Semantic Web Interest Group IRC Chat Logs for 2005-01-16

02:02:45 <djav> hi

02:04:31 <eaon> hey

02:08:15 <djav> anyone familiar with skos around?

02:09:37 <crschmidt> Best to ask your question and hope there is, typically

02:14:43 <djav> hmm well actually it's not really tied to SKOS: I noticed that in the skos-core vocab some deprecated some property have been deprecated and moved to another namespace, for instance skos-extension. Now i m wondering how to reflect that changes into my skos instances, meaning how to split properties into 2 differents namespaces while there are actually tied to only one

02:17:19 <djav> in fact i m wondering if I have to declare the new namespace, and change all the triples which are using the old namespace to use the new one?

05:09:56 <zoyd> hi

09:28:41 <anivar|away> anivar|away is now known as anivar

12:57:34 <CloCkGrr> CloCkGrr is now known as CloCkWeRX

15:59:46 * bblfish sitting in a café in Paris

16:02:12 <bblfish> anyone around? I am feeling more comfortable again with my "a simpler form of rdf xml" . Except that I am wondering if I am not just discovering the foundations of

16:02:14 <bblfish> rdf/xml.

16:03:51 <balbinus_> balbinus_ is now known as balbinus

16:18:16 <bblfish> btw. It would be really nice if I had just restated the foundations of rdf/xml, as that would point to a way to make it easier to explain it, and chime much more nicely with many peoples xml intutions

16:26:17 * bblfish got to go...

17:44:46 <md-afk_> md-afk_ is now known as PhUrl

17:45:00 <PhUrl> hi all

17:45:09 <balbinus> hi

17:45:17 <PhUrl> hi balbinus

17:45:27 <balbinus> hi PhUrl :)

17:48:24 <paolino> ciao,anybody uses python rdflib for parsing?

17:48:49 <crschmidt> Yep

17:49:03 <crschmidt> Although I do most of my work with Redland

17:49:33 <paolino> which is not rdflib I argue

17:51:17 <PhUrl> [ot] hehe: romanian couple who meets on the internet names kid "yahoo" http://news.google.com/?ned=us&ie=UTF-8&hl=en&ncl=http://www.newkerala.com/news-daily/news/features.php%3Faction%3Dfullnews%26id%3D60261

17:53:00 <paolino> anyway,I parsed http://www.semanticweb.org/library/wordnet/wordnet_nouns-20010201.rdf with rdflib and there seems no bnodes in the parsing,is it ok?

17:58:32 <libby> there's no need for there to be bnodes in every piece of rdf

17:59:00 <PhUrl> Wow DCE/RPC has been released under lgpl! http://www.advogato.org/article/817.html

18:00:40 <paolino> libby,so it's been not been a choice of the parser,there are no bnodes in that graph?

18:00:54 <libby> I dunno, it crashed browser

18:01:02 <PhUrl> free software developers are ""reinventing the wheel with things like D-Bus (which contains no IDL compiler and no networking capability) and Global File System (which is proprietary anyway, available from Redhat), and by having to reinvent or hand-build an RPC system: Bayonne, the GNU/Hurd, the KDE project - the list goes on""

18:02:32 <PhUrl> arnia should read this

18:02:41 <PhUrl> dbus is used by frege

18:03:57 <libby> at a quick glance paolino, looks like each node has a url

18:04:01 <libby> so no bnodes.

18:04:07 <libby> but not checked it all!

18:05:32 <paolino> libby,can you point me to a graph,with some complexity and bnodes in?

18:06:59 <paolino> PhUrl,reinventing the wheel makes me feel good when I know what I do

18:08:04 <libby> my foaffile has bnodes: http://swordfish.rdfweb.org/people/libby/rdfweb/webwho.xrdf

18:08:10 * libby gotta go sorry

18:08:17 <paolino> thnx

18:15:54 <paolino> crschmidt, how rdflib distinguish literals from uries?

19:07:37 <bblfish> hi

19:10:45 <bblfish> PhUrl: you still here?

19:11:14 <PhUrl> yes

19:11:18 <PhUrl> hi bblfish

19:12:25 <bblfish> hi. Just wondering if you think my last mail is a good reply to the points made by sbp yesterday

19:13:07 <PhUrl> uri?

