Re: Decibel
Calvin Walton <[email protected]>
| Newsgroups | gmane.comp.audio.musicbrainz.devel |
|---|---|
| Message-ID | <1395845983.2559.4.camel@sasami> |
On Wed, 2014-03-26 at 12:15 +0000, Sophist wrote: > On March 25, 2014 at 2:45 PM, Robert Kaye wrote: > > > On Mar 25, 2014, at 2:08 PM, Sophist <[email protected]> wrote: > > > > > 5. They have apparently accumulated data about 15 million tracks on 1.1 > > > million albums by 300k artists (https://developer.decibel.net/ ) How > > > does > > > that compare to MB? > > > > http://musicbrainz.org/statistics > > > > The numbers are amazingly similar, aren't they? > > Very similar. > > I have to admit that their description of how they "scrape their data" is > somewhat vague. On the one hand they are arguing that they have detailed, > accurate metadata, and on the other hand they are saying that they scrape > data from the internet (which IMO is not typically conducive to accuracy > without a LOT of sifting, filtering and cleaning). > > MB has a large community of editors working tirelessly to get the data > accurate over more years than Decibel has been in existence, and yet Decibel > can achieve this in a couple of years with a handful of data gatherers. > > It is a pity that their DB is not publicly available so as to see whether > they have the same minor data errors in their data as we have in ours They are currently offering access to their API for free to developers; it probably wouldn't be hard to check on this if you can find someone willing to agree to their API terms & conditions. That said, they do explicitly state that they include MusicBrainz release, track, and artist IDs as foreign IDs in their database - and any database that's linked to MusicBrainz IDs is a good thing in my view, if only because it promotes MusicBrainz a little. (Although it is better if their data is public, of course!) -- Calvin Walton <[email protected]>