IRC log of dig on 2010-05-01

Timestamps are in UTC.

00:39:47 [vinirusso]
vinirusso (~Adium@189.62.70.179) has joined #dig
00:59:36 [vinirusso]
vinirusso has left #dig
01:32:11 [kkw]
kkw (~kkw@cpe-98-14-185-162.nyc.res.rr.com) has joined #dig
01:35:14 [kkw]
kkw has left #dig
01:51:18 [ruthdhan]
ruthdhan has quit (Quit: This computer has gone to sleep)
02:36:09 [charles2]
charles2 has quit (Quit: :))
03:07:02 [timbl]
timbl (~timbl@pool-96-237-236-72.bstnma.fios.verizon.net) has joined #dig
07:24:19 [oshani]
oshani (~oshani@cpe-069-134-178-080.nc.res.rr.com) has joined #dig
07:32:22 [oshani]
oshani has quit (Quit: oshani)
12:41:39 [charles2]
charles2 (~charles2@dhcp-18-111-98-139.dyn.mit.edu) has joined #dig
12:52:15 [oshani]
oshani (~oshani@cpe-069-134-178-080.nc.res.rr.com) has joined #dig
12:53:25 [charles2]
timbl: if i were implementing the status update functionality as on facebook in a tabulator pane, do you think i should use foaf:status? or should I use things from sioc (like sioc:Microblog)?
13:07:50 [oshani]
charles2, did you look at the facebook opengraph schema? http://opengraphprotocol.org/schema/
13:08:28 [charles2]
oshani: I haven't, but i'll check it out! :)
13:10:57 [kennyluck]
I think it's not really relevant, though.
13:12:13 [kennyluck]
charles2, does sioc has microblog-related vocabulary? In that case, I would recommend sioc. I never noticed the existence of foaf:status.
13:12:32 [charles2]
kennyluck: foaf:status is new i think
13:12:36 [oshani]
yeah, opengraph doesn't seem to have anything to model the status
13:12:51 [charles2]
i built the microblog pane using sioc
13:13:15 [kennyluck]
Anyway, you can always use both :)
13:13:20 [oshani]
btw, kennyluck, I met Albert at WWW2010.. You should meet him in Japan :)
13:13:48 [charles2]
do either of you happen to know of a good way to represent a collection of photos?
13:14:18 [kennyluck]
Yup, but he asked for a formal invitation (so that he can take a business trip), I am afraid that I might not be able to do that.
13:17:21 [oshani]
oshani has quit (Quit: oshani)
13:18:45 [kennyluck]
charles2, you might want to post this question to #swig
13:18:55 [charles2]
will do
13:19:16 [kennyluck]
The community built so many ontologies...
13:20:42 [charles2]
there's bound to be one I can use :)
13:21:34 [timbl]
The modeling in the opengraphprotocol scema is rather forced as there was a reuirement that everything should be a property of the web page, when in fact it is info often about the subjcet of the page (like a movie) and also there was a requiremnt hat all should be literal strings when in fact smetimes they are URIs of reseources
13:22:14 [timbl]
Charles2, what does identi.ca use?
13:22:43 [charles2]
timbl: for status?
13:23:09 [kennyluck]
yeah, timbl, I don't like metadata. I like raw data :)
13:23:21 [timbl]
for status
13:24:01 [timbl]
I wouldn't sepnd too much time worrying. Pick one. Afterwards one can (a) change later and (b) publish triples which say they are the same
13:24:11 [charles2]
it looks like they are using sioc:creator
13:24:21 [kennyluck]
Huh? status?
13:24:26 [charles2]
er has creator
13:24:47 [charles2]
with microblogpost
13:24:55 [kennyluck]
oh I see. s a Microblog; has_creator ?
13:25:11 [charles2]
looks like it
13:25:11 [kennyluck]
hmmm 4 triples.
13:25:36 [charles2]
I wonder if the microblog pane would understand identica…
13:25:50 [charles2]
i think it would need a little tweaking
13:25:51 [kennyluck]
I would go for sioc, though. As you said, foaf:status is new.