19:13:40 <bblfish> just a sec

19:14:39 <bblfish>http://lists.w3.org/Archives/Public/www-rdf-interest/2005Jan/0111.html

19:14:40 <dc_swig> A: http://lists.w3.org/Archives/Public/www-rdf-interest/2005Jan/0111.html from bblfish

19:18:13 <bblfish> A:| A Class Object is a relation that all objects of that class have to themselves.

19:18:13 <dc_swig> Titled item A.

19:19:20 * PhUrl reads

19:19:48 <sbp> bblfish: in OWL DL, classes and properties are disjoint. so that would be inconsistent

19:20:23 <sbp> see http://www.w3.org/TR/2004/REC-owl-semantics-20040210/rdfs.html#5.4

19:20:32 <sbp> "LVI, IOT, IOC, IDC, IOOP, IODP, IOAP, IOXP, IL, and IX are all pairwise disjoint."

19:21:49 <sbp> logger: chump A

19:21:49 <sbp> A:See [http://www.ilrt.bris.ac.uk/discovery/chatlogs/swig/2005-01-16#T19-21-49|discussion]

19:21:50 <dc_swig> Added comment A1.

19:30:59 <bblfish_> I never quite finished reading the RDF Semantics, met alone starting the OWL semantics.

19:31:02 <bblfish_> sbp: so it would work in OWL Full?

19:31:02 <bblfish_> /me I can't tell if this irc channel is very brittle or if it is my connection. I keep getting disconnected

19:35:06 <PhUrl> is this valid n3: what is the equivalent rdf/xml ? :statement_pointer :stmt { :subject :predicate :statement_pointer} .

19:36:38 <dajobe> it's valid n3, but because it has {}, it is not rdf

19:36:47 <timbl> Anything with curly {braces} in N3 is outside the set of RDF graphs, and cannot be expressed (unreified) in RDF/XML.

19:37:17 <PhUrl> is there a way to write :subject :predicate :this_statement in rdf?

19:37:50 <PhUrl> this is the reification issue...

19:37:54 <timbl> You mean where this_statement is a statement?

19:38:10 <PhUrl> yes, and this_statement is the thatement itself

19:38:17 <PhUrl> the statement itself

19:38:21 <timbl> If you take the N3 you wrote above and do cwm that.n3 --reify you will get pure RDF.

19:38:23 <PhUrl> like in that example

19:38:29 <PhUrl> ok, let me see

19:38:48 <PhUrl> NICE

19:38:54 <balbinus> :)

19:38:56 <PhUrl> timbl: thanks, i did not know that option

19:39:50 * timbl thinks ... maybe --flatten is better.

19:40:31 <timbl> Yes. echo ':statement_pointer :stmt { :subject :predicate :statement_pointer} .' | cwm --flatten

19:41:22 <PhUrl> http://rafb.net/paste/results/rvsGLW20.html

19:41:36 <timbl> Flatten doesn't reify things which are already straight RDF.

19:42:50 <PhUrl> flatten http://rafb.net/paste/results/YLMLxY45.html

19:43:24 <PhUrl> this is nice stuff

19:43:29 <PhUrl> thanks timbl, dajobe

19:43:50 * danbri wanders past, waves

19:44:13 <balbinus> hi danbri :))

19:44:13 <danbri> nice work re rasqal sparql stuff, dajobe

19:44:19 <danbri> hiya

19:44:28 <dajobe>http://lists.w3.org/Archives/Public/www-rdf-interest/2005Jan/0117.html

19:44:29 <dc_swig> B: http://lists.w3.org/Archives/Public/www-rdf-interest/2005Jan/0117.html from dajobe

19:44:38 <dajobe> B:|Rasqal 0.9.5 - with SPARQL optionals

19:44:38 <dc_swig> Titled item B.

19:44:52 <dajobe> B:also made with a lot of blood, sweat and tears

19:44:52 <dc_swig> Added comment B1.

19:45:44 <danbri> how hard will "source" support be, dave?

19:45:57 <dajobe> easyish

19:46:03 <danbri> :) coool

19:46:13 <dajobe> the code's mostly there, just not connected up to the syntax, which isn't stable

19:46:37 * danbri nods

19:55:02 <bblfish_> is OWL defined in OWL DL or OWL Full? (If that question makes sense)

19:56:15 <PhUrl> rudolf/rdf is taking forever for http://introspector.sourceforge.net/2005/01/reified_test.rdf

19:56:24 <PhUrl> it there another rdf viz tool

19:58:45 <bblfish_> Or perhaps more precisely is owl:Class subject to the restrictions of OWL DL?

