Semantic Web Interest Group IRC Chat Logs for 2004-08-23

00:22:54 <dmwaters> {global notice} Hi all! please take a look at 'http://freenode.net/news.shtml' We have an outage coming up at 11:00 tomorrow morning for some maintainence. This will be a rather large outage that will affect everyone. Thank you for your patience, and thank you for using freenode!

00:54:48 * KjetilK seeks the advice of RDF hackers, anybody here?

00:56:20 <Talliesin> It's probably best just to state your question/problem/pondering and then hope for the best.

00:56:27 <KjetilK> yup

00:56:55 <KjetilK> well, I'm thinking about carefully building page titles for a website...

00:57:25 <KjetilK> ...and I figured I'd store different parts of the title in a static RDF file

00:57:43 <KjetilK> after all, parts of a title are also descriptive...

00:58:07 <KjetilK> Obviously, I have a site title, stored in dc:title element

00:58:25 <KjetilK> but....:

00:59:50 <KjetilK> if I build a page title, as in HTML title element, then the HTML title element would contain the title of the article itself'

01:00:32 <KjetilK> which should also be stored in a dc:title element of the resource itself

01:01:22 <KjetilK> but the HTML title element should contain the dc:title of the page as well as a short version of the title of the site

01:01:47 <KjetilK> to distinguish it from similar and perhaps equally titled documents on other sites

01:02:35 <KjetilK> Well, the question is simply: What would be the best way to mark up this short title for the site...?

01:02:42 <Talliesin> !

01:02:42 <wh4experimental> no command found. Commands are: , !source, !iata, !airport, !foaf:, !pic, !put, !remove

01:02:48 <Talliesin> heh

01:03:02 <Talliesin> wh4experimental butting in there.

01:03:12 <KjetilK> hehe

01:04:15 <Talliesin> Well, one approach is to not worry about the title being the same as titles on other sites. There's no law that says they have to be unique.

01:04:28 <Talliesin> There are advantages though.

01:05:04 * Arnia wonders if sh1mmer will bring up composite IFPs or whatever they're better called :)

01:05:30 <Talliesin> But in that case you could still have say <dc:title>A day in the life of an RDF Hacker</dc:title> and <html:title>MySite.com &mdash; A day in the life of an RDF Hacker</html:title>

01:05:52 <KjetilK> well, if you bookmark a lot of pages with similar subjects, it is very useful to have something distinguishing

01:06:19 <KjetilK> yup, that's the idea...

01:06:43 <Talliesin> Arguably the title is the "A day.." bit, the MySite.com only comes in a particular rendering of the same article. After all if someone paid you to let them republish the article elsewhere they'd likely want to drop that bit.

01:06:55 <Arnia> KjetilK: I don't think DC was devised to work with the concept of a 'site'... pages are placeless, URIs are atoms *ducks before someone throws something at him*

01:07:02 <sh1mmer> Arnia wha?

01:07:30 <Talliesin> DC works well enought with sites of particular sorts of content.

01:07:37 * Arnia is in a silly mood

01:07:40 <KjetilK> Arnia, yup, definately, it is not really what I was getting at...

01:07:52 <sh1mmer> Arnia "composite IFPs" or "Unambigious Property Constellations"

01:08:03 * Talliesin may object to the "Document in a loose sense" concept, but some resources are still documents even so.

01:08:06 <Arnia> Talliesin: Yeah, but the concepts are far better at marking up stuff in a placeless way.

01:08:16 <sh1mmer> .gc cifp

01:08:17 <Arnia> sh1mmer: danke

01:08:40 * Arnia can't believe he just said 'marking up' :p

01:09:08 <Aredridel> Has anyone actually -made- a vocabulary for saying X resource is_a (page|entrypoint-to-asite|document|abstract-resource) ?

01:09:09 <sh1mmer> ha

01:10:07 <Talliesin> Really I think KjetilK's question comes down to which he's describing in the RDF, the article, or a particular HTML version of the article. The title of the article (re-using my example above) is "A day in...", the title of the HTML version is "MySite.com &mdash; A day in..."

01:10:22 <KjetilK> well, say, I have a site rdfhacking.org

01:10:25 <KjetilK> yup

01:10:43 <Talliesin> Frankly I would see more point in describing the article than the HTML version.

01:10:53 <KjetilK> well, if I describe rdfhacking.org as a whole

01:10:55 <Arnia> Yeah, I agree with that

01:10:56 <Aredridel> Heh. And "url represents abstract-thing-foo". That would be immensely useful.

01:11:24 <Arnia> Aredridel: Isn't that what rdf:type allows? :)

01:11:31 <Aredridel> Oh, hehe.

01:11:31 * Arnia points to cyc ;)

01:11:33 <Talliesin> Which makes more sense in the description "This was a movingly written work", or "I hate the way the source code was indented"?

01:11:35 * Aredridel goes and reads.

01:12:16 * KjetilK admits things went a bit too fast for him now... :-)

01:13:25 <Talliesin> Of course RDF lets you do either. RDF lets you describe *anything*.

01:13:44 * Arnia huggles RDF M&S ;)

01:13:44 <Talliesin> It's not so much an RDF question as a question of what you want to have descriptions of.

01:14:25 <KjetilK> right... Uhm, I'll try again...:

01:14:38 <KjetilK> I describe my article as

01:14:40 <KjetilK> <rdf:Description rdf:about="http://rdfhacking.org/a-day-in-my-life">

01:14:40 <KjetilK> <dc:title>A day in the life of a RDF hacker</dc:title>

01:14:44 <KjetilK> ...

01:15:00 <KjetilK> but I also describe my site as

01:15:07 <KjetilK> <rdf:Description rdf:about="http://rdfhacking.org/">

01:15:07 <KjetilK> <dc:title>My adventures in RDF</dc:title>

01:15:09 <KjetilK> ...

01:15:26 <KjetilK> or something like that...

01:15:33 <Talliesin> That seems reasonable.

01:16:10 <KjetilK> but in the Real World[tm], it is useful that every HTML page has a html:title that includes something from both....

01:16:31 <Aredridel> So has anyone defined rdfs:Classes of things for "page" and "abstract" and other types?

01:16:52 <KjetilK> but My adventures in RDF -- A day in the life of a RDF hacker is too much...

01:17:06 <Arnia> Aredridel: I'm sure there are such things in Cyc, and there is definitely foaf:Document

01:17:11 <crschmidt> Aredridel: foaf:Document is a better thing for "page"

01:17:12 <Talliesin> Yes. Well, there are arguments against the site-name in title, since most times you see the title you can quickly find the sitename without following the link, I'm undecided on it.

01:17:31 * Aredridel nods

01:17:35 <Arnia> Talliesin: I'd argue its irrelevant to the information being modelled

01:17:39 <Aredridel> Aright. Hadn't seen foaf:Document.

01:17:45 <Arnia> Talliesin: Its a UI frivelry

01:17:55 <KjetilK> yup

01:18:15 <Arnia> (ie it should be generated by software rather than included in the information model)

01:18:21 <Talliesin> Yep.

01:18:43 <KjetilK> yep

01:18:53 <Talliesin> Unless you really want to describe the HTML itself. There's no law that says you can't, it's just not the most useful thing to do.

01:19:22 <Arnia> Yeah :)

01:19:33 <Talliesin> For a start machines are already good at reading HTML in and of itself. It's the article (or whatever) they don't understand.

01:19:34 <KjetilK> the HTML titles are going to be generated by software,

01:22:54 <KjetilK> it is just about when dealing with lots of similar documents, say you have a bookmarks folder with 10 "Perl Introductions", all different, but all generic...

01:23:02 * KjetilK has such a folder

01:23:17 <Arnia> KjetilK: Then they are all titled the same :)

01:23:25 <KjetilK> right

01:23:33 * Talliesin has such a folder (well not Perl) but the tooltips sort me out.

01:23:38 <Talliesin> Metadata rules! :)

01:23:53 <Arnia> Bad modelling shouldn't be used to overcome the failure of UI

01:24:07 <Arnia> Shoot UI guys instead :)

01:24:09 <KjetilK> but since they are all on different sites, having a site-name in there would be a great mnomic device so to speak...

01:24:09 <Talliesin> Yes.

01:24:27 <KjetilK> boom

01:24:33 <Talliesin> Build the best of all possible models. Build the best of all possible UIs. Then make the two work together :)

01:25:10 <KjetilK> yep, I agree, of course, but the Real World has bad UIs...

