This is a public chat log generated from the #semsol IRC channel.
14:04:24
bengee: in ARC, which is usually more performant, a CONSTRUCT or a DESCRIBE?
14:05:34
they both re-use SELECT, but CONSTRUCT usually needs less queries
14:11:08
cheers
14:11:19
np
14:27:32
danbri: was it you that asked about the only 29% of mf parse into vcard statistic I quoted at semanticcamp?
14:27:47
http://www.nabble.com/comment-on-using-RDFa-to10838379.html#a10838379
14:28:17
yes, though i didnt know you were you ;)
14:28:34
names and nicks and faces - thats why we need sparqlbot!
14:28:42
sparqlbot++
14:28:45
I can give one more advice: try to think of various ways people will
14:28:46
abuse RDFa markup... and test the tools against these test cases. I'm
14:28:47
saying this because I've recently looked at some microformat data
14:28:48
(vCard) and my quick statistics show that even after tidying only about
14:28:50
29% of manually authored hCard markup in HTML pages results in a correct
14:28:51
vCard.
14:29:01
(that's what the guy said - from yahoo research barcelona)
14:29:17
hard to know how solid the research was
14:29:58
29% is pretty poor though
14:31:21
i dont see the 29% thing in that post
14:31:30
ah
14:32:48
i'll ping Peter
14:33:39
danbri: be interesting to hear more about it :)
14:34:04
contrary to what jeremy keith said, i don't think it's because people will always get things wrong
14:34:15
i think it's because it's hard to get it right
14:34:17
if 29% of mf is good, thats no cause for rdf gloating, cos it means about 15% of rdfa will be good ;)
14:34:34
depends
14:34:45
rdfa has the advantage of being validatable
14:34:51
having a lightweight pure js parser could help
14:34:55
and you can check you get the triples you expect
14:34:56
people can test easier
14:34:57
yeah
14:35:03
there is one for rdfa
14:35:07
ben adida's
14:35:25
yup
14:35:27
v handy
14:35:56
from personal experience, it could still be made easier to debug semantic html though
14:36:20
you know, you write, then you have to go parse it to check it's ok
14:36:22
tedious
14:36:33
better if there was some way to make it more immediate and visual
14:42:02
ok i'm adding peter in skype, will bug him asap
14:42:34
one way would be to augment the wysiwyg html editors in Wordpress etc
14:42:41
to have ui for common constructs
14:44:12
danbri: there was a chap at semantic camp wanting to work on that
14:44:18
he was from ....
14:44:21
kwijibo googles
14:44:33
for rdfa, or in general?
14:44:41
yes
14:44:44
well
14:44:48
yes which? :)
14:44:50
he hadn't decided on rdfa
14:44:53
ah ok
14:45:02
he was looking at different options
14:45:38
but a semantic tinyCME wordpress plugin