20:04:20 <PhUrl> bblfish_: i will need longer to parse this mail. sorry

20:04:45 <bblfish_> ok. :-)

20:04:51 <bblfish_> It is fun. I am going to take a lot longer to parse OWL Semantics

20:09:34 * bblfish_ but perhaps I could use that cwm trick mentioned above to test out some of those ideas...

20:09:39 <PhUrl> ok, this is the graph from the rdf validator http://introspector.sourceforge.net/2005/01/reified_test.png

20:10:13 <PhUrl> it is a cyclic self referential thingie, over statement_pointer

20:11:48 <PhUrl> trying to refute http://www.arcavia.com/kyle/Analysis/References.html "RDF can not refer to it's own triples using triples"

20:12:16 <PhUrl> maybe not directly, but indirectly

20:13:09 <timbl> { preceded by itself in braces is false } preceded by itself in braces is false ?

20:13:27 <PhUrl> heheh

20:13:39 <PhUrl> this statment is false

20:13:53 <timbl> { @this a Falsehood }

20:14:18 <PhUrl> yeah, thats the ticket, is @this defined?

20:14:20 <timbl> <>!log:semantics rdf:type log:Falsehodd.

20:14:52 <timbl> @this is deprocated ... we had too many engineers dissapearing in puffs of smoke. ;-)

20:15:16 <timbl> But <> is slways the current document, as per the URI spec.

20:16:06 <PhUrl> ok, well, i think that i shown kyle that rdf can be used, we will see what he says

20:18:36 <PhUrl> thanks again timbl. learned some new things todaz

20:18:39 <PhUrl> today

20:19:16 <PhUrl> bblfish_: i have read this mail. I dont know exactly the point. you are mapping atom xml into rdf?

20:21:50 <PhUrl> i read the rest of the thread

20:22:14 <bblfish__> bblfish__ is now known as bblfish

20:22:59 <bblfish> PhUrl: hi just got cut off (Again! could be France Telecom)

20:23:03 <PhUrl> ok, bblfish, http://lists.w3.org/Archives/Public/www-rdf-interest/2005Jan/0062.html is a good start

20:23:37 <PhUrl> i am really not that up to speed on this issue, but intuitivly, i think that you can represent the xml infoset as a rdf graph

20:23:48 <PhUrl> i dont know how much that helps...

20:24:10 <PhUrl> i dont know if my brain is turned on enought to say much intelligent about this issue atm

20:24:28 <bblfish> I was trying to find a simple mapping from xml to tripples

20:24:50 <PhUrl> well, as i said before, you can make a xslt to do that for alot of cases

20:25:28 <PhUrl> but really, do you need it to be automatic? i think that you can produce lots of bloated rdf that way

20:32:34 <PhUrl> looks like freenode is under attack

20:32:41 <PhUrl> chanserv segfaulted before

20:32:54 <crschmidt> Chanserv segfaults regularly

20:32:57 <crschmidt> once or twice a week

20:33:02 <PhUrl> really?

20:33:26 <crschmidt> eh, maybe that often

20:33:27 <bblfish_> you may wish to try out myjxta which is a p2p chat system

20:33:33 <crschmidt> but it happens every now and then

20:34:17 <bblfish_>http://myjxta2.jxta.org/

20:34:17 <dc_swig> C: http://myjxta2.jxta.org/ from bblfish_

20:34:57 <bblfish_> C:| myjxta a up and coming replacement for irc?

20:34:58 <dc_swig> Titled item C.

20:35:07 <PhUrl> what about jabber_

20:35:13 <PhUrl> ?

20:35:21 <bblfish_> C: ask gonzo for more info

20:35:21 <dc_swig> Added comment C1.

20:37:22 <eaon> there's more "irc replacement" projects out there - http://alurio.amonre.org/

20:37:34 <PhUrl> C:See also, jabber vs jxta [http://jamie.ideasasylum.com/2003/08/jabber-vs-jxta-take-ii.php]

20:37:34 <dc_swig> Added comment C2.

20:38:23 <bblfish_> PhUrl: In my initial proposal I show an intuitive way to map xml to tripple space (and intuitive does not involve xslt)

20:38:42 <bblfish_> I thought this may be so general that it could apply to all xml. Then sbp pointed out that it did not seem to apply to

20:38:42 <bblfish_> xml/rdf (which many would consider non intuitive anyway). This last email I thought I showed that my mapping worked just

20:38:42 <bblfish_> as well for xml/rdf.

