Semantic Web Interest Group IRC Chat Logs for 2002-07-18

00:42:12 <oierw`> oierw` is now known as oierw|sleep

02:10:33 <jhendler>http://www.semaview.com

02:10:33 <dc_rdfig> A: http://www.semaview.com from jhendler

02:11:13 <jhendler> A: a small company in Toronto started up to do Sem Web stuff

02:11:13 <dc_rdfig> added comment A1

02:11:36 <jhendler> A: they have some nice documents on their site (intro, business case, etc.)

02:11:37 <dc_rdfig> added comment A2

02:11:59 <jhendler> A: no endorsement implied

02:11:59 <dc_rdfig> added comment A3

02:12:16 <jhendler> A:| Semaview Inc web page

02:12:16 <dc_rdfig> titled item A

02:20:47 <MarkB> met them at WWW2002. Nice guys.

02:27:59 <jhendler> very interested in this stuff, have sent me various emails from time to time - I was impressed by how hard they've worked to understand the stuff for real

04:15:34 * DanCon discovers apt-gettable blackdown java stuff, installs it...

04:15:35 <DanCon> wins!

04:15:40 <DanCon> arguml is running.

04:16:15 <DanCon> could not set package due to:java.lang.NullPointerException' at null

04:16:17 <DanCon> hmm...

04:20:34 <sbp-> sbp- is now known as sbp

04:22:27 <deltab_> deltab_ is now known as deltab

04:22:43 * DanCon gets RIC running... cool!

04:23:39 <DanCon> wow... RIC really doesn't like 404s for namespace names

04:27:56 <golbeck> RIC didn't crash did it? i just give some nasty exception

04:28:14 <DanCon> no crash

04:28:18 <deltab> RIC?

04:28:29 <golbeck> RDF Instance Creator

04:28:41 * DanCon can't figure out how to get XMI nor SVG out of argouml

04:28:48 <golbeck>http://www.mindswap.org/~mhgrove/RIC/RIC.html

04:28:49 <dc_rdfig> B: http://www.mindswap.org/~mhgrove/RIC/RIC.html from golbeck

04:28:52 <golbeck> b:

04:28:59 <golbeck> B:|RDF Instance Creator

04:29:00 <dc_rdfig> titled item B

04:29:05 <golbeck> B: From MINDSWAP

04:29:05 <dc_rdfig> added comment B1

04:29:11 <golbeck> B: Written by Mike Grove

04:29:11 <dc_rdfig> added comment B2

04:29:34 <golbeck> B: Gives users a nice form to fill out in order to create RDF instances

04:29:35 <dc_rdfig> added comment B3

04:29:51 <golbeck> general question for anyone.

04:30:08 <golbeck> i want to say, in daml, that a property has the range of a disjointUnionOf classes

04:30:33 <golbeck> <rdfs:range>

04:30:36 <golbeck> <daml:disjointUnionOf rdf:parseType="daml:collection">

04:30:43 <golbeck> <daml:Class rdf:about="Format"/>

04:30:49 <golbeck> <daml:Class rdf:about="Content"/>

04:30:55 <golbeck> </daml:disjointUnionOf>

04:30:59 <golbeck> </rdfs:range>

04:31:01 <golbeck> doesn't work

04:31:07 <golbeck> but i'm not sure how to do it otherwise

04:31:11 <golbeck> any ideas?

04:31:36 <DanCon> in what way does that not work?

04:31:51 <DanCon> ah... you need rdf:parseType="Resource" on rdfs:range

04:32:14 <DanCon> or stick an <rdf:Description> element between range and disjointUnionOf

04:32:21 <golbeck> oh. that was pretty easy. so <rdf:range rdf:parseType="Resource'>

04:32:28 <DanCon> yup

04:32:37 <golbeck> cool. thanks a bunch

04:32:48 <DanCon> it 'works' now? works in what way?

04:32:51 <golbeck> i should have figured that one out from all the examples i've been looking at

04:32:54 <DanCon> i.e. what are you up to?

04:33:07 <golbeck> just converting this nasty xmls to daml

04:33:17 <golbeck> it's been rough figuring out the exact syntax.

04:33:17 <DanCon> xmls?

04:33:24 <golbeck> xml schema

04:33:31 <DanCon> hmm

04:33:37 <golbeck> that's my thinking.

04:33:46 <DanCon> what xml schema?

04:35:50 * DanCon tries newly-minted http://www.w3.org/2000/10/swap/pim/ical.rdf in RIC

04:35:54 * DanCon waits for chacl...

04:36:14 <DanCon> !

04:36:18 <DanCon> I can't paste into RIC

04:36:28 <golbeck> paste?

04:36:37 <golbeck> like pasting in values?

04:36:56 <DanCon> like pasting in the URL field of the "Load from URL" dialog

04:37:37 <golbeck> weird - it works on my version in OS X

04:37:42 <DanCon> ugh... and it threw away the URL when it found an error, so now I have to type it again.

04:41:58 * DanCon fixes stray pointer in ns decl in http://www.w3.org/2000/10/swap/pim/ical.rdf , tries again...

04:42:58 <DanCon> er... this time no errors, but no results either

04:43:30 <golbeck> a good place to look is in the event.log file. that will list any problems.

04:44:58 <DanCon> ERROR: http://www.w3.org/2000/10/swap/util/sniffSchema#domainIntersects is an unrecognized or invalid identifier. Cannot create.

04:44:58 <dc_rdfig> Label ERROR not found.

04:45:00 <DanCon> ???

04:45:09 <DanCon> what's wrong with domainIntersects?

04:45:40 <golbeck> is that the URI you input, or is that the particular thing it's getting stuck on?

04:45:56 <DanCon> what I input was view-source:http://www.w3.org/2000/10/swap/pim/ical.rdf

04:46:04 <DanCon> er... http://www.w3.org/2000/10/swap/pim/ical.rdf

04:47:26 <golbeck> i'm not sure what that problem is. i had it do that to one of my rdf files too.

04:47:32 <golbeck> i thought it was my error, but maybe not.

04:48:08 <DanCon> well, if it's saying that sniffschema.{n3,rdf} doesn't declar domainIntersects as a property, I suppose I could fix that.

04:50:53 <golbeck> this http://www.w3.org/2000/10/swap/pim/ical.rdf gets a bunch of error in the daml validator too

04:53:58 * DanCon doesn't often agree with the daml validator

04:55:06 <DanCon> [[

04:55:07 <DanCon> Property http://www.w3.org/2000/10/swap/util/sniffSchema#intersectsDomainOf is not explicitly defined at the specified namespace URI.

04:55:08 <DanCon> ]]

04:55:10 <DanCon> fair enough...

04:55:28 <DanCon> [[

04:55:28 <DanCon> The following file substitutions were made by the DAML Validator: http://www.w3.org/2000/01/rdf-schema# -> file:/home/mdean/daml/validator/cache/www.w3.org_2000_01_rdf-schema

04:55:29 <DanCon> ]]

04:55:32 <DanCon> spiffy!

04:55:59 <DanCon> fair enough: [[ Unable to access namespace. java.io.FileNotFoundException: http://www.w3.org/2000/01/foo-X@@# ]]

04:57:48 <golbeck> danc: on that disjoint union question, should i be using a special 'Resource' (daml:Resource, rdf:resource, etc)? all the combo's i've tried still leave me with a validator error that the disjointUnionOf is a range mismatch for rdfs:range, which expects an rdfs:Class

04:57:48 <DanCon> wow... swap/log has a bogus pointer to 2000/01/pim/doc; I wonder how long that's been there!

04:59:04 <DanCon> so do <rdfs:range><rdfs:Class><daml:disjoinUnionOf>...

04:59:45 <DanCon> the silly thing should infer that the bnode is of type rdfs:Class; that's the sort of thing I don't like about the daml validator.

05:00:00 * DanCon is finding real bugs thanks to the daml validator, meanwhile...

05:00:24 <golbeck> woo hoo!

05:00:27 <golbeck> thanks :)

05:02:25 <DanCon> welcome

05:02:37 * DanCon continues bug-hunting ical and stuff it depends on...

05:09:45 * DanCon grumbles... RDF validator uses POST when it should use GET

05:13:52 <golbeck> looks like the validator doesn't enforce oneOf

05:14:00 <golbeck> but enforces a lot of domain and range stuff

05:15:03 * DanCon isn't surprised about oneOf

05:17:25 <golbeck> how likely might it be that the daml validator wouldn't support disjointUnionOf?

05:17:45 <golbeck> it accepts the code, but doesn't seem to understand it.

05:19:04 <DanCon> quite likely

05:19:14 * DanCon had no idea how much stuff my new ical namespace depended on...

05:32:13 * DanCon gets down to one niggling bug per daml validator, tries RIC again...

05:35:19 <DanCon> does RIC do anything with HTTP Accept headers when it fetches RDF files?

05:35:32 <golbeck> not that i'm aware of

05:35:49 <DanCon> hm... it's probably getting the N3 version of http://www.w3.org/2000/10/swap/util/sniffSchema then

05:36:04 <golbeck> that would definitely cause trouble

05:36:42 <DanCon> no, wget gets the .rdf version

05:37:28 <DanCon> I still don't know what to make of this diagnostic: [[

05:37:30 <DanCon> ERROR: http://www.w3.org/2000/10/swap/util/sniffSchema#domainIntersects is an unrecognized or invalid identifier. Cannot create.

05:37:30 <dc_rdfig> Label ERROR not found.

05:37:30 <DanCon> ]]

05:38:05 <DanCon> odd... Invalid URL in namespace declaration file:/home/connolly/w3ccvs/WWW/2000/10/swap/util/sniffSchema#

05:38:14 <DanCon> ah...