01:26:02 <Arnia> KjetilK: What software reads your RDF file directly?

01:26:12 * Arnia replaces the term file with model

01:26:13 <KjetilK> some XSLT...

01:26:21 * Arnia blinks

01:26:26 <Arnia> Using Treehugger?

01:26:50 <KjetilK> uhm, probably not, since I don't know what Treehugger is...

01:26:55 <KjetilK> googles

01:27:15 <crschmidt> XSLT doesn't adequately understand RDF concepts, but it works to get information out of a file structured in a prepared way.

01:27:32 <Arnia> Which is what treehugger does :)

01:27:41 <KjetilK> hehe

01:27:47 * Arnia contemplates that XSLT doesn't understand RDF at all :p

01:27:56 <crschmidt> yeah.

01:27:59 * KjetilK discovers he was at the Treehugger site a couple of days ago... :-)

01:27:59 <crschmidt> that.

01:28:00 <Arnia> 4Suite have an XSLT extension for RDF models

01:28:08 * crschmidt is tired.

01:28:22 * crschmidt decides that he should go to bed to solve that.

01:28:39 <Talliesin> A crazy plan, but it just might work.

01:28:54 <KjetilK> but I just need to grab the contents of the elements for my purpose... It doesn't really need to be RDF, but if it is, it can be reused

01:28:56 <crschmidt> I'm actually going to bed to watch a movie with my girlfriend.

01:29:01 <crschmidt> Which is more likely to happen :)

01:29:26 <KjetilK> ...and doesn't sound too bad at all... :-)

01:29:54 <Arnia> KjetilK: Collapsing graph to tree is not a one to one mapping... which makes XSLT erm fraught

01:30:00 <crschmidt> [a menow:Status; menow:watching "Moulin Rouge" ] .

01:30:10 <Talliesin> Yeah, if you need to have some XML sitting there for you to do stuff with then it might as well be understood by the RDF-savvy, even if that's not what you're doing.

01:30:13 <sh1mmer> Arnia so do wazzit, xmlspy ppl

01:30:29 <Talliesin> (At least that's what I regularly do, so I have to defend such a move :)

01:31:06 <Arnia> sh1mmer: eh?

01:31:33 <sh1mmer> Arnia altova are starting to get nice rdf tools and they have some elite xslt stuff already

01:31:51 <sh1mmer> i know i am one of the people with the w3c blanket licence on all their products :)

01:32:14 <Arnia> Ah... doesn't change the fact that XSLT is not suitable for general RDF work. You need your RDF/XML serialisation in a particular form

01:32:18 <KjetilK> Thanks for the warnings.... Except for my FOAF file, this is my first real adventure into RDF, so it is good to be warned about doing things wrong...

01:32:23 <Talliesin> There are at least two films called "Moulin Rouge", you need a universally unique identifier for your resource. Preferably one which can allow anyone to coin new ones, and with a degree of human-readability :)

01:32:33 * Arnia screeches

01:32:40 <Arnia> 'foaf file' ;)

01:33:11 <KjetilK> uhm, sorry...

01:33:15 <sh1mmer> Arnia well i got 'told' about rdf/xml and flattening earlier

01:33:25 <sh1mmer> i still maintain that nesting implies reification

01:33:36 <Arnia> Sorry... I just have a pet hate for the term :)

01:33:40 <bijan> nesting?

01:33:41 <Talliesin> ?

01:33:45 <bijan> In RDF/XML?

01:33:49 <Talliesin> Right,

01:33:52 <bijan> Implies, wha?

01:33:55 <Arnia> Its an RDF/XML file IMO, not just FOAF :)

01:34:02 <Talliesin> I'm now saying "?" to about three people in this channel at the same time.

01:34:14 * Arnia hugs Talliesin

01:34:15 <KjetilK> lol

01:34:20 <bijan> ?

01:34:41 <sh1mmer> earlier example

01:34:42 <sh1mmer> <foaf:person><foaf:nick>sh1m</foaf:nick><foaf:knows><foaf:person><foaf:knows><foaf:person><foaf:nick>sh1m</foaf:nick>some assertion</foaf:person></foaf:knows><foaf:person></foaf:knows></foaf:person>

01:34:43 <bijan> Perhaps you should upgrade to @#8!(@?

01:34:53 <sh1mmer> i am using foaf:nick as a shorthand ifp

01:34:59 <sh1mmer> i know it isn't but i was being lazy

01:35:17 <Talliesin> It's purpose is to hold FOAF data, it's therefore a FOAF file. Just like the plain text file with hosts information is referred to as a "hosts file" not as a "text file that happens to have hosts information".

01:35:56 * bijan gives up on thread tracking but doesn't believe there is any sort of RDF/XML nesting which "implies reification"

01:36:02 <Arnia> Talliesin: Nah, it describes me. What vocabularies I use are irrelevant. Its purpose is describing me

01:36:16 <sh1mmer> the general point being that if my foaf:person knows some other foaf:person who knows the first foaf:person and says some information about them. i read that as the first foaf:person is asserting the second foaf:person made an assertion about him

01:36:23 <Arnia> Which is why I call mine 'about-me.rdf' not 'foaf.rdf'

01:36:49 <Talliesin> Why read it like that sh1mmer?

01:36:50 <KjetilK> oh, so it is the about-me-file, then? :-)

01:37:14 <Talliesin> KjetilK, if it's mainly about you it is.

01:37:18 <bijan> Well, not that I followed that in detail, but that certain doesn't have anything to do with any syntactic sense of nesting in rdf/xml, which is what I thought it was about :)

01:37:21 <Arnia> KjetilK: Its just an RDF/XML file that happens to describe me... mine at least

01:37:35 <Talliesin> It's also and RDF/XML file, and also a FOAF file, and possibly also other things.

01:37:48 <sh1mmer> well the second foaf person is asserting about the first foaf person only because its the first foaf person who is making an assertion that the second foaf person is making these things it isn't authoritive

01:38:04 <sh1mmer> hence reification suggests the context

01:38:12 <Talliesin> !

01:38:12 <wh4experimental> no command found. Commands are: , !source, !iata, !airport, !foaf:, !pic, !put, !remove

01:38:22 <Talliesin> SHUT UP wh4experimental!

01:38:23 <sh1mmer> flattening in this case seems to remove stuff which is more than valid but actually important

01:38:25 <Arnia> Talliesin: It doesn't contain just FOAF though... I don't like the idea of calling it a 'FOAF' file because it misses the point that the file is describing me. It doesn't matter what vocab I use to do that

01:38:33 <crschmidt> Talliesin: stop starting lines with !

01:38:34 <crschmidt> ;)

01:38:42 <Talliesin> Whoever published the RDF/XML file is asserting the whole lot.

01:38:59 <sh1mmer> Talliesin right, but some of it they are asserting that someone else asserted

01:39:05 <sh1mmer> important differentiation

01:39:14 <sh1mmer> flattening kills that

01:39:34 <Talliesin> I like the idea of calling it a "FOAF" file, because "FOAF" is a "brand" that is getting good attention and isn't owned by anyone evil. It's a hype thing, but it's a good hype thing.

01:39:41 <sh1mmer> so it means i can say "i am a bastard" but cant say you say "you are a bastard"

01:39:45 <Arnia> Talliesin: If I change the version of FOAF used (at some point in the future) I'm changing the namespace and so I'm actually using a completely different set of concepts (although some mappings may exist)

01:39:51 <sh1mmer> Talliesin do you call your xml files xml.xml ?

01:40:06 <Talliesin> Why yes, sometimes I do.

01:40:10 <sh1mmer> bah

01:40:13 <sh1mmer> biatch

01:40:18 <sh1mmer> :P

01:40:19 <Talliesin> (Sometimes I have xml files about xml)

01:40:39 <Arnia> Talliesin: Yeah, but I don't like people getting confused by thinking its just FOAF vocab.

01:40:40 <sh1mmer> geeks can be so obtuse

01:41:07 <Talliesin> I read RDF/XML and I see triples (not necessarily the only perspective I see), so it's already "flat" to me.

01:41:16 <Arnia> Talliesin: Its a good brand, but when dealing with the technical side its good to be more explicit

01:41:22 <Arnia> I agree with Talliesin

01:41:28 <Arnia> Its 'flat' already

01:41:47 <Arnia> It just has a nice(!) serialisation that happens to use embedding

01:41:49 <Talliesin> Yeah, but people are always going to be confused one way or another. Marketing is all about controlling what way they're confused :)

01:42:30 * KjetilK puts the prototype RDF file he was thinking about using to describe his site as a whole at http://dev.kjernsmo.net/tmp/main.rdf and wonders if it looks sensible...

