------=_NextPart_000_0026_01BE6A74.AC49C960
Content-Type: text/plain;
charset="iso-8859-1"
Content-Transfer-Encoding: 7bit
Hi: I have just joined this group.
I agree with this thinking. Categories should be highly flexible. New
categories will certainly arise in the world of music.
For the last few years I have been working on a database for my own
collection of about 450 discs. My work so far has been with Microsoft Access
though I'm intending to move it to some other RDBMS that I can develop with
Java.
Some of the issues I have been dealing with are:
-"Albums" which have multiple discs.
-Works which span more than one track; a string quartet with four movements,
for example.
-Multiple "takes" of the same piece as are found on some Jazz discs.
-Changes of group members, instruments and "guest artists".
-The same recording appearing on more than one disc. For example one may
have a disc like
"The Best of Albino Wino" and also some of the albums where these tracks
were first released.
-Compilation discs with a number of separate artists.
I never even thought to include the concept of "genre" though, it seemed
too arbitrary. When the CDDB appeared I had to make some use categories in
order to move track information into my database.
Most of the messages I have read here so far seem to concerned with matters
of implementation, a much smaller number (such as this one) are talking
about database design. Should these topics be broken out into two separate
threads? The implementation stuff that I've read here I don't understand
at all. Can I be of any use in the other department?
-larry
-----Original Message-----
From: owner-cdindex nospam at freeamp.org [mailto:owner-cdindex nospam at freeamp.org]On
Behalf Of Paul Quinn
Sent: Tuesday, March 09, 1999 7:30 PM
To: CD Index
Subject: Categories
On the current topic of Categories and category granularity, CDDB
has it all wrong. It's current 'official' categories are horribly
unrepresentative of what is in its database. 50% of the entries are under
'Rock' alone. A new category system would have to either more granular or
much more dynamic. Here's an idea just from the top of my head. Each entry
has two categories, the main category from the official list, and a
secondary category that is set by submitters indicating a more appropriate
category if need be. The database would over time tally these secondary
categories and determine whether a new official category should be
generated. Bien?
-Paul
------=_NextPart_000_0026_01BE6A74.AC49C960
Content-Type: text/html;
charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable
<!DOCTYPE HTML PUBLIC "-//W3C//DTD W3 HTML//EN">