[mb-users] "The need for free and open music metadata", but MB not good enough

[Chris B]

2008/4/28 david scotson <david.scotson at gmail.com>:
> On Mon, Apr 28, 2008 at 11:12 AM, Chris B <chris at whenironsattack.com> wrote:
>  > as with the recent discogs offshoot (go and have a look in the discogs
>  >  forums today to end any doubt that making music DBs "like wikipedia"
>  >  is a good idea - discogs is essentially ruined right now), it's beyond
>  >  me how anyone can take irritations with existing DBs and solve them
>  >  with making an entirely new DB!
>
>  I had a look at the discogs forums and there appears to be some major
>  user revolt going. However, as an outsider, I had little to no clue
>  what had actually caused all the fuss. The main things seems to be
>  making it easier to get info into discogs (hence easier to get bad
>  data in) and not rewarding users with ranking points in return for
>  their work under the new system.
>
>  I don't suppose you're aware of any more comprehensive write-ups of
>  what happened, particularly one that explains stuff so that
>  non-discogs folk can get a handle on things?

what's happened in discogs over the last 2 months (or more?) is a
textbook example of how not to introduce large changes to a project:

- allowed variable data quality 'votes' ('complete & correct'/'needs
major chagnes'/'needs minor changes'/'entirely incorrect' - like
musicbrainz 'quality' concept) which is seemingly an improvement over
the Y/N (accepted/not) concept from before.

- edits to the data now get applied instantly, regardless of the
dataquality specified above. the benefits of variable dataquality are
completed irradicated, then.

- the application of such votes is broken anyway -
http://www.discogs.com/help/forums/topic/160753#2088421 - the
management seem to be totally oblivious to this fact. i've been saying
this same stuff since day 1 and no changes.

- there are no restrictions put on users who constantly add poor data.
they have to be individually reported to the one administrator, whilst
they carry on and add what they want (many are motivated by the
marketplace - getting their stock online regardless of the data, so
there are a LOT of 'bad' users). the 'entirely incorrect' votes
neither remove releases or restrict users...

- it was also thought that in amonst this carnage, it would be a great
idea to completely re-write the moderator interface, to the extent
that it is now almost completely broken. moderators used to making
many 100s of votes per day can now only effectively keep track of
their own collection.

- add to this a number of fundamental rule changes introduced at the
same time (eg, labels aren't companies now - this is a rather
multifaceted change that i could talk all day on!), without required
data structures in place to adequately control them.

all of this has almost completely destroyed the discogs moderator
community. our questions don't get answered, and theres a constant
flow of unwanted 'features' whilst the key problems don't get
addressed. bear in mind that this entire feature set literally turned
up overnight with no consultation/testing with ANY users. now the live
data is suffering a death by a 1000 tiny cuts, and all that are left
are the 'yes men'. the level of expertise and dedication many of the
moderators had on discogs was astounding, but they all seem to have
left now.

it's obvious what's going to happen now - the problems are going to be
ignored, and the voting is going to be opened up to all users, which
would be fine if done right, but the useless interface and lack of
logic to the voting options means that bad votes are going to be done
to bad data. discogs will no longer be accurate, but the votes (and
dollars) keep coming in.

rant over :) it's really upsetting, though.