01:42:46 <Arnia> Talliesin: I wish to hype the 'describe anything' abilities of RDF over a single vocabulary which I just know someone's going to try and make an XML out of :)

01:42:52 <Talliesin> At most seeing <A> <z> <B> <y> <C> makes me see the graph (A) -z-> (B) -y-> (C)

01:42:59 <sh1mmer> but rdf/xml isn't flat

01:43:03 <sh1mmer> thats the problem

01:43:06 <sh1mmer> it takes shortcuts

01:43:17 <Arnia> sh1mmer: Only as a convenience

01:43:18 <sh1mmer> as soon as it does that you get topology not just serialization

01:43:19 <Talliesin> Shortcuts good!

01:43:35 <sh1mmer> Arnia right which is why i am saying there is an implication of reification

01:43:46 <sh1mmer> or at least conveyed structure which is lost

01:43:49 <Talliesin> I take shortcuts all the time. That's one of the reasons I generally wear large boots.

01:43:57 <Arnia> KjetilK: Use a <link rel="meta" title="Site" href="http://dev.kjernsmo.net/tmp/main.rdf" /> in your HTML

01:44:26 <crschmidt> Arnia: why title="Site"?

01:44:27 <KjetilK> Arnia, yup, of course!

01:44:33 <sh1mmer> meh

01:44:35 <sh1mmer> im going to bed

01:44:40 <sh1mmer> im fucking knackered

01:44:43 <Talliesin> crschmidt, because it's describing the site?

01:44:44 <Arnia> crschmidt: I was being arbitrary tbh :)

01:44:52 * sh1mmer hopes he wakes up in time in the morning

01:45:09 <crschmidt> I was just wondering if there's a reason to choose anything specific in "title" - specific auto discovery type things that get found with special names

01:45:22 <crschmidt> since I am thinking about dropping foaf into the sites I work on for wedu

01:45:31 <KjetilK> but if this RDF looks sensible, surely, it is going to be trivial to get the title from it using XSLT?

01:45:38 <Talliesin> Arnia, I agree that "describe anything" rocks, however "Anything" and "Nothing" are closely related concepts until you've shown people something a bit more concrete.

01:45:40 <Arnia> crschmidt: I dislike using title for that purpose... it smacks of 'magic'

01:45:55 * KjetilK waves at sh1mmer

01:46:14 <sh1mmer> gnight all

01:47:06 <Arnia> Talliesin: But we shouldn't allow them to assume that 'FOAF' is a file format. It will just confuse them later and its the kind of assumption that's hard to undo

01:47:37 <Arnia> crschmidt: I use titles to separate things more for my own benefit where I can tbh

01:47:49 <Arnia> Talliesin: There has to be a middle ground

01:47:55 <Talliesin> Well it depends. The average user need never see a pointy bracket at all.

01:48:26 <Talliesin> I dislike the "file" bit more than the "FOAF" bit.

01:48:27 <Arnia> Talliesin: Its not the average user we're selling to. The average user shouldn't care their information is in RDF, XML or SQL

01:48:45 * Talliesin is selling to the average user.

01:49:03 <Talliesin> Well, I would be if I stopped chatting and started hacking.

01:49:11 <KjetilK> lol

01:49:55 <Arnia> Talliesin: You're selling them 'FOAF' as a shiny end-user product? You're not, rather, selling the concept of 'sharable self-description' (or what have you)

01:49:59 <Arnia> ?

01:51:03 <Talliesin> That tipjar thing I've had about a day's work left to do for the last 3 weeks. I want something that average users can go, Okay, so crschmidt put this FOAF thing on my LiveJournal, and then it got linked to this other thing (other thing provided by me, but I'm not on LJ so they won't know who I am) and the my friends knew what to get me for Christmas/Channuka/Yule/Kwanzaa - FOAF rocks!".

01:52:06 <Talliesin> I did do about half a days work on it over my holidays. So I'm down to half a day left.

01:52:33 <Talliesin> Right now though my ret

01:52:54 <Arnia> I don't see how that's exclusive to saying 'using FOAF and a wide-range of other vocabularies I can describe myself for everyone to share'

01:52:57 <Talliesin> retarded Norton software is protecting my IDE from my webserver, or the other way around, or both.

01:53:12 <Talliesin> It's not exclusive.

01:53:35 <Talliesin> It's all a matter of emphasis.

01:55:07 <Talliesin> FOAF is marketing. http://xmlns.com/foaf/0.1/ is one of a wide range of vocabularies.

01:57:08 <Talliesin> "marketing" in a loose sense of course.

01:57:11 <Arnia> No I agree with that... I'm just someone who gets a bee in his bonnet and the Bee de Jour is people using FOAF as if it were a file format. It feels like it misses the point of allowing graphs to merge and connect. Just me though

01:57:15 <Talliesin> It isn't evil for one thing.

01:58:00 <Talliesin> See, I already have a bee in my bonet about people talking about file formats as if they were more than a means to an end.

01:58:24 <Talliesin> So when people start saying "FOAF file" rather than "RDF/XML file" I see that as progress.

01:58:32 <Arnia> Well, I see this as a continuation of that

01:58:33 <Talliesin> Sure, I'll call it a

01:58:46 <Talliesin> RDF/XML file. But the means is my job, not the end.

01:59:16 <Arnia> Its not FOAF... its not just RDF/XML. Its a description of some concept, me, and how I connect into the world

01:59:36 <Talliesin> You aren't going to get many people to make that jump straight away.

02:00:52 <Arnia> I think it would be a powerful idea to sell though - "stop writing xml, start describing the real world"

02:00:55 <Talliesin> Not the people who think file extensions are cast in stone and assigned by some mysterious body somewhere in the pecking order between the UN and God.

02:01:07 <Talliesin> Yeah, but we need a name for that.

02:01:14 <Talliesin> I think FOAF is as good a name as any.

02:01:45 <Arnia> Talliesin: Don't mention that atm... I'm writing a Shared-Mime-Info system and apparently file extensions are higher priority than magic values :/

02:02:01 <Aredridel> Hm. Is there a vocabulary for describing conversation yet? he-said, she-said sort of things? Maybe with mode as well, he-wrote, she-wrote?

02:02:02 <Arnia> I know why they've done it

02:02:15 <Arnia> Aredridel: Yes... Thread description

02:02:21 <Aredridel> Ah, cool. URL handy?

02:02:24 <Arnia> I just don't like that it had to be done

02:02:28 <Arnia> Aredridel: Wait a sec

02:02:34 <Aredridel> Thank you!

02:03:18 <Talliesin> Arnia, I sometimes do deeply perverse things with file extensions just to play with their minds.

02:03:29 <Arnia> Talliesin: Such as?

02:03:35 * Aredridel laughs.

02:03:49 <Aredridel> .txt that's binary. .jpeg that's PNG. .avi that's MPEG.

02:03:53 <Arnia> Aredridel: http://www.schemaweb.info/schema/SchemaDetails.aspx?id=199

02:03:56 <Talliesin> Yes.

02:04:13 <Aredridel> Arnia, thank you.

02:04:15 <Talliesin> Good fun for demo sites you're only going to show to a couple of people.

02:04:26 <Talliesin> "Hey, how did you make this jpeg have selectable text?"

02:04:37 <Talliesin> "Skillz"

02:05:19 <Arnia> Isn't using FOAF like that just a direct substitute for the RDF brand?

02:06:00 <Aredridel> Hehe, that's perfect.

02:06:27 <Aredridel> (My directory index entirely eschews file suffixes, and just strips it.)

02:06:38 <Aredridel> (It may use them internally, but it doesn't expose it)

02:08:25 <Talliesin> It's a brand for a particular purpose.

02:08:49 <Talliesin> And it's a brand that's likely to be easier to sell to a particular group.

02:09:17 <Talliesin> And then they'll have bought into it when they're still saying "RDF is too complicated and needlessly verbose".

02:09:25 * Aredridel laughs.

02:09:26 <Talliesin> And then we go "Ta da!" :)

02:09:28 <Aredridel> Good, that.

02:09:37 <Aredridel> Though you run the risk of another RSS vs RSS war.

