This is a public chat log generated from the #semsol IRC channel.
23:00:09
what times is he usually found here
23:00:16
the only developer
23:00:21
ahuh
23:00:35
though some of us contribute patches and plugins from time to time
23:00:41
usually 9-5 weekdays
23:00:55
i am guessing USA time
23:00:57
afaik
23:01:03
Europe time sorry
23:01:17
he's in Germany
23:01:29
ok.. I am based in China.. I will be back tonight..
23:01:35
Thks kwijibo
23:01:36
so approx -1 from GMT
23:01:43
np
23:01:54
how big is your dataset out of interest?
23:02:00
59M triples
23:02:12
wow
23:02:50
128 m doesn't sound like enough probably
23:03:57
it would be interesting to see what bengee says: the parser streams, so perhaps memory shouldn't need to be an issue
23:04:19
however i have hit memory issues parsing large rdf docs in the past myself
23:04:48
the parser is well written, the streaming is faultless
23:05:01
its just the SQL reads and writes that are killing me
23:05:29
I need to increase the keys in the buffer for all the tables to reduce the SQL reads
23:07:48
for every triple insert there is about 6 reads and 4 writes, now multiply that by 59M
23:08:46
kwijibo, do you know how to stream a bz2 file using arc with LOAD
23:10:44
no sorry :(
23:11:42
np
23:11:45
bbl