|wc
14:26:52 <danbri>  1766 entries!
14:29:00 <libby> wow
14:32:00 <libby> I've got almost 11000 triples from that
14:33:16 <danbri> D:On the chump logs we find 1766 matches for: find . -name '*.xml' -exec grep title {} \;  (see [http://lists.w3.org/Archives/Public/www-archive/2002Mar/0018.html|list of titles])
14:33:16 <dc_rdfig> added comment D5
14:34:08 <danbri> odd, comments on D: not showing up on rdfig.xmlhack.com
14:34:30 <libby> would be nice to fix the stylesheet a bit as well to separate out the person who made the chump from the description in the rdf (I did some of the stylesheet so probably my fualt)
14:34:35 <libby> wasnt it S not D?
14:35:15 <danbri> Ah, so it was! oopsie.
14:35:24 <danbri> lucky nobody uses this data...
14:35:31 <libby> good job its the diting chump
14:35:34 <libby> editing
14:36:31 <danbri> S:On the chump logs we find 1766 matches for: find . -name '*.xml' -exec grep title {} \;  (see [http://lists.w3.org/Archives/Public/www-archive/2002Mar/0018.html|list of titles])
14:36:31 <dc_rdfig> added comment S1
14:36:48 <danbri> S:It would be good to have XSLT that created a week's view of the discussion instead of a day's. I want to run such a thing in a crontab on w3.org mailing RDF Interest a summary of the chump.
14:36:49 <dc_rdfig> added comment S2
14:36:51 <libby> oops, you overwrote the old title as well for D
14:37:16 <danbri> D:DanC says ":what I want is a month's worth of titles on a page ...with links to details"
14:37:16 <dc_rdfig> added comment D6
14:37:45 <danbri> S:|The RDF Interest Group chump, archives of raw XML (snapshotted nightly).
14:37:46 <dc_rdfig> titled item S
14:38:18 <libby> what you doing? :)
14:38:24 <libby> you did D again!
14:38:44 <tim-gone> tim-gone has quit
14:38:50 <danbri> * danbri blushes
14:38:56 <danbri> S:DanC says ":what I want is a month's worth of titles on a page ...with links to details"
14:38:56 <dc_rdfig> added comment S3
14:39:04 <libby> heheh. 
14:39:14 <dajobe> now D6:"" to delete it
14:39:30 <libby> and D5, D4....not D1 though
14:39:36 <danbri> D:
14:39:37 <dc_rdfig> http://www.ai.mit.edu/people/dquan/n3script.html
14:39:38 <dc_rdfig> The RDF Interest Group chump, archives of raw XML (snapshotted nightly).
14:39:39 <dc_rdfig> (1:sbp) A [http://www.python.org/|Python]/[http://www.w3.org/DesignIssues/Notation3|N3] convergence experiment by Dennis Quan
14:39:40 <dc_rdfig> (2:danbri) It would be good to have XSLT that created a week's view of the discussion instead of a day's. I want to run such a thing in a crontab on w3.org mailing RDF Interest a summary of the chump.
14:39:41 <dc_rdfig> (3:DanC) what I want is a month's worth of titles on a page.
14:39:42 <dc_rdfig> (4:DanC) ... with links to details.
14:39:43 <dc_rdfig> (5:danbri) On the chump logs we find 1766 matches for: find . -name '*.xml' -exec grep title {} \;  (see [http://lists.w3.org/Archives/Public/www-archive/2002Mar/0018.html|list of titles])
14:39:44 <dc_rdfig> (6:danbri) DanC says ":what I want is a month's worth of titles on a page ...with links to details"
14:40:08 <danbri> D6:""
14:40:09 <dc_rdfig> deleted comment D6
14:40:13 <danbri> D5:""
14:40:13 <dc_rdfig> deleted comment D5
14:40:15 <danbri> D4:""
14:40:17 <dc_rdfig> deleted comment D4
14:40:23 <danbri> D3:""
14:40:23 <dc_rdfig> deleted comment D3
14:40:39 <danbri> S:DanC "What I want is a month's worth of titles on a page"
14:40:40 <dc_rdfig> added comment S4
14:40:48 <danbri> D2:""
14:40:49 <dc_rdfig> deleted comment D2
14:40:55 <dajobe> it renumbers, danbri
14:41:13 <danbri> all fixed? the text log would allow the whole mess to be reconstructed, so this isn't changing history...