02:09:44 <Arnia> That's my fear

02:09:46 <KjetilK> Arnia: TDL looks very interesting, I'm surely going to be using that

02:09:50 <Aredridel> "FOAF 2.0" -- XML but not RDF

02:10:12 * Arnia can think of someone who'd do that

02:10:16 <Arnia> *coughs*

02:10:42 <Talliesin> The biggest would be rivals to FOAF are going for walled gardens. They're practically designed to be routed-around.

02:10:47 <Arnia> KjetilK: I'm using TDL in my Zope 3 comments system

02:10:56 <Talliesin> Otherwise they'll have to get with the program.

02:11:18 <KjetilK> Arnia: Cool! I'm implementing a comments system in my TABOO framework

02:11:21 <Arnia> Talliesin: Such as?

02:11:32 <Talliesin> The YASNs

02:11:42 <Arnia> KjetilK: Mine is unobtrusive and works with all existing and future content types :)

02:11:55 <KjetilK> hehe, sounds very good!

02:12:16 <Arnia> KjetilK: Just install, indicate which types you want to comment (one line of XML per type) and woo! Comment system installed and working :)

02:12:26 * Arnia strokes Zope 3 a bit

02:12:47 <Talliesin> I think the biggest reason for the RSS vs. RSS war isn't anything to do with the formats or the people involved (though that was involved in it persisting), but the fact that both RSS0.9 and RSS0.91 came from Netscape.

02:13:02 <Aredridel> Hm. Could be.

02:13:10 <KjetilK> Arnia: Got an URL I can play with...?

02:13:19 <Talliesin> The same people said "Ta da!" twice. The echoing "Ta da!"s was where the war was.

02:13:33 <Aredridel> Yeah.

02:13:44 <Arnia> KjetilK: Not for the comments system yet. For much else I've been working on though

02:14:05 <Talliesin> So, we just need to say "Ta da!" once. Future "Ta da!"s must be genuinely new improvements built on that.

02:14:36 <Arnia> KjetilK: I'm building a lot of infrastructure for using Zope3 for clients now, in the run up to release

02:14:55 * Arnia puts some Reel Big Fish on

02:15:08 <KjetilK> mmmm

02:15:12 <KjetilK> nice

02:16:24 <Arnia> KjetilK: Zope3 is a completely different animal to Zope2... its not even the same phylum in many ways. Much nicer to use :)

02:16:29 <Talliesin> BTW, Aredridel, I agree with eschewing suffixes, that's what I do with real stuff once I've had fun with selectable-text jpegs.

02:17:00 * Arnia almost gave himself the wrong injection :/

02:17:02 * Aredridel nods.

02:17:22 <Aredridel> Hm. I'll have to try Zope 3.

02:17:31 * KjetilK admits to not having checked out Zope in a long time

02:17:42 <Arnia> Aredridel: http://dev.zope.org

02:17:57 <KjetilK> I went with AxKit, the very clean separation of logic from markup is very appealing

02:18:14 <Arnia> The docs are still being written, but two books (very good books) are being written and published soon

02:18:35 <Aredridel> Mmm, nice.

02:18:42 <Arnia> KjetilK: Zope 3 is a component architecture, everything is separated and can work in blissfull ignorance :)

02:18:46 * Aredridel files it for "sometime this week when I have a spare second"

02:18:46 <Arnia> -l

02:19:31 <KjetilK> yup

02:19:34 <Talliesin> Right, I'm off to bed. See ye later.

02:19:35 <Arnia> KjetilK: Everything is structured using interfaces, utilities and adapters. Its liberating

02:19:39 <Arnia> Talliesin: Night :)

02:19:48 <KjetilK> Nighty!

02:19:59 <KjetilK> nice!

02:21:04 <Arnia> KjetilK: One of the things I wrote this week was a pagelet system in order to modularise the page rendering (for the web browser views) along interface lines

02:21:23 <Arnia> KjetilK: It works rather nicely

02:21:36 * KjetilK thinks he should be getting to bed, it is getting early [sic!] around here... But would be grateful if someone could take a quick look at http://dev.kjernsmo.net/tmp/main.rdf

02:22:23 <KjetilK> yup, I guess I should look at it once, allthough I'm not going to be leaving AxKit now that I've worked with it for quite some time, got to finish what I started

02:22:43 <Arnia> KjetilK: Yeah, understandable

02:22:50 <KjetilK> s/once/at some point/

02:23:51 <KjetilK> there are some very nice people working on Axkit too, I've come to know quite a few people there

02:24:05 <Aredridel> Hehe, yes, understandable.

02:24:47 <Arnia> I've just moved from Zope 2, so I'm keeping my friends in the community :)

02:24:56 <KjetilK> hehe

02:24:58 * Arnia was one of the core developers on Plone

02:25:14 <KjetilK> ah, ok!

02:25:24 <Aredridel> Mmm!

02:25:38 <KjetilK> It is remarkable how much the people of a good community means

02:25:45 <Aredridel> Isn't it?

02:25:55 <Aredridel> It's why I work on what I do. RDF, PLD GNU/Linux, and Ruby.

02:26:09 <Arnia> My business was the majority of the UI team... so we're making an even better markup for Zope3 :)

02:26:09 <KjetilK> yup

02:28:09 * KjetilK hasn't really liked Perl as a language that much, but always found himself returning to Perl because of all the code at CPAN, and the nice people on mailing lists and IRC

02:28:18 <Arnia> Communities are great... I love the way that you run into the same people all the time :)

02:28:23 <Aredridel> Yeah..

02:28:30 <KjetilK> but then, it is clear there are many other nice communities too

02:28:36 <Aredridel> ...

02:28:43 <Arnia> The Haskell community is fantastic

02:28:45 <KjetilK> so, we'll have cross pollination! :-)

02:28:55 <Aredridel> Has anyone ever thought of an xmlns: URI "protocol"??

02:29:18 <Arnia> Huge cross-polination between python and haskell for some reason

02:29:18 <Aredridel> ":xmlns:http://example.org/foo-namespace"

02:29:32 <Aredridel> Interesting.

02:29:32 <Arnia> Aredridel: What for?

02:29:52 <Aredridel> To help encourage separation of resource from namespace.

02:29:53 <KjetilK> hehe, I first saw haskell a couple of days ago, don't know why

02:30:03 <Arnia> (and the zope, python and twisted communities are very 'splurged' together)

02:30:16 <Arnia> Aredridel: I don't see what you mean?

02:30:17 * KjetilK went to the university where OOP was invented...

02:30:34 <Arnia> KjetilK: Cool... but I still think FP is more elegant :)

02:30:47 <KjetilK> hehe, but then, Python...?

02:30:50 <Arnia> KjetilK: Component architectures seem to merge concepts from both

02:30:59 <Aredridel> Arnia: To make a purely abstract space (ala URNs), but with a more easily understood way to make them not collide.

02:31:20 <Arnia> KjetilK: I confess to using generator comprehensions, and maps and folds and filters :)

02:31:33 <Arnia> Aredridel: Why not use URLs?

02:31:37 * KjetilK like OO, had some good teachers, but has forgotten quite a lot

02:31:52 <Aredridel> Arnia: it'd be another flavour of URL, but one that's obviously not retrievable.

02:32:29 <Arnia> KjetilK: I like it, just I don't consider most languages as being OO... Smalltalk is one of the few exceptions

02:33:35 <Arnia> Aredridel: Why not say <http://example.com/> rdf:type ex:NonRetrievableResource

02:33:46 <Arnia> No need to change the URIs

02:33:49 <Aredridel> Arnia: that's good too, but doesn't stand on its own.

02:34:04 <Arnia> Aredridel: Nothing does

02:35:09 <KjetilK> Arnia: Yup. SIMULA was the language we were taught. Very nice language, too bad it was killed by beaurocratic stupidity

02:35:46 <Arnia> I don't see how a new protocol is any better

02:36:24 <Aredridel> Mmm. Just a different version of urn, I suppose.

02:36:30 <Arnia> KjetilK: Yeah. I have a taste for 'esoteric' languages tbh

02:36:49 <Arnia> Aredridel: Yeah... one that existing systems don't understand :)

02:37:45 <Aredridel> Arnia: it'd work well that way, since there's almost no semantics to it, not even possible retrievability

02:37:56 <KjetilK> cool

02:38:02 * Aredridel shrugs.

02:38:20 <Aredridel> It was the result of a one-neuron-firing anyway.

