-
Knio has quit
2005-03-16 07535, 2005
-
Knio-sleep joined the channel
2005-03-16 07543, 2005
-
StupidMop joined the channel
2005-03-16 07516, 2005
-
StupidMop
hello hello HELLO!
2005-03-16 07516, 2005
-
cikkolata
hi
2005-03-16 07521, 2005
-
StupidMop
okay... I am curious to the benifit of having the albumjoin table. Couldn't we have albumid and squence fields put into the track table ?
2005-03-16 07556, 2005
-
cikkolata
the answer is probably something related to optimising stuff
2005-03-16 07504, 2005
-
cikkolata knows nothing about database design
2005-03-16 07529, 2005
-
StupidMop
hhmmm...
2005-03-16 07523, 2005
-
DJKC has quit
2005-03-16 07530, 2005
-
DJKC joined the channel
2005-03-16 07528, 2005
-
Knio-sleep
Knio-sleep is now known as Knio
2005-03-16 07514, 2005
-
Knio
Knio is now known as Knio-sleep
2005-03-16 07506, 2005
-
cikkolata
damn!
2005-03-16 07518, 2005
-
cikkolata
germany is winning over japan over release dates.
2005-03-16 07500, 2005
-
cikkolata
2005-03-16 07524, 2005
-
Knio-sleep has quit
2005-03-16 07533, 2005
-
Knio joined the channel
2005-03-16 07526, 2005
-
duncf has quit
2005-03-16 07542, 2005
-
dupuy
cikkolata, if there's any way to search for these:
2005-03-16 07544, 2005
-
dupuy
26 = France, Metropolitan
2005-03-16 07553, 2005
-
dupuy
should all be changed to France
2005-03-16 07500, 2005
-
dupuy
(i.e. FX -> FR)
2005-03-16 07513, 2005
-
orogor joined the channel
2005-03-16 07512, 2005
-
yalaforge joined the channel
2005-03-16 07507, 2005
-
yalaforge
StupidMop: you're right, at the moment the albumjoin table isn't really necessary because MB only uses 1:n relationships between album and track
2005-03-16 07538, 2005
-
yalaforge
the idea is to switch to a m:n relationship later where a track may be on multiple albums
2005-03-16 07508, 2005
-
yalaforge has quit
2005-03-16 07539, 2005
-
SmartMop joined the channel
2005-03-16 07513, 2005
-
Knio-sleep joined the channel
2005-03-16 07509, 2005
-
Knio has quit
2005-03-16 07557, 2005
-
StupidMop has quit
2005-03-16 07502, 2005
-
SmartMop has quit
2005-03-16 07506, 2005
-
RJ joined the channel
2005-03-16 07522, 2005
-
vald
hm, is it possible to query mb server for related artists?
2005-03-16 07531, 2005
-
Kanmu joined the channel
2005-03-16 07531, 2005
-
lindestinel has quit
2005-03-16 07521, 2005
-
orogor has quit
2005-03-16 07539, 2005
-
RJ has quit
2005-03-16 07552, 2005
-
rj_ joined the channel
2005-03-16 07531, 2005
-
DJKC
2005-03-16 07550, 2005
-
orogor joined the channel
2005-03-16 07538, 2005
-
ramoz has quit
2005-03-16 07512, 2005
-
ramoz joined the channel
2005-03-16 07526, 2005
-
Knio-sleep
Knio-sleep is now known as Knio-school
2005-03-16 07530, 2005
-
yalaforge joined the channel
2005-03-16 07543, 2005
-
enjayhch
eek
2005-03-16 07501, 2005
-
enjayhch
DBD::Pg::st execute failed: ERROR: column "current_replication_sequence" is of type integer but expression is of type character varying
2005-03-16 07531, 2005
-
yalaforge
what did you do?
2005-03-16 07551, 2005
-
enjayhch
.
2005-03-16 07553, 2005
-
cikkolata
break it, obviously :P
2005-03-16 07507, 2005
-
enjayhch
during data import
2005-03-16 07507, 2005
-
cikkolata points out the obvious
2005-03-16 07526, 2005
-
enjayhch
that was only 10mins after the load started
2005-03-16 07534, 2005
-
enjayhch
but I did a RDF dump after that
2005-03-16 07536, 2005
-
yalaforge
on the redhat machine?