14:41:16 <dajobe> OK in this case, just FYI
14:41:16 <danbri> ugh
14:41:36 <danbri> where were we! Oh yeah, XSLT. Where'd Max go?! ;-)
14:41:42 <libby> just the title wrong now
14:42:42 <danbri> D:|Dennis' RDF work
14:42:42 <dc_rdfig> titled item D
14:42:50 <danbri> * danbri forgets real title!
14:51:21 <maxoff> maxoff is now known as maxf
14:52:34 <sat> sat has joined #rdfig
14:52:45 <danbri> http://lists.w3.org/Archives/Public/www-webont-wg/2002Feb/0144.html
14:52:45 <dc_rdfig> T: http://lists.w3.org/Archives/Public/www-webont-wg/2002Feb/0144.html from danbri
14:53:03 <danbri> T:|On Standardization of the Web Ontology Language - Trends &   Controversies Section
14:53:03 <dc_rdfig> titled item T
14:53:30 <danbri> T:Discussion of [http://www.aifb.uni-karlsruhe.de/~sst/is/WebOntologyLanguage/|IEEE Trends & Controversies, Issue March/April 2002], "On Standardization of the Web Ontology Language"
14:53:32 <dc_rdfig> added comment T1
14:54:30 <maxf> * maxf bbl
14:54:33 <danbri> T:Pat Hayes' piece is my favourite: [http://www.aifb.uni-karlsruhe.de/~sst/is/WebOntologyLanguage/hayes.htm|Catching the Dreams]
14:54:34 <dc_rdfig> added comment T2
14:54:35 <maxf> maxf has quit
14:55:24 <danbri> T:It (this paper) also includes a nice account of how you need to think if you want to grok description logics.
14:55:25 <dc_rdfig> added comment T3
14:55:27 <dajobe> * dajobe remembers to catchup with webont list
14:55:53 <danbri> T:instead of saying that a equals b (sorry,
14:55:53 <danbri> we can't let you use "equals" , that is far too dangerous),  
14:55:53 <dc_rdfig> added comment T4
14:57:41 <danbri> T4:""
14:57:41 <dc_rdfig> deleted comment T4
14:57:52 <danbri> * danbri finds cut/paste of text w/ line breaks too fiddly
15:04:03 <em> * em waves, allocates himself 5 min of to discuss 'rss 1.0' issue  
15:04:39 <em> anyone know of any rss 1.0 tools that support rdf:parseType="Literal" ? dajobe, does your stuff handle this?
15:04:56 <dajobe> yes, and yes (Since I do)
15:05:24 <em> whats the url for your validator?
15:05:28 <dajobe> however only rss 1.0 tools that are real RDF tools, do that
15:05:35 <em> * em wants to test a couple rss 1.0 feeds
15:05:35 <dajobe> since rss 1.0 spec doesn't use parsetype litarl
15:05:49 <dajobe> try http://www.redland.opensource.ac.uk/rss/
15:06:49 <em> yes, i understand... the issue seems to be more if you view this as a rdf vocabulary and read perhaps more into the spec than what was written
15:06:57 <em> thanks for url... testing
15:07:42 <em> -> http://www.redland.opensource.ac.uk/rss?uri=http%3A%2F%2Fwww.w3.org%2F2001%2Fsw%2FOverview.rss&parser=raptor&box=no&Go=Go
15:07:59 <em> hmm.... 