05:38:24 <golbeck> looks like there is no code there

05:42:06 <DanCon> ??? [[ ERROR: http://www.w3.org/1999/02/22-rdf-syntax-ns#Description is an unrecognized or invalid identifier. Cannot create.

05:42:06 <DanCon> The resulting exception was:

05:42:06 <DanCon> java.lang.NullPointerException

05:42:07 <DanCon> ]]

05:42:58 <DanCon> also: [[[

05:43:00 <DanCon> ERROR: http://www.w3.org/2000/10/swap/util/sniffSchema#domainIntersects is an unrecognized or invalid identifier. Cannot create.

05:43:00 <DanCon> The resulting exception was:

05:43:00 <DanCon> java.lang.NullPointerException

05:43:00 <dc_rdfig> Label ERROR not found.

05:43:00 <DanCon> ]]]

05:43:12 <DanCon> dc_rdfig, shut up

05:44:35 * DanCon was all excited to have enough java horsepower to run RIC; is kinda disapponted that he can't get RIC to eat his shiny new schema

05:46:02 * DanCon tries # http://iama.rrecktek.com/daml/ont/agenda-ont

05:47:00 <DanCon> uh... now I have some classes; now what?

05:47:40 * DanCon reads http://www.mindswap.org/~mhgrove/RIC/readme.txt

05:48:51 <DanCon> the "Use Object" button doesn't do anythying

05:49:15 <golbeck> the classes of your ontology should appear in that window

05:49:22 <golbeck> you click one and then click Use Object

05:49:24 <DanCon> I see some classes.

05:49:30 <DanCon> I tried that. nothing happens

05:49:36 <golbeck> If you swtch to the active data area, you'll see root changed to a folder

05:49:39 <golbeck> it's subtle

05:49:49 <golbeck> if you double click the folder, you'll see the objects you've created

05:49:54 <DanCon> aha...

05:50:00 <golbeck> then, click an object to give it an ID

05:50:12 <golbeck> if you click the down arrow by the object, you'll see the properties

05:50:20 <golbeck> click one of those, and you'll get the whole form for properties

05:52:41 <DanCon> how do I connect one object (say, a Day) to a property of another (say, a Meeting?)?

05:52:59 <golbeck> well, that depends...

05:53:11 <golbeck> is there a button that says Use Object

05:53:32 <golbeck> in the form next to Day

05:53:43 <DanCon> yup

05:53:44 <golbeck> or next to Meeting (where you would specify Day)?

05:54:11 <DanCon> in the Meeting object dialog, next to day, there's a "Use object" button

05:54:38 <golbeck> so click on an object in the Class View Window

05:54:40 * DanCon thought he saw jhendler drag-n-drop stuff.

05:54:46 <golbeck> Then click Use Object in causes Event

05:54:52 * DanCon clicks on sat Day

05:54:56 <golbeck> hm...probably in SMORE, our HTML editing RDF tool

05:55:15 <golbeck> sorry... click Use Object in the form

05:55:23 <golbeck> next to Meeting , in your case

05:55:29 <golbeck> you'll see it grey out

05:55:38 <DanCon> uh... when I click on the day object in the active data area, the use object button goes away; the workspace is replaced by something showing the Day id

05:55:41 <golbeck> then, if you go back to the active data area, you'll see a little forder there

05:55:54 <golbeck> right. you can click on that day and fill out its own form

05:56:14 * DanCon filled out the sat Day object already

05:56:31 <golbeck> oh.

05:56:51 * DanCon wants :mm1 :day :sat. :sat a :Day.

05:57:10 * DanCon has :mm1 a :Meeting. :sat a :Day.

05:57:22 <golbeck> i dont know how to do it that way....or even if you can.

05:57:33 <golbeck> i've only done it working down through it.

05:57:42 <golbeck> make a meeting. that needs a day....then go make the day object

05:57:46 <DanCon> er... ok, what's the known-to-work way?

05:58:05 <golbeck> select Meeting in your Class View and click Use Object

05:58:16 <golbeck> go to the properties of the Meeting

05:58:21 <golbeck> click on Day in your Class VIew

05:58:39 <golbeck> then click Use Object next to the Meeting button in your Workspace

05:58:48 <DanCon> aha!

05:59:11 <golbeck> like i said, i'm not sure if there is a way to do references to existing objects here.

05:59:26 <golbeck> there is in SMORE, which works with drag n drop (probably what jim did)

05:59:50 <DanCon> I'm pretty sure he was using RIC; maybe he wasn't doing drag-n-drop

06:00:19 <DanCon> can I have the same Day object be a day value of 2 different meetings? or does RIC only grok trees?

06:00:52 <DanCon> also: is there a way to save RIC's state, so I can come back later?

06:02:32 <golbeck> there isn't a save yet - it's on the TODO list, though

06:02:41 <DanCon> I see.

06:03:02 <DanCon> ok, well, I think I get how to use it now. Thanks.

06:03:32 <golbeck> and from what i know, RIC only does trees. if you could do references to other objects generally, that would allow the same day for 2 meetings. but there's no support for that from what i can tell

06:03:47 <golbeck> mike grove is the expert (and author), so he may know some things i don't yet.

06:04:29 <golbeck> the focus of development has been adding more features of daml and rdf (support for oneOf, cardinality restrictions, etc), as opposed to interface improvements

07:00:17 <ChanServ> [#rdfig] This channel is logged and blogged: http://logicerror.com/rdfIRCWelcome

08:07:01 <burtonator> So what do you guys think of an RDF Schema vs XML Schema smackdown?!

08:47:18 <DMiles-TheLaptop> DMiles-TheLaptop is now known as mea_culpa_away

09:20:53 <danb_nothere> danb_nothere is now known as danbri

09:57:17 * shellac is away: I'm busy

10:40:04 * shellac is back (gone 00:42:47)

11:21:20 <dajobe-lap_> hmm

11:22:30 * shellac is away: lunch

11:56:57 * shellac is back (gone 00:34:26)

13:28:58 * DanCon waves from 1st telcon of the day...

13:41:43 * em tests the collaborative brain power of the group by posing an interesting question that he found in his inbox...

13:42:05 <em> question: "is anyone doing any work with RDF in Linda/TSpaces/JavaSpaces? Seems like sort of a natural fit, since they work with tuples."

13:42:08 * xover points DanCon at http://www.apple.com/ical/ - cf. Calendar scraping and syncing... :-)

13:42:54 <em> xover, that warrents a chump, please do so :)

13:43:06 <xover>http://www.apple.com/ical/

13:43:06 <dc_rdfig> C: http://www.apple.com/ical/ from xover

13:43:16 * em thanks xover

13:43:29 <xover> C:| iCal - Calendar Software From Apple

13:43:29 <dc_rdfig> titled item C

13:43:39 <pixel> wow. that's pretty.

13:44:11 <xover> C: Has the ability to keep multiple calendar schedules and synchronize them over the web.

13:44:11 <dc_rdfig> added comment C1

13:45:12 <xover> C: The synchronization probably makes use of WebDAV as the transport and either iCal or SyncML, or both, as the format.

13:45:12 <dc_rdfig> added comment C2

13:45:27 <DanCon> apple's iCal is out?

13:45:36 <xover> Nah, not til September.

13:45:52 <DanCon> bummer about the name; to me, ical means RFC2445

13:46:00 <xover> Yup, me too. :-(

13:46:23 <xover> Hopefully they'll have sufficient shame to at least use iCal as the exchange format for iCal[tm].

13:47:06 <kao> em: I looked at tuple systems briefly

13:47:22 <kao> Usually they require a certain format for things like labels, which often clash with URI syntax.

13:47:48 <kao> Like, they dont allow you to use : / # and other URI rreserved characters.

13:48:20 <xover> chris_s: Can publish calendars to the web, but may try to "nudge" you into using .Mac (aka. iTools).

13:48:36 <kao> But if youre willing to modify existing systems a bit to allow this, you could possibly use them.

13:48:42 * xover curses nickname expansion. (sorry chris_s)

13:49:18 <xover> C: Can publish calendars to the web, but may try to "nudge" you into using .Mac (aka. iTools).

13:49:18 <dc_rdfig> added comment C3

13:50:36 <xover> C: For those using Max OS X (10.2) and a suitable cell phone, provides the synchronization with your phone, handheld, or even your iPod (!) for free.

13:50:36 <dc_rdfig> added comment C4

13:51:41 <deltab> xover: typo

13:52:00 <xover> C4: For those using Mac OS X (10.2) and a suitable cell phone, provides the synchronization with your phone, handheld, or even your iPod (!) for free.

13:52:00 <dc_rdfig> replaced comment C4

13:52:06 <xover> Thanks deltab! :-)

13:52:12 <deltab> :-)

13:53:01 <em> keo, was your review specifically in terms of Linda/TSpaces/JavaSpaces, anything you can point me to?

13:53:11 * em thanks for the feedback

13:53:52 <xover> C: Utter vapourware so far (though the demo was pretty cool), but if Apple decides to make this tech reasonably open, this has lots of potential for making my life a *lot* easier!

13:53:52 <dc_rdfig> added comment C5

13:54:38 <xover> There. That ought to be enough free advertizing for "Jag-wire". :-)

14:01:50 <kao> em: no, sorry. It has been a while ago. I dimly remember the naming problem with some tuple system, but i cannot remember anythign more specific.

14:02:27 <eikeon> em: I have thought some about how TSpaces relate to triple stores. If I remember my thoughts correctly, the main difference to me was in regaurds to the closed world assumption.

14:03:06 <DanCon> hmm... just when I thought RDF calendaring was starting to come together, the market for it may be gone.

14:03:36 * JibberJim thinks many people don't use Mac OS

14:05:36 <xover> I see Apple's calendar stuffing iCal format to a WebDAV server; from which you can republish as RDF or even Microsoft's proprietary format (Exchange).

