Friend of a Friend (FOAF) IRC Chat Logs for 2003-12-23

01:02:41 <D[a]vey> D[a]vey is now known as Davey

01:42:54 <danbri> bryce, you _can_ say that, but I guess I'd warn that people are more widely known by indirect identifiers, eg. their homepages, weblogs, mailboxes, etc.

01:43:13 <danbri> So it'd work fine for merging with other data you also manage, and other systems which know the same URIs you're using for them...

01:45:01 <bryce> well, i *am* managing them, but my compromise is to just make up a mbox value for them... later i can equate my value with some better identifier.

01:45:29 <bryce> <foaf:depicts>

01:45:29 <bryce> <foaf:Person><foaf:mbox>Jack@NOSPAM-brycebenton.us</foaf:mbox></foaf:Person>

01:45:29 <bryce> </foaf:depicts>

01:47:20 <bryce> i made a template in Movable Type that loops through comma delimited values and plugs in wordnet and people depictions.

01:48:42 <bryce> i'd be interested to know if my template is sound. here's an example outputted file: http://tenyearsof.us/metadata/20031207

01:48:51 <danbri> yup, makes sense if you manage the whole thing. One tweak re syntax: RDF/XML syntax for properties whose values are URIs is different. We use mailto: URIs to identify mailbox values of foaf:mbox, so...

01:48:56 <danbri> <bryce> <foaf:depicts>

01:48:56 <danbri> <bryce> <foaf:Person><foaf:mbox>Jack@NOSPAM-brycebenton.us</foaf:mbox></foaf:Person>

01:48:56 <danbri> <bryce> </foaf:depicts>

01:48:58 <danbri> becomes

01:49:19 <danbri> <bryce> <foaf:depicts>

01:49:19 <danbri> <bryce> <foaf:Person><foaf:mbox rdf:resource="mailto:Jack@NOSPAM-brycebenton.us"/></foaf:Person>

01:49:19 <danbri> <bryce> </foaf:depicts>

01:50:04 <bryce> ok thanks. i forgot. i'll update the template and rebuild the files.

01:50:40 <danbri> I don't recal a foaf:annotates property... hmm libby uses or used a something-or-other:annotates property at one point

01:51:15 <danbri> where you have <foaf:depicts rdf:type="http://xmlns.com/wordnet/1.6/Face" />

01:51:25 <danbri> ...you are saying that the image depicts the general class of things 'Face'

01:51:42 <danbri> MattB used to use this style, I think he migrated away though. We generally instead say:

01:51:55 <danbri> <foaf:depicts><wn:Face/></foaf:depicts>

01:52:03 <danbri> ...ie. "the image depicts some particular face"

01:52:12 <danbri> ...this extends more gracefully, since you can say:

01:52:20 <danbri> <foaf:depicts><wn:Sheep/></foaf:depicts>

01:52:22 <danbri> or

01:52:40 <danbri> <foaf:depicts><wn:Sheep foaf:nick="larry" /></foaf:depicts>

01:52:50 <danbri> ...etc, without changing the basic style of representation

01:52:59 <bryce> i really appreciate the style tips. that's what i need. i get it. cool.

01:53:03 <danbri> ok cool

01:53:29 <danbri> can you read the rdf/xml syntax? the syntax change I showed corresponds to inserting an intermediate blank node in the pattern you originally used

01:53:41 <danbri> the RDF validator graphing option can be really handy for learning the syntax

01:54:22 <bryce> i understand the blank node. but now i understand it a little better. :)

01:55:22 * bryce learned about the intermediate blank node one day while lurking around here.

01:55:55 <danbri> glad we proved useful!

01:56:29 <danbri> what (who...) is tenyearsof.us btw? (nosy...) your family/friends/etc?

01:56:45 <bryce> totally. i didn't know anything about rdf 2 weeks ago.