15:08:03 <dajobe> yeah, so I serialise the XML to a string
15:08:22 <dajobe> which currently you can't distinguish from a regular literal string
15:08:29 <dajobe> but that code is nearly done
15:09:53 <em> * em wonders if he's pushing the rss bar a bit too high by using parsetype=literal in his feeds
15:10:15 <dajobe> nah :)
15:10:19 <em> :)
15:10:34 <dajobe> you can always dumb it down to &lt;span&gt;  rubbish
15:10:39 <dajobe> yuk
15:10:40 <em> booo....
15:10:46 <em> any eta on your code?
15:11:04 <em> any other code you know supports this?
15:11:13 <dajobe> well, I'm coding ahead of the ntriples syntax changing
15:11:21 <dajobe> other code: dunno
15:11:35 <dajobe> and there is also c14n to bring in
15:11:41 <em> hmm... anyone? 
15:11:53 <dajobe> ARP I'm sure works the same way
15:11:54 <em> ntriples syntax changing more important...
15:11:58 <dajobe> try slinging it at the validator
15:12:04 <em> ARP yes... works fine
15:12:11 <em> thats what i'm using in Glance
15:12:58 <em> * em notes Arp isn't an option in dajobe's system
15:13:28 <dajobe> I have only *4* rdf/xml parsers you mean?
15:13:40 <em> :) yep.. thats what i mean!
15:13:56 <dajobe> * dajobe really should put that paypal button on his site
15:14:00 <em> * em wants a 5th ... *after* ntriples and rss 1.0 fixes :)
15:14:20 <em> * em goes to look for beerpay click... 
15:15:48 <em> * em notes times up... thanks dajobe and returns to his regularly scheduled day already in progress
15:18:23 <maxf> maxf has joined #rdfig
15:19:44 <maxf> back
15:20:44 <tim-gone> tim-gone has joined #rdfig
15:25:00 <AaronSw> aaaargh! em put html in is rss! 
15:25:21 <dajobe> it's been there for ages
15:25:35 <AaronSw> bad bad bad
15:25:45 <jang> why why why?
15:25:59 <AaronSw> He should use the RSS content module. The spec is very clear on this.
15:33:14 <AaronSw> * AaronSw wonders if mnot is around
15:38:44 <em> * em notices blue bar in xchat and bites...
15:39:00 <em> AaronSw, only now you've realized this... getting slow in your old age, eh?
15:39:09 <AaronSw> It's true.
15:39:25 <em> * em states again and wonders if he's pushing the rss bar a bit too high by using parsetype=literal in his feeds
15:39:34 <AaronSw> it's just plain illegal, em
15:39:43 <AaronSw> but you can put them in the content module
15:40:05 <em> are there example feeds that use the content module?
15:40:27 <AaronSw> don't you want to be the first?
15:40:33 <em> you asserting its plain illegal because.... #PCDATA restriction yes?
15:40:39 <AaronSw> Yes.
15:40:42 <AaronSw> http://purl.org/rss/1.0/modules/content/
15:40:42 <dc_rdfig> U: http://purl.org/rss/1.0/modules/content/ from AaronSw
15:40:54 <AaronSw> U:|RDF Site Summary 1.0 Modules: Content
15:40:54 <dc_rdfig> titled item U
15:41:04 <AaronSw> U:Uses parseType="Literal"
15:41:04 <dc_rdfig> added comment U1
15:43:51 <em> looks to see if dc:description has #PCDATA restrictions as well... 
15:44:00 <em> arrgg...
15:44:02 <em> yes
15:44:05 <AaronSw> tsk, tsk, em
15:44:08 <em> -> http://groups.yahoo.com/group/rss-dev/files/Modules/Standard/mod_dc.html
15:44:20 <AaronSw> heh, good thing we fixed up that hole
15:44:48 <em> yeah... good thing :(
15:45:07 <dc_rdfig> dc_rdfig has quit
15:45:35 <AaronSw> uh oh... someone go call edd
15:45:43 <dajobe> done
15:45:56 <em> aaron, dajobe... so are people escaping HTML in descriptions in order to ship this around?  