14:07:30 <xover> I can take care of my phone/handheld/whatnot from the Mac; and then hopefully find some way for Evolution to talk to that WebDAV server.

14:42:38 * DanCon really likes WebDAV integration in OSX; uses Zope; wonders if RDF could be integrated at that level...

14:46:06 <xover> Heh! I mentioned repurposing the calendar as RDF on a mac newsgroup and everyone thought I was talking about "Reality Distortion Field". :-)

14:51:03 <mhgrove> this may sound stupid, but is rdf:Description defined anywhere?

14:59:01 <mhgrove> mhgrove is now known as mhgrove-afk

15:07:54 <mhgrove-afk> mhgrove-afk is now known as mhgrove

15:29:23 <mea_culpa_away> mea_culpa_away is now known as mea_culpa

15:55:53 <sandro> DanC, you were thinking about moving from Linux to OS X a while back. How did that pan out? I need a new notebook, and am wondering about going Apple..... It'll be cool, but I don't know how much grief it'll be. At least it'll mean two sets of binaries.

15:56:46 <xover> sandro: Worst case, I hear the TiBook makes a bitchin' Linux platform. :-)

15:57:56 <sandro> I'd rather have binary compatibility, if the Mac apps wont be available to me..... But yeah.

16:02:20 <eikeon> Okay... http://eikeon.com/foaf/ now has a way of browsing starting from a given mbox; a way of submiting a foaf url for it to spider (or respider);

16:02:37 <DanCon> I went with a PC for work; we since got a Mac for home. love it.

16:03:05 <eikeon> Wrong channel again... need to look before typing :(

16:03:15 <JibberJim> nah eikeon, chump it, it's good.

16:04:15 <eikeon> Need to go now... will do when I get back.

16:06:24 <sandro> Have you tried doing any work stuff on the Mac? emacs, web browsing, python-in-shell, all that stuff works fine?

16:08:55 <xover> emacs in Terminal.app works fine, you need the somwhat clunky X server running to get GUI.

16:13:00 <sandro> "clunky X server" doesn't sound so good.

16:13:16 <sandro> You mean the normal windowing system does not include X? Hrm.

16:13:19 <bijan> That's the only kind.

16:13:28 <xover> It runs as an application on top of Mac OS X' normal GUI.

16:13:38 * bijan is running at least three completely different windowing systems on the same box.

16:13:48 <JibberJim> you mean the emacs you get doesn't work in normal OS X ?

16:13:54 <xover> X == Xfree86, here BTW.

16:14:02 <bijan> MacOSX/Aqua, XDarwin with OroborX, and Squeak.

16:14:18 <bijan> None of them use scroll bars consistently :)

16:14:39 <xover> Emacs is included with the OS, but you need to run it in terminal mode unless you install Xfree86.

16:14:54 <bijan> Even then, I think the included one is x free.

16:15:04 <bijan> I had to fink another ojne.

16:15:10 <bijan> Then I saw gnu emacs in x.

16:15:16 <bijan> And finked xemacs :)

16:15:33 <xover> Hmmm. Possibly. But.. Yes, Fink makes that a breeze.

16:15:45 <bijan> Not over flaky dialup :)

16:15:56 <xover> Heh, guess not. :-)

16:16:29 * xover thanks $deity for the 1Mbps wireless connection to home!

16:18:55 <xover> But, BTW, you _can_ run the X server rootless so it /looks/ like it's part of the normal GUI.

16:22:19 <bijan> Yes.

16:22:20 <bijan> I do.

16:22:27 <bijan> But it's not quite right.

16:22:42 <bijan> OroborX doesn't fix the evil scroll bars.

16:22:51 <bijan> But it *does* have windowshade...

16:22:59 <bijan> ...which my Aqua stuff doesn't yet :(

16:23:13 <bijan> So someplaces I double click and the window rolls up.

16:23:16 <bijan> Other times it shrinks.

16:23:22 <bijan> Anyhoo. Time for phone calls ;(

16:23:38 <xover> There's a third-party hack to fix that. WindowshadeX or somesuch IIRC. :-)

17:33:35 <sandro> Is anyone aware of any serious follow-up to McDermott et al's work on layering? That work disclaims co-existance with "RDF in the wild", which IMO means it lives in Dark Triples land. Has anyone tried to get it to coexist with real (MT) RDF?

17:37:35 <pixel> is that like Dark Matter?

17:37:51 <dajobe> sandro: I think you need to look at L(base)

17:44:28 <sandro> Funny, think, dajobe, I was doing that (again) while you were sending that message. :-)

17:44:53 * sandro groks and has some quibles with http://tap.stanford.edu/sw/swmt.html. But generally it's on the right track. Is there a newer version?

17:46:03 <sandro> The layering problem, IMO, is a different one, though. Closely related. I've been using the L-Base approach (also using fol-with-equality) to find (well, look for) the paradoxes in layering.

17:46:32 <sandro> s/Funny, think/Funny thing/

17:47:57 <sandro> Biggest issue is probably with the overly simple treatment of URI-Refs and Literals. I think L-Base needs to be able to peer into their structure (which is can do, if they are composed axiomatically).

17:49:17 <sandro> The layering paradoxes are about RDF vocabularies which license inferences about RDF triples -- which very quickly leads to a liar paradox, if you're not careful.

17:50:30 <sandro> (or other, more subtle paradoxes.) I think if you axiomatize them in L-Base, though, ATP's should be able to help keep things clean. (as stated in swmt.html section 2.1, which is basically what I've been doing.)

17:50:52 <sandro> (... although I'm hoping to be able to prove consistency by actually finding a model.)

17:50:58 <bijan> Prolly russell's paradox is more common in these layerings.

17:51:10 <bijan> because of all the class stuff in RDFS.

17:51:20 <sandro> If you're using RDFS, perhaps.

17:51:41 <bijan> Well, sure. :)

17:51:52 <bijan> I tend to think of RDF as RDF(S) now adays...

17:51:57 <bijan> Jim's fault.

17:52:13 <sandro> Ah. Tim (and I) don't. I think of RDF as just triples.

17:52:22 <bijan> So no rdf:type?

17:52:37 <sandro> rdf:type is just one vocabulary term in common usage.

17:52:47 <bijan> Hmm.

17:53:02 * sandro has to run, now that Bijan has showed up. :)

17:53:06 <bijan> But that doesn't mean it doesn't have a semantics :)

17:53:11 <bijan> And that's the issue.

17:53:24 <bijan> They aren't *just* vocab terms...but vocab terms with certain meanings.

17:53:24 <bijan> Bye!

17:54:04 <mdupont> Anyone working on an SQL/RDF gateway?

17:54:26 <bijan> Good article on that in the ISCW proceedings.

17:54:35 <bijan> About compiling RuleML to SQL (and java)

17:54:42 <bijan> Maybe not exactly what you're after?

17:54:46 <larsbot> bijan, URL?

17:54:53 <JibberJim> mdupont - http://swordfish.rdfweb.org/rdfquery/ ??

17:55:02 <larsbot> it's exactly what I am after :)

17:55:07 <bijan> Heh.

17:55:11 <mdupont> Thanks.

17:55:16 <bijan> WEll, you need links acces to download the pdf.

17:55:26 <bijan> iswc.semanticweb.org

17:55:26 <larsbot> what does that mean?

17:55:29 <bijan> click on proceedings.

17:55:43 <bijan> Well, I can get the papers becaues UMD has an agreement with springer.

17:56:19 <bijan> The title is:

17:56:40 <bijan> "Automatic Generation of Java/SQL Based Inference Engines from RDF Schema and RuleML"

17:56:49 <bijan> By Andreas Eberhart.

17:57:00 <bijan> You might find a draft or other info on his site.

17:57:11 * larsbot will try that, 'cuz Ontopia does not have links access

17:57:24 <bijan> I highly recommend the ISWC 2002 proceedings. Worth getting a copy.

17:57:30 <bijan> It's chock full of good stuff.

17:57:51 <bijan> And often more polished than the otherwise available versions.

17:57:53 <larsbot> I'll try that

17:57:58 <larsbot> thanks a lot for the references :)

17:58:02 <bijan> No problem.

17:59:21 <mdupont> My issue is to take a set of Postgres SQL tables and try and reverse engineer them into some more structured format

17:59:36 <mdupont> maybe some RDFS like description of them

17:59:51 <bijan> Ah yes.

17:59:59 <mdupont> Perl DBI has some good Database Model extrator code.

18:00:05 <bijan> Just was working on a XMl Schema -->DAML+OIL thing.

18:00:06 <bijan> by hand.

18:00:09 <bijan> Ouch.

18:00:18 <mdupont> Cool, using XLST?

18:00:23 <bijan> No.

18:00:25 <bijan> "by hand" :)

18:00:29 <mdupont> Ahh

18:00:48 <mdupont> Well, thats impressive. bet you are learning alot.

18:00:50 <bijan> WE really wanted the model, not just the document model.

18:00:52 <bijan> Yes.

18:00:55 <bijan> About Pain :)

18:00:58 <mdupont> heheh

18:01:00 <mdupont> lol

18:01:43 <mdupont>http://www.ebi.ac.uk/~ugis/archive/LSR-GE/msg00187.html

18:01:43 <dc_rdfig> D: http://www.ebi.ac.uk/~ugis/archive/LSR-GE/msg00187.html from mdupont

18:01:43 <bijan> There's a pretty cool daml mode for emacs.

