I don't quite follow, for which letter is that output?
2011-08-01 21345, 2011
stereoket joined the channel
2011-08-01 21300, 2011
jacobbrett1
That's the additonal notes for U+0131 LATIN SMALL LETTER DOTLESS I
2011-08-01 21322, 2011
ocharles
hum
2011-08-01 21343, 2011
stereoket_ joined the channel
2011-08-01 21344, 2011
ocharles
Should adding a disc id for a medium immediately set track times, if they are all undefined?
2011-08-01 21349, 2011
nikki
yes!
2011-08-01 21314, 2011
nikki
modbot will only do it later otherwise
2011-08-01 21324, 2011
nikki
(but in the meantime people just open more pointless edits)
2011-08-01 21356, 2011
ocharles
yea
2011-08-01 21321, 2011
luks
ocharles: that's a very old ticket and requires a patched version of libunac
2011-08-01 21337, 2011
luks
the search server and the direct search both convert the characters
2011-08-01 21347, 2011
ocharles
ah, ok
2011-08-01 21349, 2011
luks
the patch is in the old ticket in trac
2011-08-01 21321, 2011
ocharles
I did consider just offloading the work to db
2011-08-01 21327, 2011
ocharles
SELECT musicbrainz_unaccent(?)
2011-08-01 21337, 2011
nikki
ocharles: I'm not very good at explaining, but what I meant in the original trac ticket was that when a language has ı it has I as the capital letter, and there's a separate letter i with İ as the capital letter... and you can imagine how messed up the dot/no dot gets when normally i corresponds to I... so it needs fixing a lot
2011-08-01 21345, 2011
luks
I'd consider taking over Text::Unaccent
2011-08-01 21315, 2011
ocharles
I wonder if I can request co-maint on that
2011-08-01 21324, 2011
ocharles
seeing as it hasn't been updated in years
2011-08-01 21334, 2011
luks
we've talked to the guy years ago, he said he will commit the patch to CVS and nothing happened
2011-08-01 21350, 2011
ocharles
I can probably request co-maint as it's abandoned then
2011-08-01 21354, 2011
nikki
and normally adding/removing an accent is an auto-edit, like doing o -> ö in swedish, so I don't see why i İ ı I should be any different
2011-08-01 21359, 2011
ocharles
yea
2011-08-01 21306, 2011
ocharles
nikki: I just wanted to make sure they count as the same letter
2011-08-01 21321, 2011
ocharles
(when unaccented)
2011-08-01 21331, 2011
luks
but yea, using the database would be the simplest solution for now
What't the best way to add a new multi-medium release with discids?
2011-08-01 21306, 2011
reosarevok
cjk32: I think either adding the release and then attaching the IDs or adding each disc and merging them
2011-08-01 21306, 2011
dinog1 joined the channel
2011-08-01 21315, 2011
reosarevok
But maybe it has been made simpler…
2011-08-01 21317, 2011
Xof joined the channel
2011-08-01 21329, 2011
hawke_
Pretty sure that’s right…
2011-08-01 21341, 2011
cjk32
Creating the release then adding discids seemd better than created five releases and merging - doesn't depend on edits being approved before its correct in the database.
2011-08-01 21351, 2011
reosarevok would agree
2011-08-01 21309, 2011
ojnkpjg joined the channel
2011-08-01 21313, 2011
reosarevok
The recordings probably won't get the right track lengths yet
2011-08-01 21328, 2011
reosarevok
But they will once a system for them to be taken from tracks is developed
2011-08-01 21334, 2011
reosarevok
(and that's in the plans)
2011-08-01 21336, 2011
reosarevok
So…
2011-08-01 21308, 2011
hangy joined the channel
2011-08-01 21323, 2011
cjk32
I'm not sure that's working correctly anyway. Adding a single disc release didn't provide track-lengths / recording lengths from the toc.
2011-08-01 21353, 2011
cjk32
Actually, it looks like I can start adding a release based upon one discid, then add new media to it.
2011-08-01 21336, 2011
cjk32
Gah, also looks like ISRCs aren't anywhere near as unique as one might hope!
from what he said, due to an evil global variable that kept references to past ISRCs :)
2011-08-01 21309, 2011
ianmcorvidae
needs more functional programming
2011-08-01 21310, 2011
ianmcorvidae
;)
2011-08-01 21358, 2011
cjk32
Ah, hence the requirement for a unique user agent, I guess
2011-08-01 21344, 2011
cjk32
Advocating lisp?
2011-08-01 21317, 2011
ianmcorvidae
nah, haskell's better if you really want to go functional XD
2011-08-01 21326, 2011
ianmcorvidae
I like lisp plenty, but it's not pure-functional
2011-08-01 21332, 2011
ianmcorvidae
it just supports a functional style well
2011-08-01 21300, 2011
luks
that doesn't save you from accumulating "global" state :)
2011-08-01 21326, 2011
ocharles
HASKELL
2011-08-01 21349, 2011
LeoD_
HASKELL!!
2011-08-01 21302, 2011
cparker
I have an odd scenario... there's a single track/recording that's credited to a person "and Family". Is it appropriate to create a new artist for this single collaboration? Or should the recording be attributed to just the person, with the track's artist listed with the "and Family" tacked on the end?
2011-08-01 21319, 2011
CatCat
LEARN OYOU A BOOKBRAINZ
2011-08-01 21344, 2011
cparker
I already have performance relationships added to the recording for each of the members of this person's family, as they're listed in the liner notes. Should these relationships, plus the "and Family" in only the track listing, suffice? It seems kind of silly to create a new artist just for this one-time collaboration on a single song.
