Semantic Web Interest Group IRC Chat Logs for 2002-08-30

01:16:09 <oierw> oierw is now known as oierw|away

01:28:09 <jhendler>http://www.daml.org/2002/08/dql/

01:28:09 <dc_rdfig> A: http://www.daml.org/2002/08/dql/ from jhendler

01:28:25 <jhendler> A:| Daml Query Language

01:28:26 <dc_rdfig> Titled item A.

01:28:47 <jhendler> A: Released by Joint Committee - proposed query langauge for DAML+OIL

01:28:47 <dc_rdfig> Added comment A1.

02:53:04 <oierw|away> oierw|away is now known as oierw

07:23:08 <dmiles> dmiles is now known as dmilezZzZ

08:05:39 <lilo> [Global Notice] Hi all. Fairly major rehub in process. Please bear with us.

08:10:45 <ChanServ> [#rdfig] This channel is logged and blogged: http://logicerror.com/rdfIRCWelcome

08:25:09 <lilo> [Global Notice] Hi all. Rehubbing operations should be complete. Thank you for your patience, and thank you for using freenode!

08:43:46 <blanu> Are there any RSS feeds which actually use application/rss+xml?

08:45:16 <JibberJim> Is it a registered mime-type?

08:45:36 <blanu> Yes, it's the official registered mime type.

08:45:53 <blanu> But I cannot find a feed which uses it, only text/xml and text/plain.

09:03:03 Topic now RDF/Semantic Web 24x7 chat - http://rdfig.xmlhack.com/ blog

09:03:03 Users on #rdfig: logger_1 blanu libby xmltok dajobe LotR bipolaro las1 tav deltab sbp grove dmilezZ JibberJim timmmit em shellac RM_39 eikeon oierw Skyline_ burtonator MarkB DanCon deus_x sandro datum logger pixel xover dc_rdfig

09:03:03 <ChanServ> [#rdfig] This channel is logged and blogged: http://logicerror.com/rdfIRCWelcome

10:54:22 <shellac>http://www.macdevcenter.com/pub/a/mac/2002/08/27/cocoa.html

10:54:22 <dc_rdfig> B: http://www.macdevcenter.com/pub/a/mac/2002/08/27/cocoa.html from shellac

10:55:03 <shellac> B:|Mac OS X 10.2 (phew, getting lengthy) addressbook framework

10:55:03 <dc_rdfig> Titled item B.

10:57:02 <shellac> B:A nice introduction to the OS X addressbook. Notice that properties can be added.

10:57:02 <dc_rdfig> Added comment B1.

10:59:08 <shellac> B:The addressbook does vCard already, but this method is more extensible.

10:59:09 <dc_rdfig> Added comment B2.

11:01:31 <shellac> B:Next, maybe a framework for iCal?

11:01:31 <dc_rdfig> Added comment B3.

12:40:26 <amk> Hi, all! I've updated the schema for historical figures

12:40:36 <amk> at http://www.historical-id.info/ns/person/1.0, and am now wondering whether

12:41:03 <amk> to delete the Person class, and use the FOAF Person class.

12:41:16 <amk> Should entities not be unnecessarily multiplied?

12:41:44 <JibberJim> If it's the same as foaf:Person, I'd use foaf:Person, foaf is widely deployed and understood by tools.

12:42:22 <amk> Pretty much the same, though historical figures are more often dead.

12:42:34 <Talliesin> subclass?

12:42:50 <amk> Should I use the DAML equivalentTo property to define an alias of FOAF:Person, or not bother?

12:44:07 <JibberJim> foaf:Person is already a subClassOf wordnet:Person

12:44:51 <JibberJim> If it is equivalent, I think you should just use foaf:Person, then all the foafbots can find out about your historical people.

12:45:10 <JibberJim> if it's a subset of all foaf:Person's then a subClass seems sensible.

12:47:04 <amk> It's not really a subset, just a bunch of additional properties.

12:47:09 <amk> FOAF:Person it is, then!

12:47:09 <dc_rdfig> Label FOAF not found.

12:47:29 <JibberJim> eek, schema's in N3

12:48:16 <amk> If you request content-type text/xml, you'll get the RDF version.

12:48:32 <amk> (/1.0.rdf will work, too)

12:50:06 <JibberJim> ah, that's okay then, as long as if I request application/rdf+xml too...

12:50:50 <amk> Hmm... is that the right content type for RDF, not text/xml?