18:02:29 <mdupont> D:| Announcement of the XMI parser in perl for the microarray-format Genetic Encoding project

18:02:29 <dc_rdfig> titled item D

18:03:10 <mdupont> datum: This is very similar to what I am doing with the http://introspector.sf.net except they use a standard XMI format

18:03:13 <mdupont> D:This is very similar to what I am doing with the http://introspector.sf.net except they use a standard XMI format

18:03:13 <dc_rdfig> added comment D1

18:03:24 <mdupont> D:But XMI looks really nasty.

18:03:24 <dc_rdfig> added comment D2

18:04:44 * mdupont wonders if the XMI parser from sergey is available

18:05:03 * mdupont thinks it must be part of the RDFAPI that he posted.

18:07:34 <mdupont> bijan: does the daml model need w3 mode?

18:07:45 <bijan> No.

18:07:45 <mdupont> it must be time for me to install w3 :)

18:07:49 <bijan> It can use it.

18:07:57 <bijan> And its *very* nice to have.

18:08:26 <bijan> warning: I've only just installed daml mode last night and haven't really banged on it.

18:08:49 <mdupont> ok, well I will go through the pain

18:09:04 <mdupont> if i can only remember where i downloaded it to

18:09:28 <bijan> Install was easy.

18:09:34 <bijan> It's all easier if you're using xemacs :)

18:09:36 <mdupont> Do you know what the differnece between the redland and the RDFStore and RDF API is?

18:10:08 <mdupont> xemacs? I was thinking about getting into that, but I am too used to emacs...

18:10:15 <mdupont> plus 21 is pretty nice

18:10:26 <bijan> Er...welll....ok.

18:10:39 <mdupont> I just wish that stallman was not so crazy

18:10:51 <mdupont> hehe

18:11:12 <mdupont> is there so much a difference?

18:11:17 <bijan> I think so.

18:11:21 <bijan> In X mode certainly.

18:11:32 <bijan> I'm really just getting started with emacs in general.

18:11:46 <bijan> But emacs in X is...primative.

18:11:54 <bijan> xemacs is slicker on a lot of fronts.

18:11:58 <bijan> Anyhoo.

18:12:03 <bijan> bijan-->meeting :(

18:12:19 <mdupont> OK, well I will check out daml mode.

18:12:24 * sandro returns, since Bijan is leaving. :-)

18:12:32 <bijan> Coward!

18:12:42 <sandro> I know I can't run forever. :-)

18:12:54 <bijan> I know where you live!

18:12:59 <bijan> Better, I know when you lurk :)(

18:13:25 <sandro> There's a really, really, realy interesting thread we just started: what L-Base axioms are you accepting when you use a vocabular term?

18:13:38 <bijan> On www-rdf-logic?

18:13:54 * bijan not yet on all the mailing lists he should be.

18:14:02 <sandro> no - you and I - in discussing whether rdf:type is part of RDF.

18:14:06 <bijan> hehe.

18:14:11 <mdupont>http://programmierer.freepage.de/null/nyktop/

18:14:12 <dc_rdfig> E: http://programmierer.freepage.de/null/nyktop/ from mdupont

18:14:25 <mdupont> E:| Nytop is a code generator in perl

18:14:25 <dc_rdfig> titled item E

18:14:30 <bijan> Well, my cheap answer is that rdf:type and rdf:Property are part of what RDF MT call RDF entailment.

18:14:49 <bijan> As opposed to simple entailment, which doens't consider the meaning of *any* terms

18:15:02 <sandro> who/what dictates the semantics of rdf:type? What about some other property (perhaps at geocities :-)

18:15:03 <bijan> I.e, is entailment just considering the graph as such.

18:15:17 <mdupont> geocities?

18:15:27 * mdupont wonders what sandro means

18:15:41 <bijan> patH.sexylogician.freerantpages.geocities.com

18:16:39 <mdupont> sexy?

18:16:51 <bijan> Well, we logicians take what props we can get.

18:17:03 <bijan> "On the web, nobody know that you're not sexy"

18:17:13 <bijan> Ok, meeting for real. Ta.

18:17:42 <sandro> well, if smart is sexy, .... :-)

18:18:31 <mdupont> Are you also offended by IBM alphaworks? does anyone use stuff from them?

18:19:14 <sandro> What I mean is that the whole design of RDF (using XML namespaces) has been a ridiculous exercise in stupidity if we go ahead and make rdf:type magic. The POINT is to NOT privilege any particular vocabularies.

18:19:40 <sandro> is who offended by IBM alphaworks?

18:20:03 <sandro> (I used jikes (from alphaworks) heavily when I used to use Java.)

18:22:12 <mdupont> Sure, but jikes is not one of the 90days test version or pay is it?

18:23:00 <sandro> I've never heard of something like that.

18:24:11 * mdupont hopes that he can find the link

18:24:17 <danbri> It isn't 'magic', but it is privileged.

18:24:27 * mdupont wonders why he spends so much time searching

18:25:23 <danbri> We don't want everyone creating their own variant for common notions like category membership, so we give them rdf:type. It's privileged in that it is named in a W3C spec, and has syntax stuff (typedNode) and vocab description stuff (range/domain) associated with it.

18:25:26 <mdupont> *alphaWorks technologies are available for evaluation use as always. You can download and use any alphaWorks technology for 90 days from http://www.alphaworks.ibm.com/tech/

18:25:47 <mdupont> As i said, offensive.

18:25:50 <sandro> The privilege of being W3C Recommended, sure, that's fine.

18:26:08 <sandro> And the W3C Recommended Syntax special cases it; that's also fine.

18:27:36 <sandro> The question is which L-Base Axioms go along with it? And, more to the point, how do users of other vocabulary terms specify their semantics (with L-Base axioms, PTL axioms, or whatever)....? That's the problem we need to solve -- then it becomes more clear than rdf:type is nothing more than "Hey, let's everyone try to use this same name for the same concept".

18:28:51 <sandro> To obvious answer: HTTP GET to get the axioms only works if you have strict layering, (Ontology A using Ontology B using Ontology C, with C never using A, etc). Maybe that's okay.

18:30:05 <sandro> When I visited ibm.com this morning (as a potential laptop customer), I was surprised their most eye-catching headline was about their profits. Hardly the thing one usually brags about to a customer.

18:30:33 <sandro> ... but I guess it does suggest they are smart & stable in the current economy.

18:31:12 <mdupont> Well, i think that it is very difficult to use any of the alpha-works stuff in an opensource /free project

18:31:27 <mdupont> so it remains basically just a temptation not a help....

18:31:34 <mdupont> just my 2 cents.

18:31:48 <sandro> Yeah. Painful.

18:32:06 <sandro> Have you told them you're annoyed?

18:32:27 <mdupont> i am going to try and compile this stuff from sergey with the gnu java compiler

18:32:32 <mdupont> so i dont have to install java

18:32:36 <mdupont> :)

18:32:53 <mdupont> I think the RDF java stuff would be a good target for the gjc

18:33:11 <dajobe> you not looking at Jena?

18:33:24 <mdupont> i have not installed java at all right now

18:33:31 <mdupont> just perl,python and gcc

18:33:46 <mdupont> oh and dotgnu c# compiler

18:33:55 <dajobe> well, there are RDF things for all of them

18:34:05 <mdupont> I have heard good things about jena

18:34:18 <mdupont> and i consider installing java for that

18:34:21 <dajobe> are you still making a tool choice?

18:34:24 <dajobe> sure, good idea

18:34:25 <ssyreeni> eikeon: does your FOAF spider support mbox_sha1sum?

18:35:01 <mdupont> well I dont know if I should use REDLAND which supports libxml that will give me compatibility with gnome

18:35:16 <mdupont> and has a perl layer

18:35:32 <mdupont> or RDFStore which is compatible with sergeys RDF API

18:35:37 <dajobe> yeah, you do know I wrote redland?

18:36:02 <dajobe> sergey's rdf api hasn't changed much recently compared to jena which has a lot more features

18:36:08 <mdupont> dajobe: we have chatted before

18:36:32 <mdupont> I really would like to have an API for C programs to exchange with perl and python

18:36:56 <mdupont> and the dotgnu have been working on a libxml in c# binding

18:37:21 <mdupont> so i think that it could be used to provide a transport layer for comüöplex data bindings

18:37:30 <mdupont> for comüöplex data bindings

18:37:40 <mdupont> for complex data bindings

18:37:58 * mdupont wonders why his keys are getting mangled

18:38:14 <mdupont> dajobe: do you think that i am save with redland/raptor api?

18:38:26 <mdupont> s/save/safe/

18:38:31 <dajobe> safe in what sense?

18:38:57 <mdupont> Well, is there a standard API? how different are these all,

18:39:07 <mdupont> I am a little confused where to start.

18:39:14 <dajobe> there's no standard RDF API across all languages

18:39:19 <mdupont> at least in c?

18:39:21 <dajobe> but you might consider there are defacto ones

18:39:28 <dajobe> in C, no

18:39:41 <mdupont> So raptor/redforo is my best bet?

18:39:51 <mdupont> redland

18:39:54 <dajobe> I've got the biggest C impl at the moment (that is public)

18:40:05 <mdupont> cool, i am sold. I have already compiled it.

18:40:21 <dajobe> the w3c's libwww has some rdf support for parsing, but most of it is for general web use

18:40:36 <mdupont> I want to replace my xml dumping module that I build into the gcc with the libxml/raptor API

18:40:59 <mdupont> that will allow for hooking up the perl and python modules directly across the API

18:41:10 <mdupont> without using external files.

18:41:12 <dajobe> there's also libgnurdf which is a small library for parsing manipulating fixed model used for the software packaging project

18:41:15 <dajobe> cool

18:41:26 <ssyreeni> ok. thanks to yesterdays discussion over daml:UnambiguousProperty et al, I had to take a braindump. serious amateur stuff, but still...

18:41:34 <mdupont> is that from dan v. for the rpm tool?