13:26:59 [charles2]
alright
13:27:36 [kennyluck]
Another reason, foaf:status seems to be not logical to me.
13:27:58 [kennyluck]
Microblog is a piece of message created by a person, that makes sense.
13:28:29 [kennyluck]
foaf:status depends too much on time, and hence not stable, like foaf:age.
13:28:58 [kennyluck]
In the UI you might want to do a query that looks for the most recent Microblog, I guess.
13:30:13 [charles2]
now the question is, if a person has multiple microblogs, how do I decide which gets put on the ui.
13:30:24 [charles2]
I've been pondering over how to make that relationship work
13:30:24 [kennyluck]
dc:date
13:30:37 [kennyluck]
the most recent Microblog :)
13:30:58 [charles2]
well i want to make it like facebook inthat it shows a log of them
13:31:25 [kennyluck]
Oh, OK. That sounds a lot cooler.
13:32:01 [kennyluck]
Then foaf:status obiviously won't work because it has no information about the time when the message is updated.
13:32:15 [charles2]
yup
13:32:21 [charles2]
that was a problem I was having with it
13:34:33 [kennyluck]
Oh, by the way, you might want to check the semantictweet thing, charles2
13:34:44 [kennyluck]
Hopefully your UI can be used to display that.
13:35:33 [charles2]
semantictweet? oh this is interesting
13:36:09 [kennyluck]
They made a twiiter API -> Linked Data adapter, it seems.
13:39:18 [kennyluck]
http://semantictweet.com/ (in case you haven't found it)
13:42:32 [charles2]
I think what I'm going to do is link to my identica (which i really should update more often) and see if i can get that to show up in the ui and then I'll see if i can get semantic tweet tweets to show up. that would be pretty awesome
13:43:05 [kennyluck]
Yeah.
13:43:24 [kennyluck]
I don't even have a identica account, I think. :[
13:48:45 [charles2]
oops i'm using holdsAccount which is apparently archaic. Will fix that
13:49:19 [kennyluck]
:)
13:50:24 [kennyluck]
danbri shouldn't change any URI, though. What can be changed is rdfs:label.
13:51:47 [charles2]
huh, is it normal for tabulator to not notice changes i've made to my foaf?
13:52:03 [kennyluck]
Oops, the refresh thing, huh.
13:52:16 [kennyluck]
You might want to click the green bubble.
13:52:40 [kennyluck]
to refresh (not sure what works now)
13:52:49 [charles2]
hm
13:53:25 [charles2]
it probably also doesnt help that i branched jambo's refactor branch which nobody's been using for a while
13:53:25 [kennyluck]
anyways, resort to restarting Firefox if necessary....
13:53:46 [kennyluck]
By the way, where's jambo's refactor branch?
13:54:27 [kennyluck]
Is is on http://dig.csail.mit.edu/hg/ ?
13:54:36 [charles2]
its on the mercurial repo
13:55:08 [kennyluck]
OK, not the main trunk, you mean?
13:55:54 [charles2]
http://dig.csail.mit.edu/hg/tabulator/
13:56:43 [kennyluck]
Uh... why did you say "nobody's been using for a while"?
13:56:53 [kennyluck]
Do people still commit to the svn version?
13:57:05 [charles2]
I don't think so
13:57:12 [charles2]
but everyones on the 'default' branch
13:57:22 [kennyluck]
OK, got it.
13:57:33 [charles2]
jambo was doing some cleanup. the microblog pane lives in the refactor branch.
13:58:07 [charles2]
(the functional one, despite its issues...)
13:58:21 [kennyluck]
OK
13:58:52 [charles2]
so this is strange though, tabulator doesnt seem to be updating even with the restart
13:59:19 [kennyluck]
Hmm...caching problem? I never understand this.
13:59:27 [charles2]
oh wait here we go
13:59:41 [charles2]
i pressed the green button and then restarted
14:00:05 [kennyluck]
lol, that's very complicated
14:02:35 [charles2]
the tabulator and i seem to have occasional disagreements. like right now it doesn't seem to be able to dereference the identica account that i stuck into my foaf.
