This is a public chat log generated from the #semsol IRC channel.
13:05:30
bengee dreamt of a solution to ARC's smusher issues, and that really seems to improve things
13:06:06
crazy synapses
13:34:26
you designed it in your sleep?
13:34:33
:)
13:34:52
hey bengee, do you have any figures or estimates re ARC scalability?
13:34:57
found a bug, at least
13:36:29
current version isn't optimized yet. pumped in a few Mtriples, but inserts got really slow
13:37:00
it depends quite a bit on mysql settings (available memory for indexes etc)
13:38:23
I'm working on code that'll split the triple tables. This should improve scalability and performance, early tests are very encouraging
13:40:07
I'll do scalability tests soon
16:09:42
I wrote a function to parse del.icio.us RSS feeds, but wondered if I was doing it the most straightforward way possible...
16:09:46
so posted it here: http://refactormycode.com/codes/221-parse-del-icio-us-rss-feed
16:10:44
I had to do a lot of "pick the first element from the array" for moving through the graph, wasn't sure if that was necessary
16:13:11
looks good - though maybe finding the Sequence is unnecessary if you are again checking that rdf:type is rss:item
16:14:02
bengee, have you got any grddl code for arc?
16:14:14
I found some W3C group pages have foaf in them, but grddl only
16:18:01
kwijibo: true, maybe there's no need to check for rss:item if there can only be "item"s in the "items" list
16:18:23
i think you need the sequence in order to get the list of items, anyway
16:18:30
and the order
16:23:33
danbri, no grddl, sorry
16:24:08
I wonder how tied to RAP their grddl stuff is
16:24:59
i had a look at it once
16:25:59
it seemed a bit tied if I remember - I think by default it outputs a RAP model object, rather than RDFXML
16:37:08
danbri checks license; LGPL