12:51:34 <JibberJim> I'm not sure if it's registered yet - but I'm never going to send text/xml as an accept - that's a ludicrous idea.

12:56:57 <amk> OK, now I'm returning application/rdf+xml.

12:58:57 <Morbus> hey dajobe.

12:59:08 <dajobe> hi

13:01:35 <amk> OK, I've now changed the schema to just use FOAF:Person.

13:04:02 <amk> What are people using to check that data is consistent with a given RDF schema?

13:04:54 <Code_Poet> Code_Poet is now known as Talliesin

13:06:22 * jhendler catches up w/the log

13:06:57 <jhendler> amk - historical stuff looks interesting (we're doing a project that needs some of this - sent pointer to your page)

13:07:38 <jhendler> we're working on a project to support historians working on a big document archive - not so much the "lives of the famous" but more the everyday folks from the mid 1850s US

13:09:47 <amk> Interesting. (Though I don't quite see the connection.)

13:10:55 <jhendler> framework - we need to anchor our ontologies with concepts of people and the like, and are thinking of the unique URI issue - some interesting complications

13:11:25 <jhendler> (names aren't consistent, can only be conjectured sometimes that people are same) - but we want to be able to say things like

13:11:44 <jhendler> "the person in this picture" is the same as "the person referred to in this marriage license"

13:12:33 <amk> Oh, I see.

13:12:36 <jhendler> (your RDF files link is to a forbidden page - is that on purpose?)

13:14:49 <amk> No; fixed now.

13:16:53 <jhendler> thanks, gotta go - will check out later

13:39:25 * DanC can't believe the freenode name seems to be sticking; I would have thought they'd change back by now

15:50:06 <danja>http://www.scripting.com/rss.xml

15:50:06 <dc_rdfig> C: http://www.scripting.com/rss.xml from danja

15:50:29 <danja> C:| An example of an RSS 0.94 feed

15:50:29 <dc_rdfig> Titled item C.

15:51:05 <danja> C: anyone got a spare minute to do a -> RSS 1.0 stylesheet?

15:51:05 <dc_rdfig> Added comment C1.

15:51:28 <danja> C: DC qualified should pretty much cover it

15:51:28 <dc_rdfig> Added comment C2.

15:51:39 <chrisc> is a guid element the same as dc:itentifier?

15:52:08 <danja> I *think* it's a resource ;-)

15:52:15 <danja> rdf:about

15:52:24 <chrisc> oh it's a replacement for link?

15:52:42 <chrisc> should we take this to #rss-dev?

15:53:11 <Talliesin> it would appear to be somewhat equiv to rdf:about

15:53:29 <danja> judging by http://backend.userland.com/ideasForEvolutionOfRss

15:53:31 <Talliesin> but I may be foolish in trying to think about this rationally.

15:54:39 <danja> right on, Talliesin ;-)

15:55:09 <Talliesin> okay so we have 1. duplication of dcterms:available

15:55:20 <Talliesin> 2. duplication of rdf:about

15:56:12 <Talliesin> 3. duplication of HTTP expires

15:56:38 <Morbus> 3. I don't mind.

15:56:52 <Morbus> that requires some knowledge of webserver tech, and http to understand the point.

15:56:57 <Morbus> not many people have that ability to control.

15:57:48 <Talliesin> dcterms:available can have a range, so it can do 3 as well.

15:57:49 <JibberJim> yes, but re-implementing something that is already well established, and working, should be done very much as a last resort. Too many specs especially new XML langs re-invent stuff already done at the HTTP level.

15:58:16 <Morbus> JibberJim: I don't think of it as reinventing.

15:58:22 <Morbus> i see it as making it accessible to normal people.

15:58:26 <Morbus> and that, I can't argue with.

15:58:53 <Talliesin> yes, but someone is GETting it, so that someone should be the one making that info available.

15:58:59 <JibberJim> normal people have access to the server settings - most blog software relies on server side scripts which means they can write any headers they want.

15:59:07 <Morbus> that's insane.

15:59:16 <Morbus> now you want people to learn PHP? or Perl?

15:59:23 <Morbus> or have access to use headers via Apache?

15:59:36 <Morbus> normal webhosts don't grant access to mod the headers, much less access to an .htaccess file.

15:59:51 <JibberJim> of course you have control in appache via .htaccess.

15:59:54 <Morbus> tech shouldn't get in the way of writing.

16:00:02 <Morbus> JibberJim: no, you miss my point.

16:00:06 <Morbus> JibberJim: yeah, you can do it.