18:41:53 <dajobe> separate project, www.gnupdate.org

18:42:03 <dajobe> similar idea though

18:42:24 <mdupont> I also want to add in an RDF dump into debian-sf/savannah

18:42:27 <ssyreeni>http://www.iki.fi/~decoy/texts/rdf/unproperties.en.html

18:42:29 <dc_rdfig> F: http://www.iki.fi/~decoy/texts/rdf/unproperties.en.html from ssyreeni

18:42:49 <ssyreeni> A:|Naming them DAML unproperties

18:42:50 <dc_rdfig> titled item A

18:43:15 <dajobe> howso a dump of debian-sf/savannah ?

18:43:16 <ssyreeni> A:A few thoughts about yesterdays daml:Unambiguous et al. discussion

18:43:17 <dc_rdfig> added comment A4

18:43:35 <ssyreeni> F:|Naming them DAML unproperties

18:43:36 <dc_rdfig> titled item F

18:43:42 <ssyreeni> F:A few thoughts about yesterdays daml:Unambiguous et al. discussion

18:43:43 <dc_rdfig> added comment F1

18:43:48 <mdupont> Well I have adding the XML from sourceforge 2.5 dump code back into debian-sf

18:43:49 <ssyreeni> one can be stupid but not that stupid.

18:44:04 <mdupont> and with lots of help have gotten the server running

18:44:15 <dajobe> ssyreeni: you just destroyed A:'s title

18:44:31 <ssyreeni> that's why I called myself extremely stupid.

18:44:37 <dajobe> A:| Semaview Inc web page

18:44:37 <dc_rdfig> titled item A

18:44:49 <mdupont> But I have found that RDFS/DAML really provides what i promised myself from XML

18:44:54 <dajobe> you can delete things with A4:""

18:45:33 <mdupont> The semantic web is the best plaform for metadata extraction and usage that I have found.

18:46:11 <mdupont> I mean the usage of metadata.

18:46:29 <dajobe> I tend towards the metadata ends of things in the semweb, often forgotten ;)

18:47:05 <mdupont> The coopx project hopes to define an standard package format

18:47:36 <mdupont> It has not published a fileformat, but i think that RDF/DAML will be the best bet.

18:47:49 <mdupont> it lends itself to describing projects and files,

18:48:01 <mdupont> just look at the RPM stuff dan was doing.

18:48:11 <dajobe> DanV - yeah, I used it years ago

18:48:20 <dajobe> for http://www.mirror.ac.uk/

18:48:26 <dajobe> the search system there still might be using it

18:49:38 <dajobe> heh, the cgi params say yes they are

18:49:53 * dajobe wanders off for a bit

18:52:00 <ssyreeni> A4:""

18:52:00 <dc_rdfig> deleted comment A4

18:53:16 <ssyreeni> dajobe: fixed. sorry again.

19:01:26 * mnot was reading GK's http://lists.w3.org/Archives/Public/www-tag/2002Jul/0020.html

19:02:01 <mnot> I found it counterintuitive; if I put my REST glasses on, a URI identifies a resource, a URI-Reference identifies a representation or part thereof.

19:02:56 * mdupont is following the link wondering what a URI ref is?

19:09:28 <mdupont> Do you have a definition of a fragment id?

19:10:26 <connolly> mnot http://www.w3.org/ is a URI reference; you suggest it identifies a representation?

19:10:36 <connolly> connolly is now known as DanC_

19:11:37 <mdupont> - when used in an rdf document, someurl#frag means the thing that is

19:11:37 <mdupont> indicated, according to the rules of the application/rdf+xml mime type as a

19:11:37 <mdupont> "fragment" or "view" of the RDF document at someurl. If the document

19:11:37 <mdupont> doesn't exist, or can't be retrieved, then exactly what that view may be is

19:11:37 <mdupont> somewhat undetermined, but that doesn't stop us from using RDF to say

19:11:39 <mdupont> things about it.

19:11:46 <mdupont> Hey DanC_

19:13:04 <DanC_> "doesn't stop us..." exactly; are those your words, or Grahams, mdupont?

19:14:37 <mdupont> That was a quote from GK'S post

19:15:03 * mnot was reading GK's http://lists.w3.org/Archives/Public/www-tag/2002Jul/0020.html

19:16:01 <DanC_> it's actually according to the rules of whatever MIME type you get back when you GET someurl, but the point stands.

19:16:36 * DanC_ has come to the conclusion that there are formal documents and informal documents, and one shouldn't use conneg to go between them, as rule.

19:17:08 <mdupont> conneg?

19:17:20 <DanC_> HTTP contente negotation

19:17:39 <DanC_> i.e. you shouldn't serve myDoc.html and myDoc.rdf under the same name, myDoc

19:18:01 <DanC_> ala http://www.w3.org/2000/04/maillog2rdf/email

19:18:28 <JibberJim> Are you saying http://www.w3.org/ is not a resource it purely denotes a representation, and if so how we can ever use it alone as an identifier in RDF?

19:18:50 <DanC_> the exception is: when the .html version says, in english, exactly and only those things said in the .rdf. i.e. when the HTML doc is written in very stilted, mathematical style; or is just a big table or something.

19:18:58 <dajobe> phew

19:19:02 <dajobe> logger, pointer?

19:19:04 <dajobe> See http://ilrt.org/discovery/chatlogs/rdfig/2002-07-18#T19-19-02

19:19:10 <JibberJim> because you do not give sufficient information to say which particular representation you're talking about.

19:19:13 <dajobe> 'cos that does what you just said

19:19:55 * DanC_ is confused by dajobe's remarks

19:20:18 * mdupont feels like he is drowning

19:20:32 <dajobe> the uri (name) above http:.../2002-07-18 connegs rdf, html or txt

19:20:35 * mnot wonders back

19:21:08 <mnot> it seems that conneg introduces a lot of problems all around (the caching world hates it)...

19:21:12 <xover> But that does satisfy DanC_'s "formal, stilted, etc." comment above!

19:21:28 <dajobe> mnot: cool! caching sucks

19:21:44 <DanC_> hmm... in order for the HTML version of these logs to say the same thing as the RDF version says, each of our comments would have to go in <blockquote> or some such.

19:21:55 <DanC_> caching is cool, and conneg is cool.

19:22:05 <mnot> DanC: re http://www.w3.org/ being a URI-Reference - true. I suppose RDF needs something out-of-URI to indicate whether it's a resource or representaion that is being talked about.

19:22:22 <dajobe> (oh, I'm just ranting about so-called transparent caches again. Ignore this thread)

19:23:23 <mnot> The only problem that transparent caching breaks is that it violates the trust model at the IP layer; i.e., your browser says "go to 1.2.3.4" and the router grabs it and sends it to 5.6.7.8.

19:23:47 <xover> DanC_: ref vs. res seems to me to spring from a lack of "Accept-Foo" in the _URI_; except informally by way of ".html".

19:23:53 <DanC_> RDF is always talking about resources. (note that documents/representations are resources too, e.g. in the case of http://www.ietf.org/rfc/rfc0822.txt , that URI denotes a resource, which happens to be the same as the representation you'll GET when you dereference it)

19:23:56 <mnot> I agree that conneg can be cool, but there is a lot of fuzziness around when to use it.

19:24:08 <mnot> s/transparent/interception/

19:24:15 <DanC_> fuzz: quite.

19:25:05 <DanC_> (hmm... actually, the RFC822 resource might be considered more than a sequence of bytes; it's a product of a consensus process)

19:25:08 <mnot> DanC: yes, but that information isn't captured any where; you know that they're the same because you make an assumption

19:25:33 * mdupont wonders if load-path has changed in gnuemacs21

19:25:41 <DanC_> ah; yes, we really need a bunch of utility RDF properties to relate all this stuff, mnot.

19:26:08 <xover> consensus process: Heh! Try /describing/ that process in any way that's usefull in the context! :-)

19:26:18 * DanC_ has a TAG action to formalize some URI scheme properties in RDF. (!) considers sneaking in some modelling-HTTP-in-RDF while he's at it

19:26:25 <mnot> I was asking a long time ago about whether someone had come up with properties to describe the permanence of a resource. There's something in DesignIssues IIRC

19:26:49 <mnot> Thinking of using Expires?

19:27:18 <DanC_> some modelling-HTTP-in-RDF work: http://www.w3.org/2001/03swell/http.n3

19:28:55 <mnot> The 822 resource *is* more than a sequence of bytes... it just happens that all representations returned by it are always the same sequence of bytes. There may be other resources that return the same sequence of bytes, and it would be nice if there were a way to say that they're the same in the representation (HTTP headers or HTML HEAD markup), in addition to out-of-band RDF describing it.

19:32:02 <DanC_> "The 822 resource *is* more than a sequence of bytes" <- I don't think that's observable/provable.

19:33:18 <mdupont> DamlMode is working!

19:33:21 <DanC_> i.e. I don't think you can derive a contradiction from { <http://www.ietf.org/rfc/rfc0822.txt> = " \n\n\n\n RFC # 822 ..." }

19:33:49 <mnot> As soon as you make an assertion about it beyond that, it's more than the sequence of bytes, in the context of the world defined by RDF.

19:34:01 <mnot> And the HTTP headers make assertions...

19:34:24 <DanC_> I can make lots of assertions without contradicting the = there: I can say I like it; I can say it has NNN bytes.

19:34:47 <DanC_> which HTTP header contradicts the = there?

19:35:11 <JibberJim> Vary: ?

19:35:16 <sbp> you can't say { " RFC text here" :format "text/plain" }

19:35:17 <mnot> what are the semantics of = in n3?

19:35:47 <DanC_> = in N3 is like = elsewhere in math: <foo> = <bar> means that <foo> and <bar> denote the same thing.