01:56:52 <danbri> wow

01:56:54 <danbri> you learn fast

01:57:28 <danbri> rdf has a reputation for being hard, apparently

01:57:37 <bryce> it's a way of documenting our (my wife and i and extensions) life for the next 10 years.

01:57:51 <bryce> maybe you'll make it in there...

01:57:51 * danbri finds xslt completely unlearnably hard, but there you go...

01:58:00 <danbri> :)

01:58:30 <bryce> i read the xml bible, but i never did anything with xml until now.

01:59:02 <bryce> except for modify my blog's feed.

01:59:07 <danbri> I like the one photo a day thing...

01:59:44 <bryce> yeah, it forces me to record at least something from (almost) every day.

01:59:55 <danbri> I take a lot of photos, but don't manage them well, and have run out of server space, need to buy a new hard drive over xmas...

02:00:23 <bryce> i understand.

02:01:37 <bryce> after i add the metadata to all the existing photos, the next step will be to offer interesting queries and navigation.

02:02:25 <bryce> right now, it's only chronological... although the simplicity is also nice.

02:03:48 <danbri> hey, re http://tenyearsof.us/2003/12/19/ ... congratulations :)

02:04:12 <danbri> ok i should go sleep... g'nite...

02:04:14 <bryce> that's cool, huh? the site was one reason i got it.

02:04:19 <bryce> nighty

02:04:20 <danbri> nice!

02:04:23 <danbri> take care...

02:14:09 <bryce> danbri (possibly "future log-reading danbri")... you were right. there is no foaf:annotates. but libby's javascript thingy has it. hmmm. not sure what to do.

02:15:18 <mattmcc> :annotates is an Annotea property..

02:15:24 <Davey> I was thinking that

02:15:50 <Davey> I thought that much earlier but since danbri didn't point that out, assumed he was right. heh

02:16:11 <bryce> ic. so is there a namespace or something for that (you lurkers) ?

02:16:36 <bryce> ;)

02:17:10 <Davey> hold on :)

02:17:19 * Davey is writing an Annotea client, should know this

02:18:03 <Davey> xmlns:a="http://www.w3.org/2000/10/annotation-ns#"

02:18:18 <bryce> thanks!

02:18:44 <Davey> the "spec" can be found here: http://www.w3.org/2001/Annotea/User/Protocol.html

02:18:48 <Davey> if you're interested :)

02:19:09 <bryce> that was my next question. thanks.

02:19:20 <Davey> np :)

05:41:56 <Davey> Davey is now known as D[a]vey

10:45:47 <libby> yeah my foaf:annotates and foaf:creationEvent are made-up and I should move them

11:58:11 * libby waves at jibberjim from the logs....

11:58:30 <dajobe> haha

11:58:56 <dajobe> maybe I should do a log readers top 10 for the holidays

12:24:57 <dajobe> so for November 2003 the top five are, in reverse order...

12:25:10 <dajobe> #5 somebody in San Francisco (pacbell.net)

12:25:17 <dajobe> #4 I think ldodds

12:25:39 <dajobe> #3 libbydanbrishellacvisitors

12:25:47 <dajobe> #2 me! I was probably testing things, ahem

12:25:57 <dajobe> #1 a ukolner (I can guess who)

12:26:06 <dajobe> #0 mystery person at uk.uu.net

12:26:09 <dajobe> and to prove I can't count:

12:26:24 <dajobe> the largest log reader for Nov 2003 by a huge margin is ...

12:26:28 <dajobe> mortenf

12:26:51 * dajobe waves to mortenf in the logs, congrats

12:27:01 <libby> heh

12:27:52 <libby> what about searches? top 10 searches?

12:28:02 <libby> this is #foaf right?