16:00:10 <Morbus> but you have to have access to do it.

16:00:16 <Morbus> most webhosts don't grant .htaccess control

16:00:23 <JibberJim> exactly so don't re-invent half arsed methods which tools have to learn new, to solve problems that are already solved.

16:00:23 <Morbus> much less that specific module feature to rewrite headers.

16:00:37 <Morbus> much less create docs that make it easy to understand 1) why you'd want to do this 2) how'd you do this

16:00:54 <Morbus> i should step out of this conversaton.

16:00:55 <Morbus> i'll start getting upset.

16:01:00 <JibberJim> SVG did the same invented some half arsed gzip requirement when that's been around forever...

16:01:02 <Morbus> it's the same old "I know it. It's easy to me. It should be easy to everyone else".

16:01:09 <Morbus> that's why geeks suck.

16:02:14 <JibberJim> No, I'm not, I'm saying the problem has been solved in an efficient manner understood by almost every http library around today, we shouldn't try and solve it again in a new way, we should ensure people can use the existing efficient mechanism.

16:02:40 <Talliesin> JibberJim - I don't mind the SVG gzip thing though, since it's only duplication in terms of transmission over HTTP, not for storage or other transmission.

16:03:20 <Morbus> JibberJim: people *aren't ensuring* that others know how to use it. no matter how many docs you can write, you still have the downside of "this is server side technology that can either be stopped by the server administrator, or requires knowledge of a programming language".

16:03:25 <JibberJim> But it means that Mozilla's SVG doesn't understand svgz files, even though it does understand gzip'd html.

16:03:30 <Talliesin> How about, "I know it. It's easy to me. I'll just go off and write something to do it for everyone else."

16:04:00 <JibberJim> How many RSS feeds come from hosts running static documents only?

16:04:21 <Morbus> phone.

16:05:37 <Morbus> long phone call.

16:10:13 <Morbus> JibberJim: what does that matter? Radio and MT both generate static RSS files.

16:11:14 <JibberJim> What's your point though - they can easily generate ones with controlled headers, we're looking at extensions (so the apps will need changes anyway.)

16:11:32 <Morbus> JibberJim: *how* can they easily generate ones with controlled headers?

16:11:35 <Morbus> give me an example using Apache.

16:11:53 <JibberJim> I'm not an apache user.

16:12:09 <Morbus> ok. well, gimme an example of how they would do it.

16:12:14 <chrisc> I am... what's the question again?

16:12:43 <Morbus> the point I'm getting at JibberJim: in Apache, the ability to add headers can be restricted. The ability to add headers isn't even turned on by default, I believe.

16:12:45 <JibberJim> My point is anything we can use at the http level is already supported by my RSS aggregrators. So whilst you could duplicate I'd urge you not to.

16:12:59 <Morbus> so, there's a server side restriction right there, regardless of static or dynamic.

16:13:04 <Morbus> if it's static, then they'd have to change CMSs.

16:13:21 <Talliesin> still nothing to stop you using dcterms:available.

16:13:36 <Morbus> if they move to dynamic, then the CMS has to already have a userinterface to change the header, or they'd have to learn how to code in the same lang as the CMS, and then learn where to put it in the CMS.

16:13:39 <JibberJim> an anti RDF inclination surely?

16:13:46 <Morbus> yes, I advocate dcterms: over whatever Winer wants.

16:13:57 <Morbus> because dcterms is controllable in the user's space.

16:14:01 <Morbus> http headers are not.

16:14:16 <Morbus> I'm not advocating Winer's choice - I'm advocating something that the user can control.

16:14:36 <Morbus> I'm not anti-RDF anymore, btw.

16:14:38 <Morbus> ;)

16:14:49 <JibberJim> Winer seems to be though...

16:15:11 * Talliesin implants the RDF tracking device into Morbus so he can never leave us again.

16:15:11 <Morbus> heh, heh. thanks :)

16:15:18 <Morbus> JibberJim: sure, Winer is. has been since day one.

16:15:38 <Talliesin> Winer is anti-XML, getting him to be pro-RDF will be some doing.

16:15:39 <verbosus> hola

16:15:45 <Morbus> my beef was more with "well, mod the http headers, screw Winer's thingy", not "well, use dcterms instead of Winer's thingy"

16:16:28 <JibberJim> I'm just saying only re-invent where there's good reason, so if you say dc:terms say that it should be reflected in the http headers where possible.