19:35:48 * mnot is getting SW rusty

19:36:09 <sbp> say I have a document "<doc/>" served as text/plain, and another of the same served as text/xml

19:36:42 <DanC_> well, that's not a contradiction yet, sbp.

19:37:02 <sbp> using your <URIforDoc> = "<doc/>" thing, you'd end up with "<doc/>" :format "text/plain", "text/xml" . which seems inconsistent and not very useful to me

19:37:19 <mnot> Is it legal to say <foo> = Literal?

19:37:21 <DanC_> not useful: could be. inconsistent? I don't see P and not(P) yet.

19:37:33 <sbp> well, I think it's a bad model

19:37:37 <deltab> doesn't that just modify the meaning of :format ?

19:37:38 <DanC_> yes, it's legal, mnot.

19:37:58 <sbp> just because I can't come up with a contradition of the P !(P) variety, doesn't mean to say that there isn't one

19:38:01 <DanC_> bad model, maybe. but that wasn't my point.

19:38:03 <deltab> so that :format would mean "can be interpreted as" or something like that

19:38:09 <mnot> Is = transitive in n3?

19:38:19 <DanC_> yes, = is transitive

19:38:26 <sbp> = is syntactic sugar for daml:equivalentTo

19:38:30 <DanC_> yes, sbp

19:38:50 <JibberJim> What if I have it available in 2 charsets?

19:39:19 <DanC_> you don't, JibberJim. I chose RFC822 on purpose because it has exactly one byte-sequence representation.

19:39:35 <mnot> so <http://www.a.com/> = "bag of bytes" = <http://www.b.com> , but <http://www.a.com> :hostedAt "a.com" while <http://www.b.com/> :hostedAt "b.com".

19:39:48 <DanC_> still no contradiction, mnot.

19:39:57 <DanC_> and I didn't say <http://www.a.com/> = "bag of bytes" = <http://www.b.com>

19:41:52 <sbp> :XMLDocument daml:disjointWith :TextDocument . { ?x :format "text/xml" } => { ?x a :XMLDocument } . { ?x :format "text/plain" } => { ?x a :TextDocument } . <http://somedocument.../> = "<doc/>"; :format "text/plain" . <http://someotherdocument/> = "<doc/>"; :format "text/xml" .

19:41:53 <DanC_> [as Michael Sperberg-McQueen pointed out to me on the phone just the other day, a lot of these discussions get garbled thru imprecise use of quanitifiers. I'm saying that there are (or may be) *some* http URIs, without fragments, that denote byte/character sequences. ]

19:42:12 <sbp> add in the rules associated with daml equivalence...

19:42:13 <DanC_> RFC822 isn't an XML document.

19:42:22 <sbp> did I ever say that it is?

19:42:35 <sbp> theoretically, they could publish it as such

19:42:47 <mnot> Marshall Rose would want to

19:43:20 <DanC_> but you'd need to change IETF process to do that, and it would get a new RFC number. or at least: its URI wouldn't end in .txt

19:43:23 <sbp> someone could also--if I'm not mistaken--serve up the plain text version of the RFC with the MIME type text/xml. I'm not sure if that's consistent, though

19:43:57 <sbp> IETF process could change overnight

19:44:04 <DanC_> someone could do lots of things, sbp. My only claim here is that there are reasonable, plausible contexts in which { <http://www.ietf.org/rfc/rfc0822.txt> = " \n\n\n\n RFC # 822 ..." }

19:44:13 <mnot> They could do conneg on it; there's not reason why text/xml can't be served on a URI ending in .txt

19:44:14 <DanC_> i.e. there's no architectural axiom that contradicts that.

19:44:38 * sbp thinks about it

19:45:39 <DanC_> the claim gets even stronger if I make it { <http://www.ietf.org/rfc/rfc0822.txt> = ('text/plain', " \n\n\n\n RFC # 822 ...") }. But that's almost a point of taste, not architecture.

19:46:08 <sbp> if you can prove that there is no axiom that contradicts that, then it means that there is still no known exception to HTTP URIs identifying anything, right? but the position is weak in two places since absence of contradictions put forth by us doesn't mean that there are no contractions, and also it's only one argument against the "HTTP URIs can identify anything" argument

19:46:28 <sbp> contractions? argh. s/contractions/contradictions/

19:46:58 <DanC_> I was refuting mnot's claim: "The 822 resource *is* more than a sequence of bytes". I say: maybe it is, maybe it isn't.

19:47:30 <sbp> O.K., I agree with that. The IETF hasn't issued any formal statement for what that URI identifies

19:47:58 <sbp> agree: that is, assuming that the range of HTTP argument pans out in favour of being able to identify anything with HTTP URIs

19:48:24 <danbri> eikeon?

19:48:27 <mnot> fair enough

19:48:35 <DanC_> as to modelling preferences, I'd prefer to modell it as a cyc:TemporalThing, which is disjoint from ByteSequence.

19:48:49 <mnot> But I'd say that this situation is the exception, not the rule

19:49:02 <DanC_> so where were we?

19:49:03 * danbri is thinking more seriously about transliterating the netlib/redfoot parser into ruby

19:49:10 <mnot> I.e., don't just go around saying <resource> = "representation"

19:49:19 <DanC_> not lightly, anyway, no.

19:49:19 <sbp> { timbl:GenericDocument rdfs:subClassOf cyc:TemporalThing } you do think?

19:49:28 <danbri> I want some mindless but fiddly task to do... ;-)

19:49:29 <DanC_> good question, sbp.

19:49:38 <sbp> s/you do/do you/

19:49:48 <DanC_> how about choosing a date for the 5th WebOnt ftf, danbri? ;-)

19:50:54 * danbri remembers his theory about how they pick the dates for Easter, then realises it's probably blasphemous

19:51:48 * DanC_ wonders how to leverage deployed RSS traffic patterns to deploy RDF calendaring stuff.

19:53:00 <mnot> do you have something in mind when you say 'deployed RSS traffic patterns'?

19:56:52 <DanC_> RSS traffic: yeah; all the existing RSS server feeds (W3C home page, slashdot, etc.), and the clients (evolution, meercat, ...), and all the intermediaries in between.

19:57:06 <mdupont> # The RDF Schema Specific DataBase (RSSDB): The First RDF Store using schema knowledge to automatically generate an Object-Relational (SQL3) representation of RDF metadata and load resource descriptions.

19:57:11 <mdupont>http://139.91.183.30:9090/RDF/index.html

19:57:12 <dc_rdfig> G: http://139.91.183.30:9090/RDF/index.html from mdupont

19:57:17 <mdupont> G:|The RDF Schema Specific DataBase (RSSDB): The First RDF Store using schema knowledge to automatically generate an Object-Relational (SQL3) representation of RDF metadata and load resource descriptions.

19:57:17 <dc_rdfig> titled item G

20:01:34 <oierw|sleep> oierw|sleep is now known as oierw

20:04:55 * DanC_ feels obliged to note all the new URI schemes I've seen, while I'm working on /Addressing/schemes

20:11:02 <DanC_> hmm... does \ do its thing inside """ ?

20:13:21 <DanC_> yup

20:13:23 <DanC_> longstringitem ::=

20:13:23 <DanC_> longstringchar | escapeseq

20:13:33 <DanC_> -- http://www.python.org/doc/current/ref/grammar.txt

20:14:54 <DanC_> hmm... problem using http://blogspace.com/rdf/iana-uri-schemes

20:15:05 * mdupont thinks daml-mode is cool

20:24:26 <Morbus> em, you round?

20:24:38 <Morbus> the SciAm link here: http://www.w3.org/2001/sw/EO/points is now pay-only.

20:25:46 <DanC_> pointer to daml-mode?

20:26:11 * DanC_ tries to figure out how to use HTTP last-modified in a Makefile... finds curl --time-cond option...

20:26:24 <Morbus> HEAD from LWP::Simple?

20:26:50 <DanC_> got a make recipie in mind, Morbus?

20:27:18 <Morbus> "make recipe". not a clue. never made a makefile.

20:27:41 <Morbus> HEAD http://www.w3c.org/ | grep -w "Last-Modified"

20:27:45 <Morbus> that works though.

20:28:35 <Morbus> odd. my server isn't spitting out Last-Modifieds

20:29:12 <Morbus> ah. SSI's everywhere.

20:29:12 <Morbus> bah.

20:29:44 <xover> Apache? Try "XBitHack Full".

20:29:56 <Morbus> that turns on Last-Modified for SSI files?

20:30:04 <xover> Yup.

20:30:50 <Morbus> and that won't conflict with my normal SSI operation - i've enabled it normally, sans XBit.

20:31:44 <xover> Morbus: See http://httpd.apache.org/docs/mod/mod_include.html#xbithack

20:31:55 <Morbus> yeah, this is ... hold. phone.

20:32:23 <DanC_> usually, jordan.

20:32:39 <DanC_> I'm trying to get Make to do something when an HTTP thingy is modified, but nothing if it wasn't.

20:32:55 <DanC_> I don't think it can be done.

20:32:58 <Morbus> xover: for me to do this, i'm gonna have to 755 a crap loaf of stuff I don't want to.

20:33:14 <xover> chmod g+x *.html, yes.

20:33:18 <Morbus> yeah, i don't wanna do that.

20:33:47 <xover> The no Last-Modified for you, I suppose.

20:33:57 <Morbus> unfortunately, yeah.

20:33:58 <mdupont> Anyone know of the url for the RPM RDF schema?

20:34:18 <sbp> .google RPM RDF Schema

20:34:18 <datum> RPM RDF Schema: http://rpmfind.net/linux/rpm2html/messages/0028.html

20:34:28 <em> Morbus, i'm here but deep in talk-mode...