15:46:17 <em> instead of parseType=Literal
15:46:18 <dajobe> I think I've seen it
15:46:21 <AaronSw> * AaronSw thinks em is missing the point here...
15:46:30 <AaronSw> it's plain text for a reason!
15:46:30 <em> could be...
15:46:53 <em> and that reason is... backward compatability yes?
15:47:18 <em> ... simplicity yes?
15:47:19 <AaronSw> Umm, because plain text works in things other than browsers
15:47:40 <AaronSw> There are client-side RSS readers, you know.
15:47:53 <danbri> em, yes, don't go shoving angle brackets where they don't belong! There are clientside tools, even Mozilla XUL would barf.
15:48:12 <dino> dino has left #rdfig
15:48:32 <em> * em seems to recall XUL handled this just fine...
15:49:00 <las> las has joined #rdfig
15:49:06 <danbri> The RSS 1.0 view is: if you want to extend it, use XML namespaces (and the RDF model), rather than shove misunderstandable extensions into the main properties. That goes for markup as well as pseudo-markup...
15:50:02 <em> * em recalls that exact view as well
15:50:30 <em> * em and thought that was exactly what he was doing...
15:50:55 <em> but realizing there we're deadlines, perhaps again was a bit too hasty in design
15:51:25 <em> * em reminds himself to think more in terms of XML nexttime
15:51:52 <AaronSw> plain text goes in description, html-enhanced text goes in the content module
15:51:58 <rreck> rreck has joined #rdfig
15:54:02 <em> * em can't fix this today, but will look more into content modules.... looks like a bit of a rdf:parseType=literal overkill on first glance but will hold off judgement until i get a change to re-read this again
15:54:10 <rreck> can anyone point me to jena files, it looks like jena.sourceforge.net doesnt have any.
15:54:24 <AaronSw> em, yes, but we were sort of limited by rss requirements
15:54:50 <AaronSw> rreck, sf.net/projects/jena/ ?
15:55:07 <AaronSw> hmm, that has cvs at least
15:55:40 <shellac> IIRC http://www.hpl.hp.com/semweb/jena-top.html has a source 
15:55:48 <shellac> dists 
15:56:18 <rreck> thanks
15:56:52 <maxf> maxf has left #rdfig
15:57:57 <JibberJim> JibberJim has quit
15:58:01 <DanCon> DanCon has joined #rdfig
15:58:58 <JibberJim> JibberJim has joined #rdfig
16:14:35 <DanC> DanC has quit
16:15:28 <connolly> connolly has joined #rdfig
16:20:22 <timbl> timbl has joined #rdfig
16:27:17 <neilJ> neilJ has quit
16:30:53 <DanCon> DanCon has quit
16:34:41 <em> http://www.knowledgetechnologies.net/2002/schedule.htm
16:34:56 <dajobe> wait for chumpbot to return em
16:34:59 <em> hmm... where did mr'chump go?
16:35:11 <em> err... anyone planning on being at kt2002?
16:35:43 <dc_rdfig> dc_rdfig has joined #rdfig
16:36:02 <em> http://www.knowledgetechnologies.net/2002/schedule.htm
16:36:03 <dc_rdfig> V: http://www.knowledgetechnologies.net/2002/schedule.htm from em
16:36:15 <em> V: KT2002 Schedule
16:36:16 <dc_rdfig> added comment V1
16:36:41 <em> V:|[KT2002|http://www.knowledgetechnologies.net/] Schedule
16:36:41 <dc_rdfig> titled item V
16:37:02 <em> V:Who's going?