16:16:37 <Talliesin> header only applicable to the channel as a whole anyway

16:17:09 <Talliesin> JibberJim - of course dcterms can't be used with 0.9[1-9] because it isn't RDF

16:17:53 <chrisc> There is reference to HTTP headers in mod_dcterms

16:18:02 <JibberJim> Good!

16:18:04 <chrisc> It is recomended that the Modified date should correspond to the HTTP 1.1 Last-Modified date, of the document the <link> element points to when Modified is used withing a <item> element.

16:18:11 <chrisc>http://groups.yahoo.com/group/rss-dev/files/Modules/Proposed/mod_dcterms.html#modified

16:18:11 <dc_rdfig> D: http://groups.yahoo.com/group/rss-dev/files/Modules/Proposed/mod_dcterms.html#modified from chrisc

16:18:44 <Morbus> irrelevant, of course, if the rss file is dynamic.

16:19:02 <chrisc> When Modified is used within the <channel> element it should correspond to the HTTP 1.1 Last-Modified date of the RSS feed and it is recomended that this should correspond to the date when the content of the feed changed, not the date when it was generated.

16:19:21 <chrisc> does that make sense?

16:20:38 <Morbus> yup.

16:20:47 <chrisc> :-)

16:33:03 <Talliesin> RFC822!

16:35:11 <JibberJim> RFC3023!

16:35:28 * JibberJim wonders why we're announcing the RFC we're currently reading...

16:35:51 <jang> are the rfc's up to 822! yet - I thought they were somewhere in to 3000's

16:36:12 <jang> which isn't even 7!

16:37:15 <Talliesin> I'm expressing surprise at winer's use of rfc822 date times.

16:37:36 <Talliesin> although it's not quite that bad, the examples actually use rfc2822, so at least they are y2k compliant.

16:37:54 <JibberJim> On a mime-type does the +xml suffix mean I can send the content to an xml parser and expect it to understand it?

16:38:11 <Talliesin> or rather the examples are, but the spec isn't.

16:38:57 <Talliesin> yep if it ends in +xml then it can be interpreted as "raw" xml should you wish.

16:39:12 <JibberJim> Unless it's image/svg+xml unfortunately...

16:39:21 <Talliesin> ?

16:39:30 <Talliesin> oh wait.

16:39:39 <Talliesin> because of the possibility of gzip encoding.

16:39:55 <JibberJim> yep.

16:40:41 <JibberJim> I'll raise it as issue, might force the working group to actually do the registration, that they've been threatening for a very long time.

16:42:41 <Talliesin> is there a MIME parameter for gzipped?

16:43:46 <JibberJim> there are no parameters for image/svg+xml - indications from the WG are they'll be none - including charset, which I think's wrong too, but that's an argument for when they publish a draft.

18:04:55 <danja> C: 0.94 spec now up - http://backend.userland.com/rss

18:04:55 <dc_rdfig> Added comment C3.

18:06:02 <danja> C: "We also anticipate the adoption of namespace support at some point not too far down the road..."

18:06:02 <dc_rdfig> Added comment C4.

18:06:20 <JibberJim> "It's an acronym forr Really Simple Syndication" ??

18:06:41 <Morbus> ;)

18:06:45 <Morbus> ok. that was weird.

18:06:51 <Morbus> i was talking verbally aloud with someone in the office.

18:06:57 <Morbus> she said something funny, and I had smiled in the chat.

18:07:01 <Morbus> something is wrong with this picture.

18:07:39 <danja> ;-)

18:13:05 <danja> any suggestions for this :

18:13:31 <danja> I've got a node in a graph,

18:14:42 <danja> which I'm saying is an rgml:Node

18:16:13 <danja> oops! forget it for now ;-)

18:17:10 <danja> nah, that was it - I want to associate this node one-to-one with another resource

18:17:27 <danja> of a different type... does that make sense (thinking out loud)

18:17:31 <danja> ?

18:19:58 <danja> hmm - think I'll have to bung in a bnode or two...

18:21:45 <DanC>http://www.w3.org/TR/2002/WD-webarch-20020830/

18:21:45 <dc_rdfig> E: http://www.w3.org/TR/2002/WD-webarch-20020830/ from DanC

18:21:54 <DanC> E:|Architectural Principles of the World Wide Web

18:21:54 <dc_rdfig> Titled item E.

18:22:08 <DanC> E:W3C Working Draft 30 August 2002

18:22:08 <dc_rdfig> Added comment E1.