2005-03-16 07555, 2005
-
enjayhch
yup
2005-03-16 07500, 2005
-
enjayhch
and got 1.8Gb out
2005-03-16 07508, 2005
-
yalaforge
so you didn't import yesterday. hm.
2005-03-16 07514, 2005
-
yalaforge
let me have a look
2005-03-16 07530, 2005
-
enjayhch
well I think it was towards the end of the import
2005-03-16 07539, 2005
-
enjayhch
so maybe it doesn't matter
2005-03-16 07554, 2005
-
yalaforge
phone
2005-03-16 07504, 2005
-
cikkolata
dupuy: I had wondered whether that should be done (had contemplated a mailing list post), I'll fix them now.
2005-03-16 07526, 2005
-
yalaforge
back
2005-03-16 07544, 2005
-
yalaforge
towards the end of the import? were indexes already created?
2005-03-16 07555, 2005
-
enjayhch
going to run it again, but think so
2005-03-16 07558, 2005
-
yalaforge
phone. again. argh!
2005-03-16 07543, 2005
-
yalaforge
back again
2005-03-16 07529, 2005
-
yalaforge
changing line 173 in MBImport.pl might work
2005-03-16 07539, 2005
-
yalaforge
($iReplicationSequence eq "" ? undef : $iReplicationSequence+0),
2005-03-16 07527, 2005
-
yalaforge
currently the DB driver seems to quote the $iReplicationSequence because it thinks it's a string
2005-03-16 07537, 2005
-
enjayhch
ok
2005-03-16 07554, 2005
-
enjayhch
I will wait for this import to finish and might try again in a sec
2005-03-16 07557, 2005
-
yalaforge
adding 0 makes it an integer. but if the variable contains something wrong we should at least get a warning
2005-03-16 07503, 2005
-
enjayhch
yup
2005-03-16 07552, 2005
-
enjayhch
wow, really does only take 10mins to import
2005-03-16 07505, 2005
-
yalaforge
now it worked?
2005-03-16 07516, 2005
-
enjayhch
nope, same failure
2005-03-16 07521, 2005
-
enjayhch
I will make your change now
2005-03-16 07518, 2005
-
enjayhch
also moving the postgres database to a different disk
2005-03-16 07543, 2005
-
yalaforge
shut it down first :-)
2005-03-16 07559, 2005
-
enjayhch
have done :)
2005-03-16 07559, 2005
-
ramoz has quit
2005-03-16 07519, 2005
-
yalaforge
strange. on my postgres 7.4.6 I can assign strings with a number in it to integer columns
2005-03-16 07541, 2005
-
yalaforge
I would have expected an error message
2005-03-16 07516, 2005
-
cikkolata
dupuy: I think all of the FX ones are now FR.
2005-03-16 07532, 2005
-
enjayhch has quit
2005-03-16 07537, 2005
-
cikkolata
the user europop seems to have been responsible for 19 of them.
2005-03-16 07525, 2005
-
DJKC
watch, and now you'll find that there were actually different versions released in parts of France and they were actually correct ^^;
2005-03-16 07519, 2005
-
cikkolata
heh
2005-03-16 07546, 2005
-
cikkolata
metropolitan france is covered by france though
2005-03-16 07551, 2005
-
cikkolata
so they're still right.
2005-03-16 07528, 2005
-
DJKC
yes, but it would be amusing if there was say a paris release different from the rural France release
2005-03-16 07546, 2005
-
DJKC
uncensored version for the big cities, censored for the small towns
2005-03-16 07555, 2005
-
cikkolata
I guess it would be amusing...
2005-03-16 07518, 2005
-
cikkolata
but I don't think that happened.
2005-03-16 07537, 2005
-
enjayhch joined the channel
2005-03-16 07552, 2005
-
dupuy
in any case, the FX country code has been retired (it's not longer used)
2005-03-16 07521, 2005
-
enjayhch
gah, university network lost internet access
2005-03-16 07536, 2005
-
enjayhch
wow this server is bloody fast
2005-03-16 07548, 2005
-
enjayhch
did a comeplete import and indexing in 6mins
2005-03-16 07502, 2005
-
yalaforge
what about the error?