16:37:02 <dc_rdfig> added comment V2
16:38:15 <em> V:| Knowledge Technologies KT2002 Schedule
16:38:16 <dc_rdfig> titled item V
16:56:42 <dajobe> * dajobe writes another RDF/SW talk
17:04:28 <libby> libby has quit
17:05:43 <libby> libby has joined #rdfig
17:05:56 <connolly> connolly is now known as DanC
17:14:16 <jang> "I don't know if it's existential-conjunctive, but I know what I like"?
17:15:28 <libby> :)
17:23:50 <em> jang, dajobe... as primary rdf test-case owners, are you interested in having everything under /rdf-tests/rdfcore/ default acl set to public? This would eliminate the kinds of problems DanC just fixed in his most recent email to rdf-core.
17:24:43 <dajobe> yes, but I thought w3c system couldn't do that
17:24:48 <dajobe> s/system/acl system/
17:33:28 <jang> as far as I can see, there's nothing under there that need remain private.
17:34:15 <danbri> the system was upgraded; has notion of defaults for a directory now, when new stuff checked in.
17:34:25 <em> thanks jang, dajobe... we can indeed change acls to something other than public if appropriate, but default public has now been requested 
17:34:40 <em> dajobe.... err, yeah what danbri just said
17:35:40 <dajobe> nice: thanks all round em, danbri, w3t
17:36:01 <danbri> dom and ericp, I think
17:55:28 <bitsko> bitsko has joined #rdfig
17:58:55 <bitsko> bitsko has quit
18:09:49 <dajobe> dajobe has quit
18:26:16 <libby> libby has quit
18:51:03 <dajobe> dajobe has joined #rdfig
19:30:03 <DanCon> DanCon has joined #rdfig
19:42:31 <danbri> danbri has quit
19:43:54 <shellac> shellac has quit
19:44:48 <em> http://www-106.ibm.com/developerworks/webservices/library/ws-soaprdf/?open&l=976,t=gr,p=rdf
19:44:49 <dc_rdfig> W: http://www-106.ibm.com/developerworks/webservices/library/ws-soaprdf/?open&l=976,t=gr,p=rdf from em
19:45:03 <em> W:|Using RDF with SOAP
19:45:03 <dc_rdfig> titled item W
19:45:16 <em> W:This article examines ways that SOAP can be used to communicate information in RDF models. It discusses ways of translating the fundamental data in RDF models to the SOAP encoding for PC-like exchange, or for directly passing parts of the model in RDF/XML serialized form.
19:45:16 <dc_rdfig> added comment W1
19:45:27 <em> W: by Uche Ogbuji (uche.ogbuji@fourthought.com)
19:45:28 <dc_rdfig> added comment W2
19:45:28 <DanC> DanC has quit
19:46:27 <JibberJim> em ?
19:46:44 <em> * em here (barely)...
19:46:56 <JibberJim> Any chance of http://www.w3.org/RDF/ being made to validate and actually have the charsets not different in the header/document ?
19:47:22 <em> you'd think they'd be a good chance wouldn't you :)
19:47:33 <dajobe> dajobe has quit
19:47:38 <JibberJim> fair enough ones further down the tree missing validation, but not the front page.
19:48:03 <AaronSw> J:|Tools for transforming HTML imagemaps into SVG and/or RDF, in progress. XSLT from Max F., Ruby from DanBri.
19:48:04 <dc_rdfig> titled item J
19:48:23 <AaronSw> J:Example: input [http://www.w3.org/2001/08/rdfweb/foafwho/imagemap/gimp-w34.imagemap.txt|imagemap from Gimp tool] (pre-XML HTML fragment); [http://www.w3.org/2001/08/rdfweb/foafwho/imagemap/gimp-w34.imagemap.html|tidied HTML version] (note the mailboxes in the HREF's for each person; neat trick huh?); [http://www.w3.org/2001/08/rdfweb/foafwho/imagemap/imagemap2svg.xslt|the xslt transform], and
19:48:23 <dc_rdfig> added comment J2
19:48:49 <AaronSw> J:[http://www.w3.org/2001/08/rdfweb/foafwho/imagemap/_imagemap2svg-output.svg|SVG output] ([http://www.w3.org/2001/08/rdfweb/foafwho/imagemap/_imagemap2svg-output.png|PNG rasterization] using [http://www.w3.org/2001/08/rdfweb/foafwho/|batik and some scripts]). Nearby: [http://www.w3.org/2001/08/rdfweb/foafwho/imagemap/imagemap2meta.rb|draft Ruby script].