20:34:30 <mdupont> thanks

20:34:35 <em> thanks for catching the link on /points

20:34:38 <em> will fix this

20:37:30 <danbri> hi eric

20:39:26 <mdupont> Hey, how can I specify an operation on a DAML class?

20:39:49 <mdupont> Lets say you want to spec a property that is really a function call.

20:40:01 <mdupont> or a method on an instance of a class

20:40:11 <mdupont> and want to define the methods signature.

20:41:20 <mdupont> ?

20:42:48 * em waves to danbri

20:50:10 <mdupont> would you say that a program that links to a library is a subclass of the users of the lib, or is the lib a property of that program?

20:50:45 <bijan> Er...

20:50:51 <bijan> depends on how you want to model it.

20:50:58 <bijan> but niether seem too great to me.

20:51:08 <mdupont> Well if you wanted to model a linker

20:51:10 <bijan> I'd prolly go with the latter if those were my only choices.

20:51:19 <mdupont> A property?

20:51:29 <bijan> WEll, why would the program be a subclass?

20:51:42 <mdupont> can I nest properties inside the class body in daml?

20:51:46 <bijan> or do you mean a member of the class users-of-the-lib.

20:51:51 <bijan> Why would you care?

20:51:57 <bijan> re: nesting

20:52:03 <bijan> That would be, at best mere syntax.

20:52:06 <mdupont> cause it looks nicer

20:52:09 <bijan> There's no "nesting" in daml :)

20:52:09 <mdupont> :)

20:52:10 <bijan> Ah.

20:52:14 <bijan> Dunno.

20:52:19 <bijan> I'd be a bit surprised.

20:52:25 <mdupont> hmmm... I am torn

20:52:36 <bijan> Fighting daml/rdf syntax is hard enoughw ithout trying to makeit look nice :)

20:52:42 <mdupont> heheh

20:52:47 <bijan> Or rather, dealing with modeling issues is ahrd enough.

20:52:51 <mdupont> sure

20:52:52 <bijan> hyperdaml makes it all work for me.

20:53:17 <mdupont> thks

20:53:21 <mdupont> i will try it out

20:57:26 <mdupont> Does it need java to run?

20:58:27 <mdupont> ok the webservice is also good.

20:58:38 <Morbus> whoo.

20:58:45 <Morbus> i got nearly 30 RDF bookmarks to read now.

21:01:11 <bijan> Oh, I have more for you.

21:01:13 <bijan> ANytime :)

21:01:28 * bijan only half way through ISWC 2002

21:01:39 <mdupont>http://introspector.dyndns.org/daml/introspector.daml

21:01:39 <bijan> And I just nicked jim's copy of www2002

21:01:39 <dc_rdfig> H: http://introspector.dyndns.org/daml/introspector.daml from mdupont

21:01:54 <mdupont> H:| Mikes poor attempt at breaking daml

21:01:54 <dc_rdfig> titled item H

21:03:23 <mdupont> H:there is nothing to see here right now. but the hyperdaml is crashing

21:03:24 <dc_rdfig> added comment H1

21:03:27 <mdupont>http://www.daml.org/cgi-bin/hyperdaml?http://introspector.dyndns.org/daml/introspector.daml

21:03:27 <dc_rdfig> I: http://www.daml.org/cgi-bin/hyperdaml?http://introspector.dyndns.org/daml/introspector.daml from mdupont

21:03:45 <mdupont> can someone try this link? hyperdaml does not like it.

21:03:57 <bijan> Er.../me doesn't want to crash things.

21:04:13 <mdupont> Exception in thread "main" java.net.ConnectException: Connection refused at java.net.PlainSocketImpl.socketConnect(Native Method) at

21:05:02 <mdupont> bijan: can you tell me if you can load my link at all, H: ? pretty pleaseß

21:05:03 <mdupont> ?

21:05:37 <bijan> Sure.

21:05:41 <bijan> one second.

21:05:43 <bijan> it's chumped right?

21:05:51 <mdupont> yes

21:06:11 <mdupont> Hixie: is an attempt at nesting strutures

21:06:16 <mdupont> H:is an attempt at nesting strutures

21:06:16 <dc_rdfig> added comment H2

21:06:18 <bijan> Busted for me too.

21:06:24 <bijan> Exceptions all over.

21:06:39 <mdupont> so you cannot access it? or the daml is bad

21:06:40 <mdupont> ?

21:06:59 <bijan> If I follow that link, the last one in the chump, I get a traceback.

21:07:12 <bijan> Connection refused, etc.

21:07:29 <bijan> If I follow the link itself.

21:07:35 <bijan> Ile., tha'ts being hyperdamled.

21:07:37 <chris_su> general question regarding foaf files - any takers?

21:07:38 <bijan> I get refused too.

21:07:49 <bijan> The link is unavailable to me and to hyperdaml.

21:07:56 <bijan> Depends on the question :)

21:08:04 <chris_su> well...here goes then

21:08:11 <bijan> I'll take it if I have a specific anser that I think is correct :)

21:08:34 <mdupont> ok, so my permissions are bad. thanks!

21:08:35 <chris_su> are there any plugin's for IRC clients that would let me view other people's foaf info?

21:08:39 <bijan> Sure.

21:08:44 <bijan> I can't get to that host, even.

21:08:52 <bijan> Not yet.

21:08:54 <mdupont> I chowned the file. I guess i have to chmod 777 it

21:09:02 <chris_su> ah...so the only app is http://swordfish.rdfweb.org/discovery/2001/08/codepict/

21:09:11 <bijan> mdupont: I can't get to the root site.

21:09:16 <bijan> No, there's foafbot.

21:09:19 <bijan> For example.

21:09:31 <bijan> over on #foaf.

21:09:35 <chris_su> cool

21:09:38 <chris_su> I'll have a look

21:09:40 <mdupont> ok, thanks bijan

21:09:42 <bijan> More coming soon.

21:09:47 <bijan> mdupont: no prob.

21:09:50 <mdupont> I will move it to sf.net

21:18:54 <jordan> xslt q: what is wrong with <xsl:variable name="foo" select="a:tag[@b:attr = $myvar]"/> where $myvar is a previously defined (in the toplevel) variable?

21:19:01 <mdupont>http://www.daml.org/cgi-bin/hyperdaml?http://introspector.sourceforge.net/daml/introspector.daml

21:19:01 <dc_rdfig> J: http://www.daml.org/cgi-bin/hyperdaml?http://introspector.sourceforge.net/daml/introspector.daml from mdupont

21:19:14 <mdupont> J:|Hyperdamled introspector daml

21:19:15 <dc_rdfig> titled item J

21:19:54 <Morbus> what's "Model" mean in RDF::Redland::Model?

21:20:02 <dajobe> graph

21:20:46 <mdupont> Ok so I have a Ontology introspector that describes a package DIA, but DIA has its own ontology that I want to reference

21:20:58 <mdupont> can I just embed one Ont. in another?

21:21:48 <Morbus> dajobe: so... ... what's that mean? is it a in-memory representation of the triples?

21:22:18 <dajobe> it is an abstract representation of the rdf graph. the rdf graph is a set of triples

21:22:37 <dajobe> the storage aspect of the graph is in the RDF::Redland::Storage class(es)

21:22:49 <Morbus> so I ::Model first and then pass to ::Storage?

21:22:55 <dajobe> vice versa

21:23:03 <Morbus> is there a "you're a dumb dumb" tutorial for Redland?

21:23:03 <dajobe> see the example code?

21:23:16 <dajobe> sorry, just example code

21:23:18 <Morbus> i see the example code, i just don't know whta it means.

21:23:30 <dajobe> no, that's not right. I'm not sorry that there are examples.

21:23:39 <dajobe> the perl POD documentation might help

21:24:01 <dajobe> do "perldoc RDF::Redland::Model"

21:24:06 <dajobe> or read it online

21:24:25 <dajobe> . http://www.redland.opensource.ac.uk/docs/pod/RDF/Redland/Model.html

21:25:53 <Morbus> if you have a moment, could I impose on you?

21:25:58 <Morbus> no rush, whenever.

21:26:12 <dajobe> i've a little time

21:26:38 <Morbus> i'd like to see some demo code to get a FOAF file into memory.

21:26:52 <Morbus> something that I could Dumper() out, or whatever, in an attempt to display the contents.

21:26:52 <dajobe> that's pretty easy

21:27:07 <dajobe> in what lang?

21:27:09 <Morbus> perl.

21:27:39 <dajobe> it won't be in perl's memory

21:27:59 <dajobe> i.e. the stuff will be perl objects referencing C objects

21:28:10 <dajobe> start from perl/example.pl

21:28:20 <dajobe> which already does parsing

21:28:41 <dajobe> delete the code 'Creating a statement' ... up to add_statement

21:29:03 <dajobe> change the my $storage line to be:

21:29:13 <dajobe> my $storage=new RDF::Redland::Storage("hashes", 'dummy', "new='yes',write='yes',hash-type='memory',dir='.'");

21:29:31 <dajobe> or just change hash-type='bdb' to hash-type='memory' in the existing line

21:29:45 <dajobe> now all the stuff will be in memory

21:30:11 <dajobe> change the URI to point to a foaf file that you've downloaded

21:30:17 <dajobe> from my $uri=new RDF::Redland::URI("file:$test_file");

21:30:23 <dajobe> to my $uri=new RDF::Redland::URI("file:$foaf_file");

21:30:36 <dajobe> and then you can play with the code after that

21:30:43 <Morbus> alright. i'll fiddle. thanks.

21:30:51 <dajobe> try printing out the values of methods on the $statement objects

21:31:00 <mdupont> RDFDump is telling me : The enclosing property already has an object

21:31:10 <mdupont> how can i find out what it is talking about?