2005-03-16 07509, 2005
-
enjayhch
still there :(
2005-03-16 07520, 2005
-
yalaforge
damn.
2005-03-16 07530, 2005
-
cikkolata
hmm, I have 26 FX->FR changes, so it seems like I got them all.
2005-03-16 07551, 2005
-
yalaforge
as I said, I can assing strings to integer columns in postgres 7.4.6
2005-03-16 07542, 2005
-
enjayhch
but it looks like the value is fine, most odd
2005-03-16 07515, 2005
-
yalaforge
yup
2005-03-16 07550, 2005
-
yalaforge
well, it is the last thing that the import script does
2005-03-16 07512, 2005
-
yalaforge
so you can at least use the DB without replication until we find out what's wrong
2005-03-16 07522, 2005
-
enjayhch
yeah
2005-03-16 07535, 2005
-
enjayhch
I am still in shock that it only took 6mins
2005-03-16 07551, 2005
-
yalaforge
I never heard of an import faster than 1 hour
2005-03-16 07554, 2005
-
enjayhch
took 4 hours on other server
2005-03-16 07506, 2005
-
yalaforge
well, it was all in the buffer cache after all
2005-03-16 07518, 2005
-
cikkolata
full import?
2005-03-16 07520, 2005
-
yalaforge
probably no disk io was required for reading
2005-03-16 07530, 2005
-
enjayhch
could be, yeah
2005-03-16 07530, 2005
-
yalaforge
cikkolata: almost. mbdump + derived
2005-03-16 07550, 2005
-
enjayhch
although it was unbziping it all again
2005-03-16 07510, 2005
-
yalaforge
hm. anyway, it's very fast
2005-03-16 07510, 2005
-
cikkolata
ah
2005-03-16 07512, 2005
-
cikkolata
damn.
2005-03-16 07531, 2005
-
cikkolata
I did mbdump on here in and half and a half (exactly)
2005-03-16 07556, 2005
-
yalaforge re-reads the sentence
2005-03-16 07518, 2005
-
cikkolata
errr
2005-03-16 07519, 2005
-
cikkolata
hour and a half
2005-03-16 07545, 2005
-
ramoz joined the channel
2005-03-16 07501, 2005
-
enjayhch
oh, I have been loading mbdump-moderation by accident
2005-03-16 07521, 2005
-
yalaforge
np, that's rather small
2005-03-16 07532, 2005
-
yalaforge
closed moderations are very large
2005-03-16 07545, 2005
-
enjayhch
yeah, not bothering with that for sure ;)
2005-03-16 07538, 2005
-
enjayhch
next is to see if JFS can cope with 387462 text files
2005-03-16 07558, 2005
-
yalaforge
what for??
2005-03-16 07517, 2005
-
enjayhch
going to create one file of RDF per artist and album
2005-03-16 07535, 2005
-
enjayhch
makes loading into triplestore a lot more efficient
2005-03-16 07552, 2005
-
yalaforge
but is going to bring your machine to its knees
2005-03-16 07558, 2005
-
yalaforge
if it was a sane machine, that is
2005-03-16 07505, 2005
-
enjayhch
and can also refer to an individual artist's RDF file
2005-03-16 07519, 2005
-
enjayhch
well we shall see !
2005-03-16 07542, 2005
-
enjayhch
bloke downstairs seemed to think it was a good idea !
2005-03-16 07533, 2005
-
yalaforge would generate the RDF on the fly as needed :-)
2005-03-16 07503, 2005
-
enjayhch
would be tricky to keep triplestore up to date that way
2005-03-16 07516, 2005
-
enjayhch
we have the disc space... and it is his server... !
2005-03-16 07527, 2005
-
yalaforge
good luck :-)
2005-03-16 07530, 2005
-
enjayhch
argh, I really don't know what is wrong with this replication SQL problem
2005-03-16 07548, 2005
-
yalaforge
neither do I
2005-03-16 07559, 2005
-
enjayhch
7340 looks like an integer to me
2005-03-16 07514, 2005
-
yalaforge
what is it set to atm?