19:48:50 <dc_rdfig> added comment J3
19:48:54 <AaronSw> * AaronSw pokes dc_rdfig
19:49:44 <JibberJim> aww they're polygons, not paths :-(
19:51:03 <JibberJim> ah, but your clippath works with polygons! it must've been my fault when I couldn't make it work.
19:52:25 <em> JibberJim, done.
19:52:31 <em> thanks for the feedback
19:52:55 <JibberJim> Thanks for doing it.
19:53:49 <em> http://www-sop.inria.fr/acacia/soft/corese.html
19:53:49 <dc_rdfig> X: http://www-sop.inria.fr/acacia/soft/corese.html from em
19:53:58 <em> X:|Corese : A COnceptual REsource Search Engine
19:53:59 <dc_rdfig> titled item X
19:54:31 <em> xena: The Corese platform implements an RDF/RDFS processor based on Conceptual Graphs (CG).
19:54:35 <AaronSw> T:Pat Hayes: "An aside on logic. There is a widespread misapprehension that logic is 'difficult' - like calculus is supposed to be in American high schools. In fact, basic logic is easier to use and understand than description logics; it has a simpler syntax, it has simpler inference processes, and it is closer to natural language. While there are some subtle aspects of logic, one is not obliged to use them or even to consider them."
19:54:36 <dc_rdfig> added comment T4
19:54:47 <em> X:The Corese platform implements an RDF/RDFS processor based on Conceptual Graphs (CG).
19:54:47 <dc_rdfig> added comment X1
19:55:10 <em> X:part of the Acacia project, Contact : Olivier Corby, INRIA.
19:55:11 <dc_rdfig> added comment X2
19:56:00 <em> X:demo'd at technical plenary... very cool stuff
19:56:01 <dc_rdfig> added comment X3
19:57:53 <AaronSw> em, if you've got any ideas for killer apps for me to build for my WWW2002 demo, send them my way
19:58:45 <em> AaronSw, collaborative email spaminator
19:59:01 <AaronSw> somebody else already built one
19:59:42 <em> you just going to tease us all with that statement, or provide a pointer! :)
19:59:46 <AaronSw> heh.
19:59:51 <JibberJim> Doesn't that also help spammers know what rules they need to beaT?
19:59:55 <em> has to run... but will lurk
19:59:57 <AaronSw> http://razor.sourceforge.net/
19:59:57 <dc_rdfig> Y: http://razor.sourceforge.net/ from AaronSw
20:00:17 <AaronSw> Y:|Vipul's Razor: a distributed, collaborative, spam detection and filtering network
20:00:17 <dc_rdfig> titled item Y
20:00:32 <AaronSw> JibberJim, yes, but spammers seem to slow to notice
20:00:47 <AaronSw> or maybe they're not noticing on purpose
20:01:26 <AaronSw> Y:Works best with [SpamAssassin|http://spamassassin.org/], a genetic-algorithm wated heuristics system for blocking spam
20:01:26 <dc_rdfig> added comment Y1
20:03:26 <sbp> wated?
20:03:44 <deltab> 'weighted', perhaps?