18:22:19 <DanC> E:Ian Jacobs, editor

18:22:19 <dc_rdfig> Added comment E2.

18:22:28 <DanC> E:it's a start, anyway.

18:22:28 <dc_rdfig> Added comment E3.

18:23:21 <danja> oo - nice

18:25:51 <DanC> E:from the SOTD: "This draft represents substantial input from TAG participants, but does not yet represent consensus. It is also incomplete; sections 1 and 2 are the most developed, 3 and 4 the least."

18:25:51 <dc_rdfig> Added comment E4.

19:40:32 <bijan>http://icfpcontest.cse.ogi.edu/task.html

19:40:32 <dc_rdfig> F: http://icfpcontest.cse.ogi.edu/task.html from bijan

19:40:47 <bijan> F:|ICFP 2002 programming task contest

19:40:48 <dc_rdfig> Titled item F.

19:41:43 <bijan> F: A robot control/"war" game. Are any N3 programmers going to try it? Standard, very simple AIy stuff.

19:41:43 <dc_rdfig> Added comment F1.

19:41:52 <bijan> F: (Gasp! not *AI*!!!!)

19:41:52 <dc_rdfig> Added comment F2.

19:47:17 <Seth> what are the legitimate characters that can be in rdf:Id ?

20:08:13 <Morbus> hey rael.

20:15:36 <dajobe>http://www.w3.org/TR/2002/WD-rdf-concepts-20020829/

20:15:36 <dc_rdfig> G: http://www.w3.org/TR/2002/WD-rdf-concepts-20020829/ from dajobe

20:15:50 <dajobe> G:|RDF Concepts and Abstract Data Model Working Draft, 29 August 2002

20:15:50 <dc_rdfig> Titled item G.

20:26:45 <dajobe> G:edited by Graham Klyne, Jeremy Carroll

20:26:46 <dc_rdfig> Added comment G1.

20:31:12 <dmilezZ> hi bijan

20:31:18 <dmilezZ> dmilezZ is now known as dmiles

20:31:25 <dajobe> G:Amongst many other things, defines nodes, arcs, RDF triples (node, arc, node), an RDF graph, graph equality, literals, XML literals and literal equality

20:31:25 <dc_rdfig> Added comment G2.

20:31:53 <bijan> Hey dmilesZ

20:32:18 * dmiles gets the urge to look at the new logic specified for rdf/daml

20:33:02 <dmiles> bijan: have you looked at the what sandro had been working on?

20:33:13 <bijan> No, not yet

20:33:38 <dmiles> sandro: what did you come up with for logical operators for resolutionish inference?

20:34:46 <dmiles> bijan: ah i am going to dedicate next to working on writing extra Cyc infernce for completness now they have exposed the SubL language to programmers

20:35:34 <dmiles> bijan: for example i can load hack a prolog infernce engine into it

20:36:07 <dmiles> that will live and work with the Cyc blackboard of information

20:36:41 * dmiles just is excited about the possiblities now opening up

20:39:31 <sandro> My language is pretty much just FOL. Current version schema is http://www.w3.org/2002/08/LX/RDF/v2.n3 although I'm playing with a few ways to make it more readable.

20:39:58 * dmiles looks

20:40:15 <sandro> The big tricky part is in defining the semantics of the truth predicate to avoid paradoxes, but most people never have to pay attention to that.

20:40:47 <dmiles> ah conjLeft,ConjRight ;P very 3ish

20:40:57 <sandro> I checked in a version of cwm last night with "--flatten" to produde this stuff; I'm working on cleaning it up a bit more.

20:41:30 <sandro> I'm also thikning about renaming conjLeft and conjRight to "both" and "and". *shrug*

20:41:38 <sandro> 3ish?

20:41:40 <dmiles> it looks good

20:42:21 <dmiles> oh i see

20:42:55 <dmiles> i thought for a conjunction you had to recursivly define fomula using those as glue

20:43:08 <dmiles> so that it would flatten

20:43:30 <dmiles> as in into predicate + 2 enttiies

20:43:39 <dmiles> but i see now its better then i thought

20:43:51 <dmiles> and vs both! yay!

20:44:31 <dmiles> so we can specify a diffence between a logical and and a huristic both?

20:44:37 * dmiles requotes

20:44:47 <dmiles> so we can specify a diffence between a logical 'and' and a heuristic 'both'?

20:45:42 <sandro> No, you misunderstand "both" -- it's just a name for the left operand to "and". It's silly.