14:03:42 [kennyluck]
Hmm.. probably owl:sameAs problem? Or it's just about content negotiation not working right?
14:04:12 [charles2]
it might actually be identica? it looks like when i try to hit it with the browser i'm redirected to a webpage
14:04:39 [kennyluck]
Sounds like a content negotiation problem.
14:06:00 [charles2]
hm. what would i do about that?
14:06:41 [kennyluck]
test it with curl?
14:06:55 [kennyluck]
Hmm..is the tool -> data brower -> make tabulator request RDF on?
14:09:13 [charles2]
yup
14:09:21 [charles2]
i'll try curl
14:12:36 [charles2]
well that's an interesting response. it says doctype is xhtml and follows that with an link to identi.ca/cm2
14:13:19 [kennyluck]
Huh? What's the URI you are trying with?
14:14:47 [charles2]
http://identi.ca/user/57496#acct
14:17:03 [kennyluck]
curl http://identi.ca/user/57496 -I -H 'Accept:application/rdf+xml'
14:17:16 [kennyluck]
I got http://identi.ca/cm2/foaf
14:18:18 [kennyluck]
curl http://identi.ca/user/57496 -I -H 'Accept: application/rdf+xml, text/html'
14:18:28 [kennyluck]
got http://identi.ca/cm2
14:18:54 [kennyluck]
That is, the service prefers HTML.
14:19:20 [kennyluck]
which is correct I think, but a trouble for tabulator.
14:22:00 [charles2]
that's unfortunate :(
14:23:51 [kennyluck]
very unfortunate :)
14:24:04 [kennyluck]
curl http://identi.ca/user/57496 -I -H 'Accept: application/rdf+xml, text/html; q=0.5'
14:24:15 [kennyluck]
got http://identi.ca/cm2/foaf
14:24:41 [kennyluck]
so you might want to change the accept header with about:config
14:25:20 [kennyluck]
I think timbl had q=0.3 for HTML content type before.
14:27:16 [charles2]
hmm
14:29:40 [charles2]
oh also, I have a question: when i use owl:sameAs to link myself to my identica foaf, I get double of everything. do you know what causes that?
14:29:54 [kennyluck]
sigh
14:30:32 [kennyluck]
Hmm... you might want to add some kb.canon calls to your code.
14:30:56 [kennyluck]
For example kb.any(s, p, o)
14:31:23 [kennyluck]
-> kb.any(kb.canon(s), p, o)
14:31:42 [charles2]
aha
14:31:47 [charles2]
I'll try that right now
14:32:09 [kennyluck]
I hate sameAs's just because of this reason.
14:33:14 [charles2]
haha. yeah… I was trying to figure out why when i tried to use my pane on timbl's foaf everything was duplicated
14:34:22 [charles2]
that did the trick
14:34:32 [kennyluck]
Hmm.. You need to assume, for example, the kb has more information than you think.
14:34:59 [kennyluck]
duplicate triples, and so on. All the dirty thing...
14:35:38 [kennyluck]
"that did the trick", you mean kb.canon?
14:35:46 [kennyluck]
charles2?
14:35:55 [charles2]
yes
14:36:02 [charles2]
it got rid of the duplicates
14:36:05 [kennyluck]
Hurray...
14:48:10 [oshani]
oshani (~oshani@cpe-069-134-178-080.nc.res.rr.com) has joined #dig
14:56:23 [oshani]
oshani has quit (Ping timeout: 246 seconds)
15:24:49 [charles2]
I don't think semantictweets does this right
15:24:56 [charles2]
i follow people on twitter that i do not know
15:29:31 [charles2]
or perhaps i do know them, but that's not the same as friends
15:33:16 [kennyluck]
Yeah, I agree.
15:33:31 [kennyluck]
They kind of choose the wrong vocab.
15:33:59 [kennyluck]
On the other hand, there is not yet a standard vocab for followers.