02:38:48 * KjetilK -> bed

02:38:53 <Arnia> Night :)

02:39:03 <KjetilK> nighty, night!

02:39:14 <Aredridel> Night!

02:50:46 <Aredridel> Hm. What's the latest best way to ~quote things?

02:51:13 <Aredridel> so one can say "John said "<foo> is_a dog"" and "Sarah said "<foo> is_not_a dog""

02:52:03 <Arnia> Aredridel: The official means is reification

02:52:27 <Aredridel> That's what I thought.

02:52:32 <Aredridel> (that's always how I'd leaned)

02:54:47 <Arnia> Right... I need to sleep

02:54:51 <Arnia> Night

02:55:22 <Aredridel> Night!

03:13:51 <mbafk> is anyone awake?

03:33:34 <lilo> [Global Notice] Hi all. It's 03:33, we're going to do a bit of rehubbing over the next few minutes to get us ready for our outage at 11:00 (http://freenode.net/news.shtml). Thanks for your patience, and thank you for using freenode!

03:34:00 <golbeck_> golbeck_ is now known as golbeck

03:41:31 <ChanServ> [#rdfig] This channel is logged and blogged: http://logicerror.com/rdfIRCWelcome

03:42:03 <khaled> hi

03:42:38 <khaled> how do i group a class tree ?

03:43:44 <lilo> [Global Notice] All done! It's 03:43, and our next activity will be the actual outage scheduled for 11:00 and lasting about one hour - http://freenode.net/news.shtml - thanks and have a great morning!

05:25:36 * sanctius Bonjour tout le monde / Good Morning all

06:18:29 * Aredridel gets RDF4R working again.

06:21:50 <qmacro> hey libby - good luck with the driving test

06:22:15 <libby> heh, thanks qmacro!

06:22:27 <libby> I think I'll view this one as a practice...

06:22:46 <libby> how you doing? back home?

06:22:57 <qmacro> I was thinking of you as I was getting lost in Antwerp - you should have come with me and driven - good practice!

06:23:25 <libby> in antwerp?

06:23:26 <qmacro> yes - drove back via the channel tunnel yesterday - not a bad journey at all

06:23:37 <libby> ah yes, I forgot you were driving

06:24:06 <qmacro> major longterm roadworks in Antwerp - usually go through them but this time tried to avoid them via the ring road #2, but went wrong and ended up going the wrong way out of Antwerp

06:24:17 <libby> erk

06:24:36 <qmacro> indeed. The air in the car was blue

06:24:40 <libby> sorry to rush off bt. I was going to show an annotation of your photos...

06:24:49 <libby> s/bt/btw/

06:24:49 <qmacro> oo

06:24:53 <qmacro> no probs

06:25:26 <qmacro> the foaf braindump talk was very useful for me

06:26:09 <libby> that's good. danbri's summary was very nice I thought. just could have donw with more time

06:27:50 <libby> poor paula radcliffe :(

06:27:57 <qmacro> time - most precious commodity

06:28:00 <qmacro> yes

06:28:04 <libby> yeah

06:29:10 * qmacro at work printing tons of RDF background stuff out

06:29:17 <libby> heh

06:29:40 * libby up early cos builders coming...

06:43:07 <qmacro> qmacro is now known as qmacro_work

06:43:21 <qmacro_work> teleconf already :-S

06:51:52 <libby> :/

06:55:41 * qmacro_work sees http://www.flickr.com/photo.gne?id=232491 and is shocked to find he has a semi-bald head!

06:56:04 <Aredridel> Hehe.

06:58:41 <Aredridel> Night, all!

06:58:56 <Aredridel> I'm headed to bed, now that I have RDF4R working.

07:34:27 <balbinus_> balbinus_ is now known as balbinus

08:04:53 <CaptSolo> hi all

08:07:03 <balbinus> hi

08:08:55 <qmacro_work> hi

08:46:12 <shellac> Hi Jipp!

08:55:54 <_jeen__> _jeen__ is now known as jeen

09:37:02 <Jipp> ls

09:45:25 <dajobe> wrong channel.txt

09:45:25 <dajobe> $

09:47:24 <Jipp> focus-follows-eyes on

09:55:27 <grirgz_> plop

09:55:33 <grirgz_> grirgz_ is now known as grirgz

10:30:31 <dmwaters> {global notice} Hi all! In about 30 minutes 3 main servers are going to go down for maintenence for about an hour. Please see 'http://www.freenode.net/news.shtml'

11:15:16 <chrisc_> chrisc_ is now known as chrisc

11:42:46 <CaptSolo> silencio

11:43:30 <balbinus> hi :)

12:07:49 <Emmy> Good morning :)

12:21:42 <balbinus> good morning Emmy :)

12:21:58 <Emmy> hey balbinus :)

12:22:20 <balbinus> wassup?

12:24:24 <Emmy> balbinus: in private...

12:24:46 <balbinus> k :)

12:31:13 <oommoo> aloha

12:31:35 <balbinus> hi

12:32:03 <oommoo> is anyone here interested in redland ?

12:32:24 <balbinus> well, normally yes :)

12:32:32 <oommoo> kewly

12:32:48 <oommoo> balbinus: I amlso like the Jena bits . . .

12:33:19 <balbinus> k :)

12:33:58 <oommoo> balbinus: what are your particular interests ?

12:34:10 <balbinus> about parsers?

12:35:33 <oommoo> I derive happiness from scanning documents and using ontology to annotate them

12:35:52 <balbinus> aoh... cool

12:38:36 <oommoo> balbinus: have you played w/ the platypus wiki implementation ?

12:38:58 <balbinus> nope

12:39:46 <oommoo> ya, it is pretty funky for a wiki, and uses jena of course

12:40:58 <balbinus> yep

12:49:12 <oommoo> danja: aloha

12:50:07 <danja> helo-a

13:05:03 <oommoo> nice, BTW, we have also started to use #semagix

14:04:08 <KjetilK> hi all! I have an issue that I've been working on since last night (there's something in the backlog), but I find it boils down to describe that the resource can have a two different titles, a short one and a long one....

14:04:31 <KjetilK> so, the question is how to mark up that...?

14:04:42 <KjetilK> I though about using a Alt container...

14:05:44 <jeen> why not simply two different properties? my:shortTitle and my:longTitle, or sth like that...

14:06:01 <KjetilK> well, it is an option, but...

14:06:15 <crschmidt> jeen: What 'my' schema includes a shortTitle and longTitle?

14:06:32 <jeen> none that I know of :)

14:06:45 <KjetilK> since dc is so well developed for the purpose, I figured it is best to stick to that...

14:06:54 * crschmidt nods

14:07:02 <crschmidt> KjetilK: you can always just use two dc:title properties

14:07:07 <KjetilK> yup

14:07:08 <jeen> and the long title does not quite qualify as a description I gather?

14:07:16 <KjetilK> nope...

14:07:32 <KjetilK> I have a different description...

14:08:03 <KjetilK> Example: http://www.osi-orientering.com/

14:08:11 <KjetilK> eh, whoops, wrong URL...

14:08:17 <KjetilK> New error to follow:

14:08:27 <KjetilK>http://dev.kjernsmo.net/tmp/main.rdf

14:08:46 <dajobe> hmm

14:09:04 * dajobe goes hunting dc_rdfig

14:09:13 <KjetilK> does it look sensible at all...?

14:12:05 <jeen> looks good to me

14:12:11 <KjetilK> ok!

14:12:15 <dajobe> yeah

14:13:09 <jeen> concerning titles: is perhaps rdfs:label or rdfs:comment something you might use? it depends a bit on your purpose with the long/short titles...

14:14:07 <KjetilK> hm, perhaps, looking

14:14:38 <KjetilK> BTW, I've updated the http://dev.kjernsmo.net/tmp/main.rdf file with something that has two titles and an Alt box, does that look sensible

14:16:10 * KjetilK obviously needs to read the rdfs spec properly...

14:18:35 <KjetilK> but at first sight, it seems not right, since I'm not describing a vocabulary....

14:29:34 <KjetilK> what I could do, though, is define a vocab where I define shortTitle element as a rdfs:label, that seems about right...

14:30:01 <balbinus> as a subPropertyOf rdfs:label, you mean?

14:30:03 <KjetilK> but then I'm trying to get away with this without creating a new vocab...

14:30:13 <KjetilK> right (?)