20:46:01 * DanC looks around for Hendler

20:47:06 <dmiles> i think i see now, it is types of syntactic glue

20:49:31 <dmiles> # same for Conditional/if/then

20:49:42 <dmiles> one question i wasnt able to glean for myself

20:49:44 <sbp> haven't seen Jim on #rdfig for a while, but he's certainly emailing :-)

20:49:55 <dmiles> you have both kif => and <=>

20:50:34 <sandro> yes, <=> is called Biconditional

20:50:41 <dmiles> do you have a => that disallows Modus Tolens ?

20:51:03 * sandro has to go buy a horse. ciao.

20:51:35 <dmiles> ( more if a if/then then that is only as strong as prolog's ':-')

20:52:10 <dmiles> ttyl

21:16:08 <verbosus> hi

21:22:12 <DanC> G:oh my goodness... "an RDF URI reference" is defined in section 3.1. Hmm... I guess we couldn't just use IRI.

21:22:12 <dc_rdfig> Added comment G3.

21:24:11 <DanC> G:phpht. nodes are distinguished from node labels. Why?

21:24:11 <dc_rdfig> Added comment G4.

21:25:04 <DanC> G:stuff like "The set of nodes of an RDF graph is tidy" (from 3.4) would just go away if we just merged the concepts of nodes and labels

21:25:04 <dc_rdfig> Added comment G5.

21:26:31 <DanC> G:section 3.6 Graph Equality should be defined in the more familar way, where the isomorphism I just covers the bnodes.

21:26:31 <dc_rdfig> Added comment G6.

21:27:32 <DanC> G:"RDF uses well established ideas..." I'd add codd's paper on the relational calculus to that list of citations.

21:27:32 <dc_rdfig> Added comment G7.

21:27:58 <DanC> G:ah... we cite Gray, at least.

21:27:58 <dc_rdfig> Added comment G8.

21:36:23 <dajobe> C:userland-rss (TM) returns to the original RSS idea - RDF and namespaces for modularity

21:36:23 <dc_rdfig> Added comment C5.

21:37:46 <Morbus> C:Sam Ruby [http://radio.weblogs.com/0101679/2002/08/27.html|should be happy]

21:37:46 <dc_rdfig> Added comment C6.

21:38:04 <dajobe> C5:Userland-rss returns to the original RSS idea - RDF and namespaces for modularity

21:38:04 <dc_rdfig> Replaced comment C5.

21:38:11 <dajobe> wouldn't want to cite him wrongly, lol

21:38:28 <Morbus> heh, heh.

21:39:20 <DanC> hmm... anybody know where to find a list of the pages with the highest google PageRank?

21:39:33 <dajobe> if anyone does, aaron would

21:39:54 <Morbus> i'll ask.

21:42:07 <dajobe> note the userland-rss scary copyright for extra laughs

21:43:31 <Morbus> DanC: 01,9910: 0112http://www.google.com/search?q=the0101,99

21:43:40 <Morbus> searching for a stopword.

21:44:18 <dajobe> they don't have them anymore IIRC

21:44:50 <Morbus> yeah, so searching for "the" would be a decentish indication.

21:44:57 <DanC> ha!

21:45:18 <DanC> top page in "the" search results is Macromedia flash download page; the google summary is "Your web browser does not support frames."

21:45:33 <dajobe> least useful link ever?

21:46:06 <dajobe> I thought so http://google.blogspace.com/archives/000221 "Google Indexes Stopwords"

21:46:09 * DanC has to admit there are some cool shockwave games

21:46:17 <dajobe> from Aaron's Google Weblog http://google.blogspace.com/

21:49:50 <Talliesin> dajobe - not really RDF and maybe namespaces in the future would be more accurate.

21:50:43 <dajobe> it's his format, nobody else can tell what he'll do with it. As the copyright says.

21:51:05 <DanC> "Last night the DAML crawler found over 4,200,000 million DAML statements on close to 20,000 web pages - not bad for a couple of years!!"

21:51:35 <Talliesin> dajobe - I know, I just don't know why he bothers to tell anyone else about it.

21:51:42 <Talliesin> DanC - kewl!

21:52:26 <dajobe> when I last looked (aug), most of the crawled pages came from 2-3 sites

21:53:13 <dajobe> ok, 6 sites

21:53:16 <dajobe> see http://www.w3.org/2001/sw/Europe/reports/rdf_scalable_storage_report/#table1

21:53:26 <deltab> did those pages have over 210 million statements each back then?