12:28:12 <dajobe> I'll look, but it might be r00d

12:28:30 <libby> heheh

12:29:25 <dajobe> ok the top five for some random period are

12:29:31 <dajobe> #5 error

12:29:36 <dajobe> #4 nodeid id

12:29:38 <dajobe> #3 foafer

12:29:40 <dajobe> #2 inkel

12:29:43 <dajobe> #1 spanish

12:29:45 <dajobe> hey, go figure

12:30:01 <dajobe> (they are lowercased and munged a bit)

12:30:10 <libby> weird

12:30:56 <libby> what about log reader for dec?

12:31:50 <dajobe> it's churning the ips to hostnames right now

12:32:25 <Dorward> Top five searches for arriving at my site are: (5) mouseover menus, linux intellimouse, microsoft intellimouse linux, intellimouse explorer linux, and the top search is: intellimouse linux (for this month so far) ... and the main focus of my site is the web, not OSes.

12:33:43 <dajobe> I guess I could do some #foaf chat stats like I do for #rdfig; http://www.ilrt.bris.ac.uk/discovery/chatlogs/rdfig/stats.html

12:36:52 <libby> heh, that's neat

12:36:57 <dajobe> there we go http://www.ilrt.bris.ac.uk/discovery/chatlogs/foaf/stats.html

12:38:56 <libby> cool.

12:39:15 <libby> on the by nick bit, does that include joining and leaving?

12:39:26 * dajobe fixes the imagaes

12:39:37 <dajobe> hmm, I don't know

12:39:43 <libby> damn, jibbs is just ahead of me

12:41:01 <JimH> bizarre stuff, I've said more than some far more illustrious people, and I thought I mostly lurked in here

12:41:35 <dajobe> that's automated now, it'll be updated daily

12:43:41 <dajobe> the stats for dec are nearly churned

12:44:40 <DynaD> the trouble with stats like that is that they make people like me want to talk a lot

12:49:09 <DynaD> pah! who needs friendsreunited when you have... http://beta.plink.org/pivot.php?q=http%3A%2F%2Filrt.org%2F

12:51:05 <dajobe> ok, the scores for dec are in

12:51:11 <dajobe> in reverse order

12:51:15 <dajobe> and this time I can count

12:51:19 <dajobe> #5 mattb

12:51:25 <dajobe> #4 danbrishellaclibby

12:51:33 <dajobe> #3 ukoln person

12:51:37 <dajobe> #2 mortenf!

12:51:46 <dajobe> #1 mystery person at uk.uu.net

12:51:53 <JimJibber> I win, I win! probably...

12:52:25 <dajobe> the #1 has 2x the #2 slot

12:53:00 * JimJibber does have a script...

12:53:19 <dajobe> I suspect mortenf does also

12:55:42 <JimJibber> hmm, although maybe it's not me, as I can't get rDNS info for 159.... that is me.

13:40:36 <libby> yay go jimjibber!

13:43:06 <libby> have fun xmas all

15:15:04 <bryce> has there been any discussion about identifying groups of people... eg. a rock band?

15:15:35 <Dorward> I think there is foaf:group

15:15:42 <bryce> ah.

15:16:17 <Dorward>http://xmlns.com/foaf/0.1/#term_Group

15:17:53 <bryce> thanks. but then the question is how to uniquely identify the group. just a string description?

15:19:32 <bryce> nevermind. the link has the info.

15:19:41 <JimH> you could use a dc:title or dc:description for the group

15:20:35 <JimH> I've used foaf:group for a multi contributor weblog before so the foaf:group has a foaf:weblog

15:20:36 <bryce> the example uses foaf:name and then allows to list foaf:members

15:21:36 <JimH> yeah foaf:name is probably better than dc:title or dc:description

15:41:54 <bryce> perhaps the official homepage for a band would make a good identifier.

15:42:51 * danbri thinks we probably need foaf:groupWeblog for that, ie. a relationship between a Group and a Document

15:43:01 <danbri> foaf:weblog is a relationship between an agent and a document

15:44:34 <JimH> I'll blame JibberJim for my bad use of foaf:group - it seemed far better than a load of people each claiming the weblog though

15:45:27 <danbri> yep. it was a not-quite-cornercase to address 'later' when we added foaf:weblog

15:46:03 <bryce> ok, but in the cases that a group does have a homepage, it seems like that would be the way to go. in terms of a group, the distinction between homepage and weblog seems more clear.