20:39:05 <PhUrl> ok, i understand now. thanks.

20:40:17 <bblfish_> I thought my solution (at least the diagram showing how a class is a relationship between every instance of the class and itself, was

20:40:19 <bblfish_> a really neat way of solving that problem :-)

20:41:25 <PhUrl> the class extent

20:42:14 <PhUrl> yeah, each member of a set has a relationship (reference) to the set itself.

20:43:23 <bblfish_> that is one way of looking at it. The other way of looking at it

20:43:28 <bblfish_> is to think of the set as a relationship between every member of the set and itself

20:44:00 <bblfish_> so that means one has to think of <Entry> as both an object: the set

20:44:19 <bblfish_> and as a relationship that holds between every member of the set _mem and itself ie _mem

20:44:22 <PhUrl> yes, it can also be that a set is a relationship between every member of the set and every other member :)

20:45:04 <PhUrl> so you would have a unique property for each set that gives you all the ordering and members of each set

20:45:30 <bblfish_> yes, but that relationship would not help me in my transformation...

20:45:54 <PhUrl> just an idea, i like the idea of monsterous sets...

20:46:19 <PhUrl> what is _mem?

20:46:28 <PhUrl> ah, the set.

20:46:34 <bblfish_> no a member

20:46:54 <PhUrl> ok, i see.

20:47:04 <bblfish_> so PhUrl isa Human

20:47:13 <bblfish_> and the Human relationship relates PhUrl to PhUrl

20:47:48 <bblfish_> but you are right. It perhaps also realtes PhUrl to me

20:47:48 <bblfish_> and perhaps that is also ok for my transformation.

20:47:49 <PhUrl> so a self referencial node

20:47:59 <bblfish_> yes

20:48:08 <PhUrl> so you have a reference to the set and itself as a member of that set

20:48:22 <PhUrl> and via thet set a reference to each other node in the set

20:48:36 <bblfish_> that's it

20:48:40 <PhUrl> but do you need a self reference? can that not be infered

20:49:12 <bblfish_> if it can be inferred and it works in my tranformation then that would be ok... But I have to look at the transformation again.

20:50:16 <PhUrl> let me ask one more silly question : you are trying to find a rdf rep of atom, or a generic model?

20:50:54 <bblfish_> <Feed>

20:50:56 <bblfish_> <entry>

20:50:57 <bblfish_> <Entry>

20:50:59 <bblfish_> <title>my first entry</entry>

20:51:00 <bblfish_> </Entry>

20:51:02 <bblfish_> </entry>

20:51:03 <bblfish_> </Feed>

20:51:05 <bblfish_> I am trying to find something that works for Atom, but being ambitious I would be happier if it solves every problem in the rdf universe :-)

20:51:06 <bblfish_> _F --Feed--> _f

20:51:08 <bblfish_> |--entry-->_E

20:51:09 <bblfish_> |--Entry-->_e

20:51:11 <bblfish_> |--title-->_t

20:51:12 <bblfish_> |--xxx:string--+

20:51:14 <bblfish_> |

20:51:16 <bblfish_> "My first entry" <-------+

20:51:26 <dajobe> er, that's a bit large for irc

20:51:27 * eaon begs bblfish to learn n3

20:51:52 <bblfish_> ok

20:52:25 <eaon> [ a :Feed; :entry [ a :Entry; :title "my first entry"; ]; ] . :)

20:52:31 <PhUrl> bblfish_: you should use http://rafb.net/paste/

20:52:35 <eaon> ^^xsd:string maybe too

20:52:44 <bblfish_> F_ :feed _f .

20:52:45 <bblfish_> _f :entry _E .

20:52:47 <bblfish_> _E :entry _e .

20:52:48 <bblfish_> _e :title _t

20:52:50 <bblfish_> _t xxx:string "my first entry".

20:53:24 <PhUrl> bblfish_ do you have redland and cwm installed?

20:53:30 <bblfish_> yes

20:53:41 <bblfish_> but I needed the named anonymous nodes.

20:53:42 <PhUrl> is there not a atom to rss converter?

20:54:05 <PhUrl> rss being the rdf version of rss, there so many, i dont know which?

