User:AaronCooper/Spring Cleaning Classical Releases

From MusicBrainz Wiki
  • Attention.png Orphaned, no category, no link, no named editor, no real interest as far as I can tell. Suggesting DeletedPage. Please object if I'm wrong. -- dmppanda 16:57, 01 April 2007 (UTC)
    • Should this not be added to CommunityProject? With a status notice that this was an attempt which did not take off (but might still make sense)? -- DonRedman 11:38, 06 April 2007 (UTC)
      • I'm a bit more crude than you about this page. IMO, this doesn't contain anything interesting (it just articulates a serie of obvious evidences about routine editing, and doesn't try to coordinate any tidying/community effort). I would be more than glad that the ClassicalEditors do something about their mess in the database, but actually this doesn't look like it, and I see no point in keeping pages that are mostly "somebody speaking to him/herself" :-). Author of this page: please ProveMeWrong as I would really love to see a joint ClassicalTidyingEffort. -- dmppanda 20:04, 06 April 2007 (UTC)
        • I authored this page to explain why I think we should be able to remove "useless" classical releases. -AaronCooper

Preface

I've been thinking that we, the MB users, should consider doing some tidying of the existing classical releases currently in the database. Obviously, we are editing and improving data regularly but I think that there are several releases that may not being used, may never be used, and probably should not be used for tagging or other purposes (such as referencing how to enter a specific work). I propose we set out to do some spring cleaning and make our existing, correct data more available and "findable".

What defines a "useless" release?

  • No performer information exists in release titles/track titles or AdvancedRelationships to disambiguate from similar releases.
  • No related URLs exist to lookup the performer information.
  • No proof was provided in the edit history which might lead us to performer information.
  • The original editor does not respond to our request for more information [after X amount of time].
  • The FreeDB disc ID does not lead us to any valuable information.

Why are "useless" releases bad for MB?

Without performer information, we end up with several... and I mean *several* releases with no information to disambiguate performances (excluding perhaps track times which are very hard to look up, are not all that trustworthy, and are not always present). For example, our friend Wolfgang Amadeus Mozart has 12 releases of 'Requiem in D minor' without any performer information. A particularly brutal example has ELEVEN (11) completely different disc IDs!

These "useless" releases:

  • make it harder to find "useful" releases,
  • are being used to tag music incorrectly,
  • are being referenced to write incorrect track/release titles for new releases, and
  • cannot be improved since we have exhausted all means available to obtain useful information.
  • may be used as examples of track names, leading to more track names that do not follow the CSG
  • may be used as examples for release titles, leading to more bad releases without performer info, etc.

What should we do about "useless" releases?

Once a release has been determined to be "useless" it should be removed from the database. These releases are only likely to degrade in quality as more and more disc IDs, TRMs, and PUIDs are erroneously applied to them. So, let's help all the users who primarily use MB for tagging purposes and clean up our classical database to make it more accessible, more efficient, and a more useful tool.

How do we keep the database in good order?

Of course, all this would be in vain if we did cleaned out the database and then allowed new releases to be entered in a "useless" state. Perhaps we will need to write up some documentation on how to successfully add a classical release. By this, I do not mean more style guidelines but rather a checklist of what is required before a classical release will be accepted into the database. This could include performer information such as any orchestras, conductors, or notable soloists, among other pieces of information deemed necessary (some sort of proof verifying tracklist and performers would also be beneficial).

Discussion

  • Some releases do not have performer information. -- davitof