14:30:49 <balbinus> well, you could use DOAP (since your RDF seems to be describing a project)

14:31:10 * KjetilK googles

14:31:20 <balbinus> google for DOAP?

14:31:28 <crschmidt>http://usefulinc.com/doap

14:31:28 <dc_rdfig> A: http://usefulinc.com/doap from crschmidt

14:31:29 <balbinus> s/google/googling/

14:31:38 <crschmidt> A:| DOAP

14:31:38 <balbinus> .g DOAP

14:31:38 <dc_rdfig> Titled item A.

14:31:42 <phenny> DOAP: http://usefulinc.com/doap

14:31:43 <dc_rdfig> Label DOAP not found.

14:31:47 <balbinus> :)

14:31:52 <crschmidt> A: Vocabulary for describing a project

14:31:52 <KjetilK> hehe

14:31:53 <dc_rdfig> Added comment A1.

14:32:00 <balbinus> A:From Edd

14:32:01 <dc_rdfig> Added comment A2.

14:34:06 <KjetilK> hm, but really, it isn't the project this file is supposed to describe (though I should be publishing some DOAP for it too), it is pretty much a news and article site as a whole

14:34:20 <balbinus> aok :)

14:34:38 <KjetilK> http://mouettes.balbinus.net/doap-creator.php <- 404...

14:34:39 <dc_rdfig> B: http://mouettes.balbinus.net/doap-creator.php from KjetilK

14:35:03 <KjetilK> uhm, did I hit some bot...?

14:35:12 <crschmidt> yeah, the chump bot

14:35:15 <balbinus> B:=http://doapy.bonjourlesmouettes.org/doap-a-matic

14:35:16 <dc_rdfig> Replaced URL of B.

14:35:17 <KjetilK> hehe

14:35:18 <balbinus> :)

14:37:52 <KjetilK> very interesting, I'm creating a DOAP now...

14:38:12 * balbinus must now fix that 404...

14:40:40 <KjetilK> yup, 301 would have been cool

14:41:05 <balbinus> KjetilK: what browser do you use?

14:41:38 <KjetilK> Firefox

14:42:02 * KjetilK gets white on gray in the boxes...

14:42:07 <balbinus> well... mine works... i'm gonna telnet

14:42:42 <balbinus> what OS?

14:42:51 <KjetilK> Debian Sarge

14:43:50 <balbinus> I got a 301 with telnet

14:44:21 <KjetilK> hm...

14:45:19 <balbinus> oops

14:45:30 <balbinus> where did you find that url?

14:45:38 <KjetilK> edd's page

14:46:01 <balbinus> oh! :(

14:46:16 <balbinus> he has the very first url, that lived less than a day :(

14:46:23 <balbinus> anyway, i will make a redir :)

14:46:30 <balbinus> and tell him :)

14:47:00 <balbinus> B:|DOAP-a-matic

14:47:01 <dc_rdfig> Titled item B.

14:47:23 <balbinus> B:New URL (http://mouettes.balbinus.net/doap-a-matic.php should give you a 301)

14:47:23 <dc_rdfig> Added comment B1.

14:48:18 <KjetilK> it does!

14:48:31 <balbinus> B1:New URL (http://mouettes.balbinus.net/doap-a-matic.php gives you a 301)

14:48:32 <dc_rdfig> Replaced comment B1.

14:48:35 <balbinus> :)

14:49:44 <balbinus> B:Also "corrected" Useful, Inc. URL (now gives a 301)

14:49:45 <dc_rdfig> Added comment B2.

14:52:58 <KjetilK> Konqi also has white on light gray in the forms, though, that's a bit hard to read...

14:53:24 <balbinus> white on grey??? never seen it... gotta see what i can do

14:55:20 <KjetilK> I'd recommend always setting color and background in the same CSS block to avoid user stylesheets doing funny things...

14:56:45 <balbinus> wait a sec :)

14:56:51 <KjetilK> http://www.htmlhelp.com/tools/csscheck/ gives nice warnings about stuff like that

14:56:52 <dc_rdfig> C: http://www.htmlhelp.com/tools/csscheck/ from KjetilK

14:57:53 <balbinus> better now?

14:58:15 <KjetilK> yup, much better!

14:58:18 <KjetilK> Thanks a lot

14:58:19 <balbinus> :)

14:58:28 <balbinus> thx for signaling the bug :)

14:58:35 <KjetilK> :-)

15:06:10 <KjetilK> balbinus: The typical Perl GPL/Artistic dual license isn't represented in the form, any idea how to set that...?

15:06:24 <balbinus> aoh...

15:06:34 <balbinus> do you *truly* need it ;)?

15:06:54 <balbinus> that's possible. here's how to do it

15:07:09 <KjetilK> :-)

15:07:13 <balbinus> well

15:07:23 <balbinus> just use Artistic in DOAP-a-matic

15:07:24 <balbinus> then

15:07:39 <balbinus> create your License file for Perl GPL :)

15:08:27 <balbinus> well, i could create the license file you miss

15:08:41 <balbinus> but i cannot be sure what its URI will be

15:09:02 <KjetilK> OK, I think I'll look at the RDF I get out and try to figure it out from there...

15:09:18 <balbinus> to add another license is easy

15:09:24 <balbinus> copy/paste the existing one

15:09:28 <balbinus> and modify the URI

15:10:07 <balbinus> can you replace it with GNU GPL until it gets approved and fixed?

15:10:42 <KjetilK> yup.

15:10:45 <balbinus> well, if you need it, anyway, nothing forbids you from creating your RDF file

15:10:51 <balbinus> and putting it wherever you want

15:11:01 <KjetilK> yup

15:11:09 <balbinus> and linking to it in your DOAP

15:11:41 <KjetilK> it is not too uncommon with multiple licenses, so I guess a good solution would simply be to a multiple type box there...

15:12:02 <balbinus> the Perl+Artistic is not a license in itself, is it? it's just the fact to use both the GPL and the Artistic License, right?

15:12:10 <balbinus> KjetilK: yep :)

15:13:52 <KjetilK> AFAI understand, they are two licenses, independent, yes

15:16:48 <KjetilK> yup, now my DOAP is created and linked! Cool!

15:17:29 <balbinus> :)

15:17:32 <balbinus> where is it?

15:18:14 <KjetilK>http://www.kjetil.kjernsmo.net/software/TABOO/doap.rdf

15:18:16 <dc_rdfig> D: http://www.kjetil.kjernsmo.net/software/TABOO/doap.rdf from KjetilK

15:18:44 <balbinus> D:|The first DOAP document of KjetilK

15:18:50 <balbinus> D|:The first DOAP document of KjetilK

15:18:50 <dc_rdfig> Titled item D.

15:18:51 <KjetilK> :-)

15:19:36 <balbinus> D:Created using balbinus's DOAP-a-matic

15:19:37 <dc_rdfig> Added comment D1.

15:19:55 <KjetilK> so, the bot is creating RDF for the links posted?

15:20:11 <balbinus> yep

15:20:14 <KjetilK> cool!

15:20:16 <crschmidt> KjetilK: http://rdfig.xmlhack.com

15:20:24 <balbinus> see it at http://rdfig.xmlhack.com

15:20:25 <balbinus> :)

15:20:34 * crschmidt winnar, and still champ-ene

15:20:58 <balbinus> champ-ene?

15:21:00 <KjetilK> veeeery cool!

15:21:22 <KjetilK> can I give the CSScheck link I posted a title?

15:21:38 <balbinus> D:Who is very proud to see his little DOAP light on (thanks crschmidt) in his Firefox :)

15:21:39 <dc_rdfig> Added comment D2.

15:21:41 <balbinus> yep

15:21:43 <KjetilK> uhm, trying...

15:21:48 <crschmidt> balbinus: In boxing matches, they do a little thing at the end for who won. The voice that sticks in my head (in english) says "Winner, and still Champion" in a way that sounds like "Champ-ene"

15:22:07 <balbinus> aok

15:22:08 <KjetilK> C|: A checker for Cascading Stylesheets

15:22:18 <crschmidt> KjetilK: flip : and |

15:22:24 <KjetilK> C:| A checker for Cascading Stylesheets

15:22:25 <dc_rdfig> Titled item C.

15:22:44 <KjetilK> neat

15:23:50 <balbinus> woahaargh! there's an enormous css bug in DOAP Viewer...