21:53:35 <Talliesin> Actually DanC you're just the man to talk to.

21:54:00 <Talliesin> I'm looking for Glenn Adam's response to your charset considered harmful

21:54:26 <Talliesin> Any idea where I might find it (the link of the w3c i18n area doesn't work).

21:54:43 <DanC> hm... Tallesin, I have a pretty handy index of my mail back to about '98, but I think that message from Glenn predates that by quite a bit.

21:55:03 * DanC hunts for older mail archives.

21:56:08 <dajobe> google finds something from html-wg archives

21:56:35 <deltab> at http://ksi.cpsc.ucalgary.ca/archives/HTML-WG/html-wg-95q2.messages/thread.html

21:56:35 <DanC> do tell, dajobe

21:56:39 <dajobe> above

21:58:26 <dajobe> I assume this is the original DanC message http://ksi.cpsc.ucalgary.ca/archives/HTML-WG/html-wg-95q2.messages/0188.html

22:00:56 <AaronSw> http://www.webmasterworld.com/forum3/2222.htm

22:03:29 <sbp> wow, Bobby made the list?

22:04:17 * DanC wonders if would be good or bad for all HTML pages to carry a little "made with HTML!" thingy at the bottom... supposes the "valid HTML!" button is kinda close.

22:05:04 * Talliesin has seen such a button as DanC theorises.

22:05:15 <sbp> perhaps browsers should automatically put "made with HTML!" in the status bar

22:05:18 * niq thinks "made with HTML" is asking for "HTML is naff" FUD

22:05:24 * DanC wonders how long it will be before W3C stops recommending that people put <!DOCTYPE at the top of their HTML documents. surely <html xmns="http://www.w3.org/1999/xhtml"> is enough.

22:05:45 <sbp> difficult to validate it, then

22:05:49 <DanC> why?

22:05:59 <DanC> xsv will validate it just fine.

22:06:00 <sbp> what are you going to validate it against? 1.1? basic?

22:06:27 <DanC> against whatever W3C says, at http://www.w3.org/1999/xhtml. If you want a less popular dialect, you can say so, using xsi:schemaLocation

22:06:56 <Talliesin> yep doctypes have their place for profiles, but I would like to see xsv considered enough to "make the grade" so to speak.

22:07:22 <sbp> it seems likely that XSD validation will be an option for XHTML 2.0

22:07:47 <DanC> it's an option for XHTML 1.0. er... maybe you mean an endorsed option.

22:08:35 <sbp> recommended option

22:09:54 <sbp> interesting quote: "Let's keep XHTML2 simple, and hopefully less stupid." - Masayasu Ishikawa, http://lists.w3.org/Archives/Public/www-html/2002Aug/0222

22:10:11 <sbp> there have been loads of interesting responses on www-html to the XHTML 2 draft

22:10:24 <DanC> hmm... why does the whitehouse have such a high pagerank? does anybody actually use that web site?

22:10:24 <Talliesin> what's the response to xforms been like?

22:10:40 * DanC doesn't know much about xforms

22:11:05 <DanC> 36,000 links to the whitehouse home page...

22:11:51 <sbp> there's not been a great deal about XForms on www-html lately...

22:11:52 <Talliesin> I keep changing my mind about whether I think it should be brought in so early. It is technically a better solution, but I can see resistance to xforms causing resistance to xhtml2.0

22:12:19 <Talliesin> I suppose maybe html forms will be in the legacy mod

22:12:42 <sbp> someone mentioned that it's odd that support for XForms is required in XHTML 2, and XLink isn't even in there

22:12:44 <DanC> hoot! a web counter is one of the highest pageranked sites.

22:13:19 <Talliesin> lol

22:13:32 <Talliesin> sbp - yes that's true

22:13:51 <DanC> anybody got any thoughts on what will cause deployment of xforms? of XHTML 2?

22:14:22 <DanC> Health On the Net Foundation http://www.hon.ch/ <- I had never heard of that before. anybody else familiar with it?

22:14:33 <Talliesin> in fact that may well have been me forwarded on from wai-ig

22:14:45 <DanC> 47,500 links to hon

22:15:47 <sbp> Hmm... only if you're Chris Mannall

22:16:23 <DanC> (for reference: 53,200 links to w3.org)

22:16:37 <sbp> that reminds me...

