[mb-style] Composition/Performer/Production ARs at Release or Track level?

[Olivier]

2008/1/2, Brian Schweitzer <brian.brianschweitzer at gmail.com>:
>
> > I've like to ask you to not do this. Currently we have the data we have,
> > and we can say they are correct. If you move all of them automatically
> > to tracks, you will add many incorrect ARs to the database. Leave this
> > to people who actually know about the releases/tracks.
> >
>
> Sure, ok, I won't do it.  :)
>
> Now what this tells me is something interesting, especially that there were
> 3 messages saying essentially the same thing all so quickly following my
> email.  It's essentially the same thing luks tried to say to me in IRC last
> night, which, once I'd thought on it, convinced me to change my mind.
>
> In essence what these three messages imply is this:
>
> Even for editors who are willing to state that they 100% never varied from
> viewpoint #1, and only ever added release-level ARs if the ARs applied to
> every single track, people still don't want those ARs auto-converted to
> track-level ARs.  In essence, even in the absolute most extreme case of
> viewpoint #1, people still view that data as suspect, or at least, in need
> of verification before those ARs are attached to tracks.

I'm not sure I'm counted in that or not :-)
Either way, I don't actually view any data I entered as suspect :] -
though double checking sure can't hurt.
Now, there are data I entered at release level *which* just should be
applied to all tracks (entered on release back in the time because it
was simpler), and there also are data that I entered at release level
that *must stay* there (producer and others).
You seem to imply my reserve was about data quality - it's not. It's
about your conclusion that any data at release level is "fuzzy".

>
> In other words, any release-level AR, no matter who added it, is
> automatically viewed as being fuzzy, even if the editor is categorically
> willing to say that he or she has never added a fuzzy release level AR.

No, again, see above.

>
> In even simpler terms?  Viewpoint #2 is already in effect, no matter what
> those of us using Viewpoint #1 thought.  It just then remains for us to
> actually put that into a guideline, so everyone knows exactly what is going
> on.

I stated it earlier, I just don't agree. You're deliberately ignoring
the fact some ARs have a specific meaning on release level, and it's
not fuzzy.

- Olivier