15:47:07 <bryce> the homepage may not be their voice, but it is a representation of them.

15:57:40 <oPless> [OT] (#freenode) Jibbler: yeah, but at least there would be more distros using grsec and/or openwall :)

15:58:50 <JimH> danbri: I've just skimmed the spec and http://xmlns.com/foaf/0.1/#term_Group says "The foaf:Group class represents a collection of individual agents (and may itself play the role of a foaf:Agent, ie. something that can perform actions)."

16:00:43 <JimH> is it intended that foaf:Group becomes something other than an agent at some point?

16:03:25 <danbri> it represents a collection of agents. only agents make up a group, usually people.

16:03:48 <danbri> the claim that it is itself an Agent is a bit fancy/experimental.

16:04:04 <danbri> sorry cant hack on this right now, busy...

16:05:07 <JimH> ok, np

16:09:30 <JimH> a few thoughts from a hackers perspective is that making a group and agent is a nice simple hack which instantly gives the group all the capabilites of any other agent without having to have a seperate set of stuff like groupWeblog, groupHomepage, groupThis, GroupThat, and groupTheOther

16:10:44 <JimH> I guess the counter argument is that groupWeblog (et al) are clearly understandable for a simple user and should prevent more bad Foaf in the wild

16:12:20 <bryce> why would it matter if someone used weblog vs. groupWeblog for a group?

16:12:33 * bryce is a newbie.

16:13:14 <JimH> I think Dan's not convinced that making a Group have all the capabilities of an Agent is the right way to go

16:13:37 * JimH isn't a newbie, but that doesn't mean he's right either

16:14:40 <bryce> can you provide a reference for what *is* an agent? i thought an agent was like an evolved spider.

16:15:03 <JimH> yep - http://xmlns.com/foaf/0.1/#term_Agent

16:15:15 <bryce> no links to whatisthematrix, please. :)

16:15:42 <JimH> ok, you know what a foaf:Person is?

16:15:49 <bryce> yes.

16:16:08 <JimH> so foaf:Person is a class representing a person

16:16:27 <bryce> The foaf:Agent class is useful in a few places in FOAF where foaf:Person would have been overly specific. For example, the IM chat ID properties such as jabberID are typically associated with people, but sometimes belong to software bots.

16:16:35 <JimH> that's it

16:16:45 <bryce> ok.

16:16:53 <JimH> it's just that step vaguer than being necessarily human

16:17:20 <bryce> what about "people" that can no longer do things... eg dead people?

16:17:49 <JimH> I guess they're still foaf:Persons

16:18:19 <bryce> so person or agent doesn't imply currently active.

16:18:41 <JimH> probably not

16:20:06 <JimH> "We don't nitpic about whether they're alive, dead, real, or imaginary." it says about foaf:Person

16:20:24 <bryce> ah. good quote.

16:20:44 <bryce> maybe i should rtfs.

16:22:09 <JimH> it's not a bad spec, it does cover most of the corner cases - I should also rtfs more often :-)

16:22:55 <bryce> yeah, actually the documentation is pretty good. props to the authors.

17:16:31 <JimH> Jim Ley's been following the Group is/isn't an Agent stuff - http://jibbering.com/blog/.1?1072199408376

17:17:07 <JimH> Foaf as an Xmas pressie list is rather amusing, but damn useful

17:18:58 <bryce> hehe i doubt i'll be receiving a gift from both my wife *and* a girlfriend.

17:20:03 <deltab> foaf:Group is listed as a subclass of foaf:Agent

17:22:55 <bryce> well, i think the point was that danbri said that might be fanciful, but jibbering is saying that it is indeed useful.