22:16:40 <sbp> .googlecount "Semantic Web"

22:16:41 <datum> "Semantic Web": 87,800

22:17:06 <Talliesin> .googlecount the

22:17:07 <datum> the: 2,740,000,000

22:17:24 <Talliesin> oh, IRC-based ego-surfing:

22:17:25 <Talliesin> .googlecount "Jon Hanna"

22:17:26 <datum> "Jon Hanna": 1,810

22:17:39 <Seth> .googlecount web semantic

22:17:40 <datum> web semantic: 354,000

22:17:45 <sbp> it's good for settling arguments

22:17:49 <sbp> but... um, open to abuse

22:17:57 <Seth> .googlecount semantic web

22:17:58 <datum> semantic web: 354,000

22:18:04 <niq> .googlecount "cementic web"

22:18:04 <datum> "cementic web": 0

22:18:10 <niq> shame...

22:18:16 <Talliesin> lol

22:18:18 <Talliesin> .googlecount "Jon Hanna" -psychedelic -"christian rock" -car

22:18:18 <datum> "Jon Hanna" -psychedelic -"christian rock" -car: 1,630

22:18:20 <DanC> wow... www-zig is remarkably active. http://lists.w3.org/Archives/Public/www-zig/2002Jul/thread.html

22:18:23 <sbp> well, give it a while

22:18:33 <Seth> googlecount "semweb"

22:18:44 <Seth> .googlecount semweb

22:18:45 <datum> semweb: 3,650

22:19:00 <Seth> .googlecount "sem web"

22:19:01 <datum> "sem web": 788

22:19:26 <Talliesin> .googlecount googlecount

22:19:26 <datum> googlecount: 56

22:20:29 <Seth> here is a uri for the context of the semantic web http://robustai.net/ContextPhrase#semantic_web

22:20:43 <DanC> ha! switching from "the" to "le", the #1 hit is a microsoft page counter!

22:25:55 <Talliesin> .googlecount "title goes here"

22:25:56 <datum> "title goes here": 21,600

22:26:04 <sbp> heh, heh

22:26:41 <Talliesin> .googlecount "meaningful title goes here"

22:26:42 <datum> "meaningful title goes here": 6

22:26:59 <sbp> there was a proposal on www-html to merge <h1> and <title> into a single element

22:28:10 <Talliesin> .googlecount "Document title goes here - used by web search tools to identify this document."

22:28:11 <datum> "Document title goes here - used by web search tools to identify this document.": 11

22:28:31 <Talliesin> sbp, would only encourage people to not use it me things.

22:29:00 <sbp> well, it'd probably be required as <title> is now

22:29:17 <sbp> and it would be better than having people set the h1 and leave out the title

22:29:21 <dajobe> I thought <title> was the only required html element?

22:29:22 <raelity> hey folks

22:29:47 <sbp> (Aaron proposed something similar about a year ago, oddly enough)

22:29:48 <dajobe> hi

22:29:53 <Talliesin> dajobe - before 1999 it was

22:29:56 <sbp> hey Rael

22:30:07 <sbp> yeah. not in XHTML

22:30:27 <Talliesin> well, it wasn't the only required element, but it was the only required element that had to be explicitly tagged in the source

22:31:27 <Talliesin> html, head and body were also required, but their opening and ending tags where optional.

22:32:18 <dajobe> it must be web history day today ;)

22:33:15 <Talliesin> shoulders of giants

22:34:03 <Talliesin> At least this is once channel were one is safe from "I was on the web before any of you" arguments.

22:38:26 <DanC> G:see [review of earlier draft by timbl|http://www.w3.org/2002/07/29-rdfcadm-tbl.html]

22:38:26 <dc_rdfig> Added comment G9.

22:42:07 * DanC is away: family time

23:01:38 <MysticOne> hey guys, question ... does timmmit ever actually get on IRC and talk? :)

23:01:57 <sbp> quite often, yes. I think he's on holiday or something, now

23:02:21 <MysticOne> ahh... I've been trying to get up with him to ask him a few things, but he hasn't been on

23:02:26 <MysticOne> oh well, holiday would explain it :)

23:02:29 <MysticOne> thanks for the info :)

23:02:53 <sbp> you're welcome. and happy birthday once again :-)

23:02:58 <MysticOne> hehehe, thanks :)

23:03:02 * MysticOne waves

23:11:47 <LotR> Talliesin: doesn't TimBL gloat about that? ;)

23:12:27 <Talliesin> yeah, but he's never actually online when I am, just connected to discourage such gloatting in others ;)

23:13:39 <LotR> heh