20:54:06 <dajobe> rapper can do atom to rss

20:54:07 <eaon> _:F

20:54:48 <PhUrl> so, do we need another rdf representation of atom, if the rss representation is equiv, or am i missing something?

20:55:22 <dajobe> that doesn't use the terms that the atom world uses - feed, entry

20:55:35 <dajobe> but the ones that rss1.0 (and earlier) uses - channel, item etc.

20:55:47 <bblfish_> _:F :Feed :f.

20:55:49 <bblfish_> :Feed a owl:Class

20:55:50 <bblfish_> I want to deduce that

20:55:52 <bblfish_> _:F == _f

20:56:19 <PhUrl> ok, so you can define a query or cwm transformation that renames the predicate to atom and you are done?

20:56:21 <dajobe> deduce? with whose semantics?

20:57:02 <PhUrl> is this just a question of naming the things, or something fundamentally different?

20:57:26 <bblfish_> the semantics of what a owl:Class is hopefully

20:57:43 <PhUrl> bblfish_: i dont think i can add anything here, i am confused and suffering from a hangover. I got some bugs to fix. :)

20:58:03 <bblfish_> ok :-)

20:59:01 <bblfish_> dajobe: I describe what I want to do more clearly here: http://lists.w3.org/Archives/Public/www-rdf-interest/2005Jan/0111.html

20:59:50 * dajobe is baffled

21:00:29 <dajobe> I think you are heading down a well trod and ultimately rather fruitless path of trying to get something out of "simple XML" apart from just a tree of elements.

21:01:36 <bblfish_> I just show how to get a tree of elements. Then I add an ontology and hey presto: I get what one gets with rdf/xml

21:01:58 <dajobe> is there an OWL DL machine behind the curtain? or what?

21:02:45 <bblfish_> two steps:

21:02:46 <bblfish_> 1. xml -> very simple tree of elements

21:02:48 <bblfish_> 2. OWL machinery + ontology

21:03:07 <bblfish_> --> an interpretation of xml

21:04:39 <bblfish_> the quickest explanation is here: http://lists.w3.org/Archives/Public/www-rdf-interest/2005Jan/0090.html

21:04:58 <dajobe> it's not I you need to worry about selling that to, but atom people. I think that's very very hard.

21:05:21 <bblfish_> oh. I think it is more general that atom

21:05:38 <bblfish_> In fact I was thinking this is so general that xml/rdf falls under it too.

21:08:26 <bblfish_> but that generality was questioned by spb here: http://lists.w3.org/Archives/Public/www-rdf-interest/2005Jan/0109.html

21:08:45 <bblfish_> and I tried to respond here: http://lists.w3.org/Archives/Public/www-rdf-interest/2005Jan/0111.html

21:09:16 <PhUrl> bblfish_: if this is to work, then you will be able convert xmlschema docs into rdf. then you will have a rdf representation of any give xmlschema? so then you want to be able assign that document to an xmlschema ontology and then later be able to transform an xmlschema into an owl ontology?

21:09:45 <PhUrl> will you be able to create an ontology out of the structure of any xml document or what?

21:10:08 <bblfish_> no. that is too general and I don't think can be done.

21:10:28 <PhUrl> but this tool would be able to convert an xsd into some form or rdf?

21:10:49 <bblfish_> I am not sure. I don't know xsd

21:11:05 <PhUrl> ok, well, i think that would be more interesting as an example than atom :)

21:11:35 <PhUrl> or, be able to convert an xslt doc into rdf? now that would be interesting.

21:11:45 <bblfish_> yes, but I know Atom very well now :-)

21:12:25 * eaon doesn't see why xslt as rdf would make sense...

21:13:09 <PhUrl> eaon: it would not... not right away. would be interested in what his tool would produce

21:13:28 <PhUrl> i mean, just a more complex example than atom.