15:24:33 <KjetilK> Interestingly, DOAP has bumped into the same problem I'm addressing with long and short names

15:24:43 <KjetilK> whoops

15:25:06 * balbinus thinks he'd better link to mortenf's FOAF Explorer instead of trying to display them in DOAP Viewer

15:26:37 <KjetilK> DOAP has addressed the short/long names by separate elements, but I don't feel it is a good solution for me...

15:27:05 <KjetilK> balbinus: what's the CSS problem?

15:27:43 <balbinus> it makes tables' contents enormous sometimes without any reason

15:29:55 <balbinus> argh... that's horrible

15:30:49 <balbinus> oops

15:31:11 <balbinus> another problem: didn't strip slashes (now corrected)

15:35:50 * KjetilK hasn't seen the problem

15:36:18 <KjetilK> C: The CSS Checker is somewhat outdated

15:36:19 <dc_rdfig> Added comment C1.

15:37:27 * KjetilK finds there is probably something not right about his FOAF....

17:06:59 <Cloud^Home> Cloud^Home is now known as Cloud_

17:11:09 <Davey> Hey, anyone around? :)

17:11:29 <balbinus> yep

17:11:29 <Cloud_> lo

17:11:46 <Aredridel> Yup.

17:12:43 <Davey> OK, we're (PEAR, http://pear.php.net) going through the voting process for a port of RAP (RDF API for PHP)... but somebody is contesting the package name, RDF::RDF_($subpackage) (which would mean an new RDF category with several sub-packages in)

17:12:58 <Davey> they're saying it should go into Web Services, and I can't find a substantial reason why it shouldn't ;)

17:13:16 <Davey> but Web Services currently has like SOAP/UDDI stuff

17:13:19 <balbinus> because it's not?

17:13:42 <Davey> balbinus: RSS is...

17:13:49 <Davey> sorta

17:13:52 <balbinus> parsing RSS isn't

17:14:10 <Davey> I'm looking for a piece of text that says "RDF isn't a Web Service because... "

17:14:17 <Davey> that I can show these people

17:14:44 <balbinus> this chan log? ;)

17:15:11 <Davey> you can only use RDF in web services if you use it *with* SOAP or XML-RPC, therefore it, in itself is no more a web service than any other XML (I know about n3/ntriples, keeping it simple) vocabularly is on its own

17:15:37 <balbinus> RAP is for parsing RDF, isn't it?

17:15:46 <Davey> balbinus: yes

17:15:48 <Aredridel> Yeah... RDF:SOAP::HTML:XMLRPC

17:16:39 <balbinus> well, parsing is not a web service

17:16:44 <Davey> balbinus: thank you ;)

17:16:57 <balbinus> Davey: well, if i can help pear :)

17:17:24 <Davey> what are other semweb technologies? I can only think of OWL

17:17:32 <balbinus> and btw, RSS is in XML_RSS :)

17:17:39 <balbinus> RDFS ;)

17:17:41 <Aredridel> DAML+OIL

17:17:48 <Aredridel> FOAF

17:17:53 <balbinus> FOAF is RDF :)

17:17:54 <Davey> FOAF is RDF, no?

17:17:56 <crschmidt> FOAF isn't a technology, just RDF

17:18:01 <Aredridel> Ah, yes.

17:18:06 <Aredridel> Well.. what's the difference?

17:18:11 <DanC> FOAF isn't a technology? bizzare.

17:18:19 <balbinus> it isn't a tech in itself

17:18:23 <balbinus> it is included in RDF

17:18:24 <Aredridel> Heh.

17:18:27 <balbinus> a part of

17:18:31 <Davey> then neither is RDF, its just text ;)

17:18:32 <Aredridel> It is, though!

17:18:39 <balbinus> true

17:18:42 * crschmidt shrugs.

17:18:44 <balbinus> text isn't either

17:18:47 <balbinus> it's bits

17:18:50 <DanC> what's the argument that RDF should go under web services? (why are PHP packages arranged in a hierarchy at all, rather than a web?)

17:19:04 <Davey> balbinus: in the end, its all just natural ;)

17:19:15 <Davey> DanC: its PEAR, http://pear.php.net

17:19:16 <balbinus> bits aren't either. they are electrons. electrons aren't either. they're particles.

17:19:21 <balbinus> we are the void.

17:19:23 <balbinus> lol

17:19:50 <DanC> yes, why does PEAR have a hierarchy? (btw... all the folks who ask why w3.org uses dates in URIs... it's to avoid silly arguments like "doesn't RDF go under web services?")

17:20:28 <Aredridel> Hehe.

17:20:29 <Aredridel> Yes.

17:20:30 <Davey> DanC: its alot easier to find what you want this way ;)

17:20:30 <balbinus> Davey: well, RAP only deals w/ XML serialization of RDF, true? you could put it under "XML"

17:20:37 <Davey> DanC: also avoids class naming conflicts :D

17:20:38 <Aredridel> So there is ~one URI, and many categories.

17:20:47 <Davey> balbinus: nope, N3 and NTriples too

17:20:52 <balbinus> oops

17:20:55 <balbinus> didn't try :)

17:20:59 <Davey> has a full RDQL parser and DB datastore stuff too

17:21:16 <balbinus> oooh...

17:21:20 <Davey> DanC: we use the categories as namespaces, except PHP doesn't support real namespaces

17:21:25 <DanC> yes, Davey, categories are handy fro browsing, but they needn't impact the name/URI of a package.

17:21:32 <balbinus> just used the XML part for DOAP Viewer :)

17:22:14 <Davey> DanC: well, in this case it does ;)

17:22:23 <DanC> using categories as a namespace is antipattern #N in http://www.w3.org/Provider/Style/URI

17:22:29 <Davey> cbleh

17:22:31 <Davey> bleh

17:22:45 <Aredridel> Hehe.

17:23:03 <Davey> DanC: this way *works* ;)

17:23:21 <Davey> we just need to decide what the categories are ;)

17:23:45 <DanC> hmm... this discussion suggests to me that this way does not work.

17:24:28 * Aredridel uses dates personally, and very broad, flat category space professionally, with dates for things likely to be relatively static.

17:25:38 <Davey> DanC: pffft

17:25:54 <Davey> DanC: thats like saying that because people can't plug in their blender right, it doesn't work.

17:26:20 <DanC> not really.

17:26:32 <Davey> sure it is

17:26:37 <Davey> we're just deciding where the package fits.

17:26:41 <DanC> I've seen this "where does my thing go in the hierarchy?" discussion 10000000 times. it doesn't work;.l

17:26:50 <Aredridel> Sure it is. And if people can't plug in their blender right, plugs are too complex ;-)

17:27:01 <Davey> DanC: it does, it the hierarchy can change :)

17:27:21 * Davey grumbles about finding a decent WSDL validator

17:27:31 <Aredridel> Cool URIs don't change -- http://www.w3.org/Provider/Style/URI#xpointer(//h1)

17:28:04 <DanC> BLURB: Why does PEAR require each package to go in exactly one category?

17:28:09 <dc_rdfig> E: Why does PEAR require each package to go in exactly one category? from DanC

17:28:32 <Davey> DanC: thats for legacy code reasons ;)

17:28:35 <Davey> something we're working on ;)

17:28:48 <DanC> E:communities make this mistake over and over. The debian categories are pretty much invisible to the user, finally.

17:28:48 <dc_rdfig> Added comment E1.

17:28:50 <Davey> Aredridel: the URIs don't change ;)

17:28:58 <Aredridel> Hehe.

17:29:04 <Aredridel> That's a good start.

17:29:34 <DanC> E:cf *What to leave out... Subject* in [http://www.w3.org/Provider/Style/URI|Hypertext Style: Cool URIs don't change]

17:29:35 <dc_rdfig> Added comment E2.

17:29:35 <Davey> Aredridel: the URL is always pear.php.net/Category_Package and for the packages themselves pear.php.net/get/Category_Package-Version.tgz

17:29:58 <mattb> i was discussing perl CPAN with one of the maintainers on saturday evening

17:30:14 <mattb> he was saying that if he could go back and change one thing set in stone about it, he would not use hierarchy

17:30:18 <Aredridel> So if the category changes, the URL does.

17:30:19 <mattb> it just don't work

17:30:32 <Aredridel> Heh.

17:30:48 <DanC> E: i was discussing perl CPAN with one of the maintainers on saturday evening...

17:30:48 <DanC> <mattb> he was saying that if he could go back and change one thing set in stone about it, he would not use hierarchy

17:30:49 <dc_rdfig> Added comment E3.