15:34:19 [charles2]
identica uses sioc:follows
15:34:32 [charles2]
i used that for the microblog pane as well
15:34:34 [kennyluck]
Oh, so it exists.
15:34:54 [charles2]
somehow I was able to see the posts i made on identica
15:34:56 [charles2]
and now i can
15:34:57 [charles2]
t
15:34:58 [kennyluck]
Hmm..sorry I didn't know that, one of us should report about this to semantictweets people.
15:35:27 [kennyluck]
Anyway, I don't think that thing will be well maintained, anyway.
15:35:37 [charles2]
but i guess the idea of the semantictweets thing is to bootstrap a foaf out of twitter relationships
15:35:37 [kennyluck]
All these third party scraping things...
15:35:46 [kennyluck]
Yup,
15:42:54 [charles2]
OH!
15:43:40 [kennyluck]
OH what, charles2?
15:44:08 [charles2]
so i figured out what was going on with the posts. I looked in the header at http://identi.ca/cm2 and its got an alternate link to the rss feed
15:44:29 [charles2]
which describes all of the posts
15:44:36 [charles2]
so i needed the rss
15:44:42 [kennyluck]
Ah, right.
15:45:05 [kennyluck]
But it's not using sioc or something, I suppose? Or it is ?
15:45:26 [charles2]
its got sioc in there
15:45:36 [kennyluck]
That's very cool!
15:45:46 [charles2]
i never would have though
15:45:59 [charles2]
s/though/thought/
15:46:14 [kennyluck]
SIOC in RSS, you mean?
15:46:56 [charles2]
yeah, it didnt occur to me that someone would put it in there
15:47:43 [kennyluck]
wow, that's totally cool.
15:48:08 [kennyluck]
timbl, do you know who's the developer doing this (ref. SIOC in RSS)
15:54:06 [timbl]
No, I don't
15:54:33 [timbl]
By the way .. getting dupliacets of everything ... usng cnon() should NOT be necessary on kb.each(a,p,o)
15:59:30 [kennyluck]
timbl, have you played with relfinder already? I think it's very interesting, and it's a very general LOD application, like the tabulator. It relies on SPARQL endpoints, however.
15:59:54 [kennyluck]
Just wonder how relfinder can help government discover more hidden facts.
16:05:16 [oshani]
oshani (~oshani@cpe-069-134-178-080.nc.res.rr.com) has joined #dig
16:07:02 [oshani]
oshani has quit (Client Quit)
16:09:11 [kennyluck]
Hmm... is WWW2010 all over?
16:16:23 [kennyluck]
Oh, by the way, timbl, are you going to join the discussion on the potential W3C forum system? http://www.w3.org/2008/OUCD/wiki/Main_Page
16:21:43 [timbl]
WWW2010 is all over .. had 5 sem web people on the same plane back to W3C/DIG
16:22:41 [kennyluck]
:)
16:26:43 [timbl]
ian, mat, me,sandro, eripc
16:28:52 [kennyluck]
Hmm.. sadly sem web people are still few.
16:36:54 [kennyluck]
Based on some rough estimations, there are 22 sem web people in Japan.
16:37:14 [kennyluck]
And based on my estimation, there are 5 sem web people in Taiwan.
17:08:22 [vinirusso]
vinirusso (~Adium@189.62.70.179) has joined #dig
17:13:00 [vinirusso]
vinirusso has quit (Client Quit)
17:13:40 [vinirusso]
vinirusso (~Adium@189.62.70.179) has joined #dig
18:17:41 [vinirusso]
vinirusso has left #dig
18:29:48 [charles2]
got it!
19:16:15 [lkagal]
lkagal (~lkagal@pool-98-110-160-220.bstnma.fios.verizon.net) has joined #dig
20:59:50 [oshani]
oshani (~oshani@31-35-247.wireless.csail.mit.edu) has joined #dig
22:24:53 [vinirusso]
vinirusso (~Adium@189.62.51.2) has joined #dig
22:25:07 [vinirusso]
vinirusso has left #dig
23:23:01 [charles2]
charles2 has quit (Quit: charles2)