21:13:51 <PhUrl> trying to think of more challenging xml types

21:14:07 <bblfish_> I think the first step is quite uncontroversial:

21:14:07 <bblfish_> 1

21:14:07 <bblfish_> 2

21:14:07 <bblfish_> 3

21:14:07 <bblfish_> 4

21:14:08 <bblfish_> 5

21:14:10 <bblfish_> 6

21:14:12 <bblfish_> 7

21:14:14 <bblfish_> 8

21:14:16 <bblfish_> 9

21:14:18 <bblfish_> 10

21:14:20 <bblfish_> 11

21:14:22 <bblfish_> 12

21:14:24 <bblfish_> 13

21:14:26 <bblfish_> 14

21:14:28 <bblfish_>

21:14:30 <bblfish_> <feed>

21:14:30 <PhUrl> bblfish_: you should use http://rafb.net/paste/

21:14:32 <bblfish_> <entry>

21:14:34 <bblfish_> <title>hello</title>

21:14:36 <bblfish_> <id>urn:id:id1</id>

21:14:38 <bblfish_> </entry>

21:14:40 <bblfish_> </feed>

21:14:42 <bblfish_>

21:14:43 <crschmidt> PhUrl: probably a bit late

21:14:44 <bblfish_> maps to

21:14:53 <bblfish_>

21:14:53 <bblfish_> _f -feed-> _e

21:14:53 <bblfish_> |--title-->_t

21:14:53 <bblfish_> | |--xx:text--> "hello"

21:14:54 <bblfish_> |--id----->_id

21:14:56 <bblfish_> |--xxx:text-> "urn:id:id1"

21:14:58 <bblfish_> oops. I tried that nopaste tool. sorry

21:15:02 <bblfish_> ah I see. I am meant to point to that page...

21:15:28 <PhUrl> and because they are temp pages, put a space in front of the uri, so it dont get chumped

21:15:38 <bblfish_> yes. I did not understand the point. I do now :-/

21:17:16 <bblfish_> anyway the point is rather to make it much easier to model xml

21:17:57 <bblfish_> because if it makes immediate sense how xml and ontologies interact, then one can explain how to do OO xml development

21:18:42 <bblfish_> ie: how to correctly use foreign name spaces in an xml document (of which atom is a good example)

21:19:32 <PhUrl> well, i would be interested in how things like xmlschema and relaxng map onto ontologies

21:20:02 <bblfish_> well you could try with a simple xmlschema and post it to the thread

21:20:03 <eaon> xmlschema contains logic? it'd rather map to rdf-schema

21:20:19 <PhUrl> because if you can map a schema onto an ont, then you can transform that xml document matched by the schema into an the beginnings of a ont

21:20:43 <PhUrl> ok, good point. i mean a rdf vocab. i should not use ont unless i mean it

21:21:13 <PhUrl> iirc, this all has been well trodden by the rdfig people in the past as dajobe has mentioned.

21:32:18 <bblfish__> bblfish__ is now known as bblfish

21:36:38 <bblfish> Here is a procedure for creating an ontology from some simple xml.

21:36:41 <bblfish> 1. start with some reasonably simple well known xml

21:36:41 <bblfish> 2. try the procedure of mapping every element and attribute to a rdf relation between blank nodes (as in my examples)

21:36:41 <bblfish> 3. for each property specify what the type of the domain objects and the range are

21:36:41 <bblfish> This will allow one to do some simplifications on the graph such as:

21:36:42 <bblfish> if

21:36:44 <bblfish> _blank rdf:type xxx:String .

21:36:46 <bblfish> _blank xxx:string "hello I am joe" .

21:36:48 <bblfish> then

21:36:50 <bblfish> _blank == "hello I am joe"

21:36:52 <bblfish> 4. you will want to work out if any of the properties are functional, inverse function, symmetric, or transitive

21:36:54 <bblfish> this will alow you to do further graph transformations

21:36:56 <bblfish> as you do this you will build an ontology of xmlschema

21:37:16 <bblfish> sorry: as you do this you will build an ontology of your xml

21:40:47 <sbp> what about context sensitive XML?

21:41:18 <sbp> <p><q><p><q><r/></q></p></q></p> the outermost <q> element means something entirely different to the innermost one

21:41:22 <eaon> i'm afraid he's gone?

21:41:29 <bblfish> no I am back

21:41:39 <eaon> oh, damn irc

21:42:06 <sbp> just one of the many points that the plan falls short on

21:42:40 <sbp> <PhUrl> iirc, this all has been well trodden by the rdfig people in the past as dajobe has mentioned.

21:42:40 <bblfish> well I am trying to perhaps deal only with well thought out xml

21:42:49 <kota_> kota_ is now known as kota

21:43:05 <sbp> there have been a lot of efforts to do XML/XMLSchema <=> RDF mappings, but none of this slightly saneless approach

21:43:21 <sbp> e.g. DanC, EricP, and myself have worked on infoset-in-RDF and schema annotation technologies

21:43:43 <sbp> you think context sensitive XML isn't well thought out?