17:31:16 <DanC> phpht.

17:31:18 <DanC> E3:""

17:31:19 <dc_rdfig> Deleted comment E3.

17:31:23 <Davey> Aredridel: but that'snot allowed ;)

17:31:26 <DanC> mattb, pls tell the chump

17:32:04 * DanC swats away a cookie from http://pear.php.net/

17:32:23 <DanC> E:see [http://pear.php.net/|PEAR - PHP Extension and Application Repository]

17:32:24 <dc_rdfig> Added comment E3.

17:32:44 <mattb> DanC: ok

17:32:48 <Davey> DanC: its been working for several years, and is the best solution for this.

17:33:10 <Davey> DanC: because the URIs correspond directly to the package and its description

17:33:19 <Aredridel> Davey: I wish the author luck when his project changes categories ;-)

17:33:27 <Davey> Aredridel: that doesn't happen ;)

17:33:30 <Aredridel> Heh.

17:33:33 <Davey> then it becomes a new package ;)

17:33:40 <DanC> you've tried all the other solutions, Davey? in particular, you've tried allowing packages to connect to as many categories as they like, ala freshmeat?

17:33:46 <Aredridel> I guess it is PHP. Packages become unmaintainable before they grow in scope ;-)

17:33:50 <DanC> logger, chump E:

17:33:50 <DanC> E:See [http://www.ilrt.bris.ac.uk/discovery/chatlogs/rdfig/2004-08-23#T17-33-50|discussion]

17:33:51 <dc_rdfig> Added comment E4.

17:33:51 <mattb> E:i was talking to one of the perl CPAN maintainers and he was saying that if he could go back and change one thing set in stone about it, he would not use hierarchy

17:33:52 <dc_rdfig> Added comment E5.

17:33:53 <Davey> DanC: we're working on that additionally

17:34:16 <Davey> Aredridel: well, heres the thing, we don't remove the original releases of a package which DID fit into the original category

17:35:07 <Davey> so if XML_Parser-0.1 turns into Text_Parser-0.2, XML_Parser-0.1 is still available at http://pear.php.net/XML_Parser though it may have a deprecated statement on it ;)

17:35:22 <DanC> E:see also [SoftwarePackaging|http://esw.w3.org/topic/SoftwarePackaging]

17:35:23 <dc_rdfig> Added comment E6.

17:36:02 <Aredridel> Yeh.

17:36:06 <Aredridel> It's just ugly to rename.

17:36:29 <Aredridel> "You need XML_Parser or Text_Parser or HTML_Parser or Generic_Parser" to run this package.

17:37:57 <DanC> E:hmm... software package names are somewhat like [http://c2.com/cgi/wiki?WikiWord|WikiWords] in that they're both user interface and URI.

17:37:58 <dc_rdfig> Added comment E7.

17:39:08 <DanC> E:i.e. the way one software package refers to another is by a name that's visible in the source code. renaming is somewhat costly, but it's not completely avoided.

17:39:09 <dc_rdfig> Added comment E8.

17:42:35 <Aredridel> E: I think URIs are always on the edge of user interface. Not their primary function, but it's tempting to use them that way.

17:42:35 <dc_rdfig> Added comment E9.

17:44:38 <ear1grey> Davey: is the Structures category not more appropriate for PEAR?

17:45:10 <ear1grey> Davey: (Given that there appears to be a graph structure package there already)

17:45:30 <DanC> E:yes, quite; it's a tension that's always there.

17:45:31 <dc_rdfig> Added comment E10.

17:45:47 * sanctius vous souhaite une très bonne nuit à tous et plein de beaux rêves dans les bras de Morphée ???

17:47:01 * DanC wanders off, greatful for the opportunity to rant a bit ;-)

17:47:12 * Aredridel smiles.

17:47:13 <Aredridel> Tata.

17:50:02 <danja> where do DB things go in PEAR?

17:50:40 * danja just noticed 'Structures' suggestion - not bad

17:51:21 <danja> is the idea to classify it properly, or to make it easy to find for people likely to be looking for it..?

17:51:37 <danja> if the latter, bung it in XML

18:16:48 * KjetilK just added a <rdfs:seeAlso rdf:resource= -link to his DOAF from the currentProject of his FOAF... Does that make sense?

18:17:01 <KjetilK> s/DOAF/DOAP/

19:14:42 <balbinus> KjetilK: you can also use <doap:Project> inside of a <foaf:currentProject> :)

19:18:07 <KjetilK> balbinus: Oh, thanks, checking out

19:19:01 <balbinus> if you speak a little french, you can have a look at http://doap-fr.org, the first post (at least the example)

19:20:23 <KjetilK> uhm, my french is pretty much limited to knowing that Norwegians generally pronounce "pommes frites" incorrectly... :-)

19:20:50 <balbinus> well, the example is in RDF, so it's international :)

19:21:20 <KjetilK> OK!

19:21:37 * KjetilK finds that he understands more french than he thought... :-)

19:21:41 <balbinus> lol

19:22:43 <KjetilK> I did have one year of french in school, though, so I should understand a bit, but it was far too little to get anywhere

19:23:04 <balbinus> :)

19:38:06 <KjetilK> balbinus: but I can't have both foaf:Project and doap:Project within the same currentProject?

19:38:16 <balbinus> no

19:38:20 <KjetilK> ok

19:38:21 <balbinus> why?

19:38:58 <mattmcc> A doap:Project is a subClassOf foaf:Project.

19:39:10 <balbinus> yep

19:39:43 <balbinus> i don't see why you would need both foaf: and doap:Project...

19:42:08 <balbinus> KjetilK: hey?

19:43:18 <KjetilK> nah, ok, not really...

19:43:35 <KjetilK> I just like to use dc-stuff when I can...

19:43:39 <balbinus> you can

19:43:47 <KjetilK> yup, and I did... :-)

19:43:50 <balbinus> :)

19:44:01 <KjetilK> but it is redundant...?

19:44:55 <balbinus> why?

19:44:55 <balbinus> nah

19:45:17 <KjetilK> in the sense that the contents of the dc:description element and doap:description element is likely to be identical

19:45:40 <balbinus> yep

19:45:59 <balbinus> i think that doap:desc is subPropertyOf dc:desc (or will become)

19:46:21 <KjetilK> ah, ok.

19:46:32 <KjetilK> how about doap:name?

19:46:42 <balbinus> err... wait a sec

19:47:30 <balbinus> subPropOf rdfs:label

19:47:53 <KjetilK> ok, good too

19:48:50 * KjetilK noticed that FOAF explorer didn't give a title for the project with only doap:name defined

19:52:55 <KjetilK> BTW, while I'm at it: Is there a way to say that a depiction (i.e picture of me) is the primary depiction that should be used if an app is to choose a single pic?

19:53:13 <balbinus> put it first :)

19:54:25 <KjetilK> hm, FOAF explorer doesn't think that...

19:54:37 <balbinus> well...

19:55:55 <KjetilK> FOAF explorer reverses the order and plink takes a random, it appears...

19:57:07 <balbinus> plink is random, that's true

19:57:20 <balbinus> well, put last for FOAF Ex :)

20:03:36 <KjetilK> hehe

20:07:24 <KjetilK> Hm, I see that '-s are escaped in the doap -o-matic gave me...

20:08:59 <balbinus> yep

20:09:07 <balbinus> i corrected this bug a few hours ago :)

20:09:12 <KjetilK> hehe

20:10:01 <KjetilK> now corrected in my doap too! :-)

20:10:15 <balbinus> :)

21:53:01 <mortenf> KjetilK, re "primary depiction", use foaf:img

21:54:19 <KjetilK> mortenf, thanks a lot!

21:54:51 <KjetilK> BTW, mortenf, I pretty much copied your FOAF to form the basis for my own...

21:55:05 <mortenf> good idea ;)

21:55:08 <KjetilK> yep!

21:55:23 <KjetilK> but I noticed you claim to read swahili :-)

21:56:12 <mortenf> erh, no

21:57:20 <KjetilK> yep, I know you read no too, but... :-)

21:57:32 <KjetilK> sv is swedish...

21:58:48 <mortenf> ah, right, a typo (wonder why i didn't catch that before), thanks!

21:59:01 <KjetilK> :-)

22:01:32 <KjetilK> it is very natural to think sw(edish), but ISO639 (IIRC) uses codes close to native