20:03:51 <AaronSw> Y1:Works best with [SpamAssassin|http://spamassassin.org/], a genetic-algorithm weighted heuristics system for blocking spam
20:03:51 <dc_rdfig> replaced comment Y1
20:10:19 <danbri> danbri has joined #rdfig
20:14:34 <jonb> jonb has joined #rdfig
20:15:40 <jonb> danbri: just read pat's paper that you chumped -- i couldn't agree more
20:17:01 <danbri> he writes nicely too
20:17:39 <AaronSw> i like how he even left his .sig on there. ;-)
20:19:45 <jonb> yes it is a very well written paper
20:21:13 <jonb> * jonb admits that he didn't realize the difference between DL and the logic he learned in high school until becoming involved in WebOnt
20:22:55 <jonb> well though DL is a big deal among biomedical folks
20:53:36 <DanCon> * DanCon looks around for em and danbri
20:53:36 <mnot> mnot has quit
20:54:03 <danbri> * danbri notices the #rdfig tab turn blue
20:54:07 <danbri> * danbri eating his dinner
20:54:54 <DanCon> OK if I skip RDF Core telcon tomorrow?
20:55:41 <DanCon> i.e. does the chair have what he needs? have you looked at his proto-agenda? it looks very full; could use some pruning/prioritizing.
20:56:46 <DanCon> did you notice that we decided to fold datatypes into RDF schema?
20:57:13 <danbri> I've talked w/ Brian about the details of the agenda, but did talk with him earlier in the week about datatypes, RDFS, Guha getting back in the loop and related stuff.
20:57:50 <danbri> We talked about having a draft 'datatypes doc to be folded into RDFS' done separately for now, then bolted in. Meanwhile, I've been revisiting my editors tasklist for RDFS proper.
20:58:19 <DanCon> anyway... the ftf resulted in lots of momentum; but I don't feel in a good position to carry it forward. it would be a shame if the momentum was lost because ericm/danbri/danc flaked out.
20:58:20 <danbri> It's up to you re tommorrow; can't decide for you. I won't be annoyed if you're not there, but I don't know what other's expectations might be.
20:58:29 <danbri> Indeed.
20:58:47 <danbri> I'll be there anyway, and try to talk w/ Brian beforehand.
20:59:13 <DanCon> Jeremy seems to be carrying lots of balls forward. seems reasonable to proceed on his stuff. But I think EricM has a steak in lang stuff, so I'd like to ping him before I tune out.
20:59:58 <DanCon> datatypes seems to have gone back-burner, now that PatH is travelling or whatever. Maybe it's good to take a little break from it, but I dunno.
21:00:23 <danbri> Datatpyes seemed to be the most difficult and exhausting of topics...
21:24:17 <connolly> connolly has joined #rdfig
21:24:36 <dmiles> dmiles has quit
21:25:00 <dmiles> dmiles has joined #rdfig
21:31:27 <tav> tav has quit
21:33:16 <niq> niq has joined #rdfig
21:39:24 <tav> tav has joined #rdfig
21:40:55 <DanCon> DanCon has quit
21:54:54 <JibberJim> JibberJim has left #rdfig
22:08:45 <jonb> jonb has quit
22:12:24 <tim-gone> tim-gone has quit
22:12:29 <timbl> timbl has quit
23:20:30 <las> las has quit
23:27:26 <DanCon> DanCon has joined #rdfig
23:39:05 <sat> sat has quit
23:44:05 <connolly> connolly has quit
23:45:46 <SethR> SethR has joined #rdfig
23:54:58 <SethR> T: I still dont understand how using negation ie (:not_p :negates :p} so we can use {:A :not_p :B} leads to inefficient inferences
23:54:58 <dc_rdfig> added comment T5
23:56:20 <SethR> Can anyone can shed some light on that ?
23:59:16 <SethR> anyway I made a mentograhic study of negation at
23:59:19 <SethR> http://robustai.net/mentography/notArrow.gif
23:59:20 <dc_rdfig> Z: http://robustai.net/mentography/notArrow.gif from SethR
23:59:48 <SethR> Z:| mentographic study of negation
23:59:48 <dc_rdfig> titled item Z