yeah...reading the replication data in the background doesn't seem to be that err simple
hrmm
FauxFaux
Go look at how the Lucene index is maintained?
Compressed
im using lucene
well ferret is the ruby port of lucee
err lucene
wait..is there an already maintained lucene index for mb?
luks
yes, but it has the same problem
Compressed
ah
hmm
how often is the database updated? every day?
luks
you mean the existing lucene index?
that is rebuild every day
Compressed
oh
luks
replicated databases are updated hourly
Compressed
how long does the index rebuild take to run?
luks
a couple of hours, if i remember correctly
Compressed
oh wow
hmm
i wonder why ferret is taking so long
luks
I've never done it myself, so I don't know any more detailed number
Compressed
although i dont have that heavy hardware
right
well when i ran it on an opteron 185 it took about 4 days...
FauxFaux
It's taking so long 'cos it's written in ruby, duh.
Compressed
hah u think?
FauxFaux
Compared to Java? Hell yes.
Compressed
hmm maybe i just just try running the java version of lucene
ojnkpjg
is java considered fast now?
Compressed
i could care less how i get the index, as long as i get it in less than 4 days heh
FauxFaux
Tee hee, I know this is a downhill argument,but, in general, for long-running tasks, some things > sensible vm'd languages > static languages > scripting languages in terms of speed.
stochasticism has quit
Compressed
hmm looks like ferret can work with an existing lucene index .. :)
maybe java will hold the key hehe
aCiD2 has quit
thanks for the help
stochasticism joined the channel
dalen has quit
guigouz has left the channel
Apollozeus has quit
luks has quit
stochasticism has quit
catgruff joined the channel
Kerensky97 joined the channel
catgroove has quit
catgruff has 茉莉花茶
Kerensky97
I wish trillian wasn't so unicode deficient. ^^looks like spaghetti to me.