21:43:57 <sbp> even XHTML uses the "title" QName twice

21:44:15 <bblfish> don't know.

21:44:16 <bblfish> is it?

21:46:02 <bblfish> well anyway. Thanks for all the feedback. I have learnt a lot along the way, but probably should learn a lot more eg: OWL semantics, rdfschema and other things before I continue

21:47:15 <dajobe> you seen the work swbpd has been doing?

21:47:15 <sbp> you're welcome. I only wish you'd heed it a bit! :-)

21:47:37 <dajobe>http://www.w3.org/2001/sw/BestPractices/XSCH/xsch-sw/

21:47:38 <dc_swig> D: http://www.w3.org/2001/sw/BestPractices/XSCH/xsch-sw/ from dajobe

21:47:42 <dajobe> D:|XML Schema Datatypes in RDF and OWL

21:47:42 <dc_swig> Titled item D.

21:48:07 <bblfish> sbp: well I was going to give up yesterday because of your remarks, but then I thought I found a really clever workaround

21:48:14 <dajobe> D:this is tricky stuff and I think it's still not entirely clear how it all works together. work in progress for a couple of years

21:48:14 <dc_swig> Added comment D1.

21:48:38 <dajobe> D:very *DRAFT*

21:48:38 <dc_swig> Added comment D2.

21:48:40 <bblfish> dajobe: thanks for that link. I should look into that.

21:49:31 <dajobe> I'm not sure it's possible myself, to get all of RDF, OWL and XML Schema in your brain at one time ;)

21:51:46 <bblfish> There are lots of fun things. I thought the other fun one is the mapping of Java to OWL

21:52:11 * jsled shudders

21:55:31 <bblfish> jsled: I have an implementation http://lists.w3.org/Archives/Public/www-rdf-interest/2005Jan/0031.html

21:55:31 <bblfish> I think it is really cool

21:56:17 <jsled> Oh yes, I do remember seeing that. Java just gives me the shakes, especially on the weekends. :)

21:56:29 <bblfish> Oh. Come on.

21:57:40 <bblfish> thinking about the relation between OO and OWL is really helpful (for people coming from the OO community at least)

21:58:21 <dajobe> it is

21:58:27 <dajobe> the uml world is interested too

21:59:31 <bblfish> and that is why I was interested in my simple xml->rdf mapping. Cause I thought then you can explain how to extend xml in an OO way. Pretty weird. but it strikes me as true

22:01:27 <bblfish> anyway. There"s a lot of virgin territory here ...

22:01:28 <bblfish> Got to go to bed now. Byee..

22:02:02 <dajobe> bye

22:03:11 <PhUrl> jsled: hehe, java makes me shudder as well..

22:03:32 * dajobe thinks happy thoughts

22:03:46 <dajobe> well at least eclipse reduces the pain somewhat

22:04:24 <sbp> nurofen's better

22:05:15 <PhUrl> i remember using netbeans on a 800mhz laptop. after a certain point, you just will trade all the gui for speed.

22:05:43 * sbp bookmarks eclipse

22:06:43 <dajobe> it's got good C and python IDEs (so I hear)

22:08:06 <PhUrl> dajobe: how is the speed for you ?

22:08:24 <dajobe> it's fast enough here, 2GHz pc java1.5 linux

22:08:50 <dajobe> only took 10 years to make a pc that runs java fast

22:08:57 <PhUrl> lol

22:09:10 <dajobe> pc & jdk actually

22:11:11 <sbp> looks worth investigating further, from a skim o' the whitepaper

22:11:17 * PhUrl renames structure fields to include a class prefix... improves readablily of code when you are hung over.

22:12:31 <PhUrl> if you want to see how good your code is documented, drink a bottle of burbon and try and fix bugs the next day.

22:16:00 <jsled> heh

22:18:20 <jsled> "18 millions users of RDF -- By Andrew Newman"

22:18:55 <jsled> Hmm. This is going to be like OS/2. "18 million OS/2 seats [17.9 million inside ATMs]."

22:19:45 <dajobe> it's rather daft. how many users of xml are there? What would that number help you do?

22:20:13 <jsled> .w daft

22:20:19 <phenny> daft 1. informal or slang terms for mentally irregular; "it used to drive my husband balmy"

22:20:31 <dajobe> that's a daft definition

22:20:35 <jsled> heh

22:20:38 <dajobe> barmy even