I am from India and as I saw there weren't many matches found on picard for hindi songs. So I wanted to fetch songs data from the sites from where I can download the songs and update the database.
Fetching part is almost over.
Ok, a question. Let's take this release http://musicbrainz.org/release/97a99ab0-816f-4d.... On back cover `label #` is clearly written as `YSE 017` (just above the barcode). DiscoGS uses strange `YSE *-(D)CD`-like # for volumes I've seen util now, while yet elsewhere it is yet again different permutation of YSE, 017 and random letters. Would it be sensible to simply use (and change already entered releases) to use `YSE xxx` or `YSE-xx
TrigonaMinima: if you're going to add releases, I'd suggest not doing it fully automated. Either write an importer (you can seed the release editor, see http://musicbrainz.org/doc/Development/Release_...) and make a last check by hand before entering, or at least add only a few dozen per day and review that they look good
Having more data is certainly good, mind, but we'd just love to ensure it looks fine :)
simukis_: that specific one at least seems to clearly be without -CD
Maybe the medium has -CD written on it?
If not, no idea where it came from
Both of mediums have.
But then, again, it's the release label #, not medium.
Well sure, but anywhere on the release that's still a catno
I don't really have a particularly useful opinion here - I'd be fine with standardising to only one of them, but no idea which one really
Someone who does edit this kind of music would be more helpful :)
Is there any particular reason for not having an API ?
One is that someone would have to write it
The other is that (at least until recently) the community didn't particularly trust mass-editing anyway
I'm not sure whether that has changed
But mostly, really, is that, that we're tiny and nobody has written it
(it's in the plans, eventually)
Yes for editing. And if there's a doubt for correctness of the editings (for autoediting on a mass level) then it can always be limited to a few number.
_5moufl joined the channel
drsaunders joined the channel
\o/ for more non-western centric editors
Can this be done as a GSoC Project ? I mean is making an API big enough for a proj ?
I imagine it would be big enough, yes - I don't know if it would fit for this year because I don't remember what other changes we were supposed to make before it
But you can certainly suggest it
Is there any feature list or bug list ?
TrigonaMinima: For mass editing/data importing/digesting, the current goal is to use Geordi. I'm not sure if that's applicable for wherever you're importing your data from though.
TrigonaMinima: However, mass-adding of data is generally untrusted, since it has a bad history of adding... less than pristine data.
E.g., capitalisation for titles in Latin script, formatting of titles, proper use of attributes, etc.
Thus, an importer, as reosarevok suggested, is far preferred, though a bot that adds a few releases that can then be looked over before making a new batch I guess could also work.
This is unlikely to change with an API, though an API would enable better/custom importers (e.g., instead of just feeding the release editor) as well as client-side editing, such as queing up edits while offline and then sending them off when reestablishing a connection.
But more or less blindly mass-adding data will probably never be favoured by the community.
Yeah, I am asking for that. I mean instead of needing human hands to add more data, they can be used in checking them. And i think this will make it easier and quicker only (provided the API or bot adds a few releases only ).
And having an API will bring more people in editing , I guess.
murk_ joined the channel
The same thing could be achieved by letting people simply provide an URI to other metadata website and MB pulling down + parsing all the data before editing process.
peaveyman joined the channel
I am sorry, but I am not following...
reosarevok: in what way iTunes are capable of enforcing their ToS in this aspect?
Well, they're certainly able to block us from their affiliate program
They'll probably not go and sue us for it or anything (but if they did we'd have no money to fight back and would be forced to shut down, which means it probably won't happen anyway :p)