21:31:17 <dajobe> have a look at perl/test.pl and perl/rssdump.pl for more hints

21:31:26 <dajobe> mdupont: it should give you a line number

21:31:31 <Morbus> dajobe: any issues compiling this on os x?

21:31:39 <Morbus> (redland, that is0.

21:31:39 <mdupont> dajobe: just the file

21:31:42 <dajobe> Morbus: I just checked, it builds ok on OSX just now

21:31:54 <dajobe> mdupont: darn

21:31:58 <Morbus> alright :)

21:32:34 <dajobe> mdupont: it is complaining about <foo:property><foo:property2>value</foo:property2></foo:property>

21:32:48 <dajobe> but it's odd it didn't give a line no

21:32:54 <mdupont> ok

21:32:58 <dajobe> wait, not quite that

21:33:00 <mdupont> I can try and debug

21:33:11 <dajobe> it is complaining about <foo:property rdf:resource="http://someuri/"><foo:property2>value</foo:property2></foo:property>

21:34:13 <mdupont> cool. so you want me to see if I can get it to tell me the line number?

21:34:36 <dajobe> it's puzzling that it doesn't

21:35:06 <Morbus> dajobe: any weird licensing i need to know about concerning shipping redland libraries?

21:35:16 <Morbus> ie, if I were to include a foaf parser in amphetadesk, using redland as the backend.

21:35:22 <Morbus> (i do the same with expat, for example).

21:35:36 <mdupont> i have * $Id: rdfdump.c,v 1.18 2001/10/08 22:30:11 cmdjb Exp $

21:37:10 <dajobe> Morbus: can't say, it's LGPL/GPL/MPL tri-licese, like Mozilla is

21:37:13 <dajobe> license

21:37:37 <Morbus> oh! i though you coded it all.

21:37:52 <Morbus> yeah, that should be fine.

21:37:55 <Morbus> amphetadesk is artistic.

21:38:05 <dajobe> that'll work

21:38:56 <Morbus> my difficulty is gonna be figuring out how to bind it in a win32 exe.

21:39:11 <dajobe> it builds on win32 so I hear

21:39:37 <Morbus> yeah, but for some reason or another, i can't take a new expat and hook it into the perl stuff

21:39:50 <Morbus> so, i fear the same will be said of redland.

21:39:54 <dajobe> possibly

21:39:58 <dajobe> you can always use libxml

21:40:13 <Morbus> (i compile amphetadesk into an .exe runtime for win/mac so enduser doesn't need perl installed)

21:40:20 <dajobe> sure

21:40:59 <mdupont> dajobe: is the raptor code under public read cvs somewhere?

21:41:08 <dajobe> of course

21:41:22 <dajobe> here: http://cvs.ilrt.org/cvsweb/redland/raptor/

21:41:35 <dajobe> building from cvs details: http://www.redland.opensource.ac.uk/raptor/INSTALL.html

21:42:01 <mdupont> thanks

21:42:22 <mdupont> I had a small problem building and want to diff to see if i had to change anything

21:43:31 <dajobe> I've gotta take a break for today

21:43:35 <dajobe> back tomorrow

21:44:17 <mdupont> ok, thanks for helping out. I am looking forward to using raptor/rdf.

22:20:46 <DanC_> Dave? can you help with the daml:collection test?

22:21:25 <DanC_> DanC_ is now known as dawaco

22:21:46 * dawaco tries on a dajobe-style jedi name

22:22:03 <dawaco> dawaco is now known as DanC

22:31:04 <mdupont>http://www.omg.org/docs/orbos/1997/97-02-19.txt

22:31:05 <dc_rdfig> K: http://www.omg.org/docs/orbos/1997/97-02-19.txt from mdupont

22:31:19 <mdupont> K:| An old copy of the subset of the ODL

22:31:19 <dc_rdfig> titled item K

22:31:33 <mdupont> K:THis is to show aj how cool rdfig is

22:31:33 <dc_rdfig> added comment K1

22:32:01 <mdupont> dc_rdfig: what is the homepage?

22:32:02 <dc_rdfig> Not understood: what is the homepage?

22:33:04 <DanC> <daml:subClassOf rdf:resource="GNOMEApp"/> <- missing a #

22:33:06 <mdupont> dc_rdfig: rdfig chat page?

22:33:06 <dc_rdfig> Not understood: rdfig chat page?

22:33:56 <mdupont> thanks danc ++!

22:34:29 <DanC> F:ooh..

22:34:33 <dc_rdfig> added comment F2

22:35:55 <mdupont> F:

22:35:55 <dc_rdfig>http://www.iki.fi/~decoy/texts/rdf/unproperties.en.html

22:35:55 <dc_rdfig> Naming them DAML unproperties

22:35:56 <dc_rdfig> (1:ssyreeni) A few thoughts about yesterdays daml:Unambiguous et al. discussion

22:35:57 <dc_rdfig> (2:DanC) ooh..

22:41:40 <DanC> F:this was discussed, breifly, in today's telcon. The issue owner's [proposal|http://lists.w3.org/Archives/Public/www-webont-wg/2002Jul/0177.html] still calls for OWL:FunctionalProperty OWL:InverseFunctionalProperty.

22:41:41 <dc_rdfig> added comment F3

22:42:16 <mdupont> Hey Dan, you gotta check this out :

22:42:18 <mdupont>http://www.javaskyline.com/database.html

22:42:18 <dc_rdfig> L: http://www.javaskyline.com/database.html from mdupont

22:42:21 <DanC> F:I'm not going to object. I can see the point that saying that something is both a ManyToOneProperty and a one to one property is a little goofy.

22:42:21 <dc_rdfig> added comment F4

22:42:46 <mdupont> L:Object relational mapping, JDO, and Database engineering Broadly speaking Object Relational Mapping tools map objects onto a relational database. The goal of these tools is to reduce development time, database applications. O/R Mapping and database engineering tools generally obtain metadata in some form - by reverse-engineering from the database, from XML, or from Java classes. Barry and Associates have a set of presentations on Objec

22:42:46 <mdupont> elational mapping tools and what they are used for. JDO (Java Data Objects) is an emerging standard within Object Relational. JDO provides the developer with a common design-time interface to O/R capabilities that are implemented at run-time. Eric Sampson provides an introduction to JDO and history of JDO. And also see, JDO Central.

22:42:46 <mdupont> They then use the metadata to generate code aDDL (data definition language) and possibly data transforms. They may also provide an accessible repository for the metadata. The table below lists the general features, APIs, and interfaces for each tool.

22:42:46 <dc_rdfig> added comment L1

22:43:40 <DanC> most RDB/OO systems are uninterestingly closed word.

22:43:44 <DanC> not to mention fragile.

22:43:48 <mdupont> www.ccs.neu.edu/home/futrelle/bionlp/ismb2000/mcentire.pdf

22:43:48 <DanC> closed world.

22:43:55 <mdupont>http://www.ccs.neu.edu/home/futrelle/bionlp/ismb2000/mcentire.pdf

22:43:55 <dc_rdfig> M: http://www.ccs.neu.edu/home/futrelle/bionlp/ismb2000/mcentire.pdf from mdupont

22:44:24 <mdupont> M:|Ontology for Bioinformatics

22:44:24 <dc_rdfig> titled item M

22:44:28 <MorbusIff> MorbusIff is now known as Morbus

22:53:34 <mdupont> so how can i rdfdump a file in the current directory

22:53:41 <mdupont> file:/./x.rdf?

22:59:17 <ssyreeni> F:that's just the point. under the E-R terminology, one-to-one relations are a subclass of many-to-one, one-to-many and many-to-many ones. whatever one wants to say is done by rdfs:subPropertyOf.

22:59:17 <dc_rdfig> added comment F5

23:01:50 <ssyreeni> F:so, the unproperties correspond precisely to the established meaning of one-to-many and many-to-one.

23:01:51 <dc_rdfig> added comment F6

23:09:57 <mdupont> I:

23:09:57 <dc_rdfig>http://www.daml.org/cgi-bin/hyperdaml?http://introspector.dyndns.org/daml/introspector.daml

23:10:17 <mdupont> I:this is broken, how can i delete it?

23:10:17 <dc_rdfig> added comment I1

23:10:21 <mdupont> J:

23:10:21 <dc_rdfig>http://www.daml.org/cgi-bin/hyperdaml?http://introspector.sourceforge.net/daml/introspector.daml

23:10:22 <dc_rdfig> Hyperdamled introspector daml

23:11:37 <mdupont> J:Has been cleaned up

23:11:37 <dc_rdfig> added comment J1

23:11:39 <mdupont> K:

23:11:39 <dc_rdfig>http://www.omg.org/docs/orbos/1997/97-02-19.txt

23:11:40 <dc_rdfig> An old copy of the subset of the ODL

23:11:41 <dc_rdfig> (1:mdupont) THis is to show aj how cool rdfig is

23:11:55 <mdupont>http://introspector.sourceforge.net/daml/introspector.daml

23:11:56 <dc_rdfig> N: http://introspector.sourceforge.net/daml/introspector.daml from mdupont

23:12:11 <mdupont> N:|Introspectors experimental DAML

23:12:11 <dc_rdfig> titled item N

23:12:31 <mdupont> N:|has been cleaned up and tested against rdfdump

23:12:32 <dc_rdfig> titled item N

23:13:00 <mdupont> N:|thanks to DanC and Bijan and Danjobe

23:13:01 <dc_rdfig> titled item N

23:16:10 <mdupont> is it fair to put multiple ontologies in one file?

23:16:24 <mdupont> how can i associate an ontology to the classes

23:16:25 <mdupont> ?

23:16:39 <mdupont> and how can i comment a subclassof property?