Brian Schweitzer wrote: > > Sure there are composer-pianists who have worked for musical academies > teaching composer-pianists-to-be. I do not see why MB should be rich > enough to let you say: TeacherA taught StudentA SubjectA and SubjectB, > TeacherA taught StudentB only SubjectA. > > > Exactly the type of thing I was thinking of. I've been reading a book > by Steve Reich, where he talks about various teachers he had and his > musical relationships with classmate John Cage. That's the kind of > richness that's quite interesting to me. If I can see that Foo taught > both Bar and Pez, at the same time, that's interesting - it's not > conclusive, but from that I can also guess that either Bar and Pez knew > each other at that time, or at least that they had many of the same > influences, which would likely influence their later output. So I guess > I would rephrase your question: why should MB *not* be that rich in > detail, even if it is not directly tied to a specific recording or > work?
I said I found the Teacher-Student relationship interesting, so given a bona fide artist teacher of both Cage and Reich, I'd be happy to see them linked like you describe :) (and thereby enabling users to "discover completely new music!" as About_MusicBrainz says) What I don't get for the teacher-student AR is why is it good to split them up? It seems to me more a task for a prose biography, e.g. page on wikipedia, to have that level of detail, because it will be easier to make the wording come out right and also since you will be putting a further strain on the voting system. Come to think of it, one could perhaps argue that a nicely formatted list of students on the annotation for TeacherA, and similar lists of teachers for StudentA ... StudentZ would do this even better than you propose because you no longer have any AR wording constrains to worry about. For human users, this is just as good. (After all, we track who made travel arrangements for a band to > get to the studio to record a track... :D) > > (A couple of random thoughts about adding these ARs and making the > expressivity of MB richer > > - what happens when we make voting difficult for other than dedicated > fans/experts? won't it just be noise putting other voters off? > > > To be honest, the voting situation's gotten even worse since I ran > numbers 3 years ago; back then, we had 10-15% at 14,000 edits per 2 > weeks, we've recently been at more like 5% on 20-25,000 edits per 2 > weeks. Given that so little data at all gets verified, I think the > better place to look for verification would come from the person using > the data who notices it's wrong, rather than the person who's voting on > the data. (I was reading http://www.nybooks.com/articles/21131 earlier; > the bit about fixing an incorrect '40 people die of this' struck me as > quite similar to how I see our current situation.) Sounds a little like you're moving towards a position saying edits should no longer need votes (perhaps except for the high quality stuff). > - what quality can we expect of this data, what's the low threshold for > what we find acceptable?) > > > Autobiographies, biographies, class yearbooks, artist/teacher-attributed > statements, liner notes, concert bios, etc. - perhaps anything > non-tabloidish? Personally, I'd go with anything that's 'trustworthy', > however you define that - if we go too specific, we'll just end up > blocking good info. The sources of this info could come from many places, I agree. I was thinking more in terms of broadness/coverage and reliability. With fewer votes given and unverifiable sources, this data will perhaps not be of much value for non-human agents. Leiv / leivhe _______________________________________________ MusicBrainz-style mailing list MusicBrainz-style@lists.musicbrainz.org http://lists.musicbrainz.org/mailman/listinfo/musicbrainz-style
