[mb-users] Re: Clean vs. Explicit versions

[Frederic Da Vitoria]

2007/10/26, Chad Wilson <chad.wilson at gmx.net>:
>
> Frederic Da Vitoria wrote:
> > How is the presence of absence of beeping/squelching/silence/etc not a
> > rather audible difference between the two versions?  A 10 second
> > difference could simply be that the master was final mixed slightly
> > slower on one version of a CD release, etc.  The presence or absence
> > of beeps, blurps, squelches, silence, whatever, would seem at least
> > as, if not more, significant in differentiating the 2 releases as two
> > versions of the same release.  I've not seen the reverse addressed
> > though - why *should* the two versions be merged as a single listing?
> >
> >
> >
> > I agree this is not cool. I had issues with PUIDs being too
> > discriminant (I had different PUIDs for different compression levels
> > of the same tracks), now we seem to have not discriminant enough.
> > Maybe there is currently no better solution, but we must stay aware of
> > the problem.
> That's exactly my point, I'm sure I've seen several occasions where the
> PUIDs have been the same for a clean vs. explicit track that has the
> same track length.
>
> The problem with keeping multiple releases is that PUIDs and disc ids
> etc will /always/ be incorrectly attributed even if there is some funky
> annotation explaining it because very few users actually look at that
> info. It means you have two logical entities in theory, but you get a
> big mess of actual data relating to them making them essentially the
> same release after any period of time. That mess is difficult for anyone
> to clean up because the data is difficult to obtain reliably and few
> people have both copies.
>
> Remember that for most of us arguing the "other way", our argument only
> applies if the track times are the same. The majority of clean vs.
> explicit versions around have differing track times and/or tracks
> anyway, and I don't think anyone would argue to keep those instances in
> a single release.
>

I agree that if the PUIDs are the same, separating them may not be a good
idea. OTOH, when we have a conflict between data reliability and technical
limitations, we usually choose the data... If we agree on merging them if
the PUIDs are different (but how will we check this?) the fact that both
versions exist and that the MB release applies to both should be clearly
stated in the annotation IMO.

-- 
Frederic Da Vitoria
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.musicbrainz.org/pipermail/musicbrainz-users/attachments/20071026/9cb93c65/attachment-0001.htm