[Table of Contents]


[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[ARSCLIST] Clarification (was Re: A wiki-discopedia?)



Based on a private reply, I need to clarify what I meant by a
"wiki-like" system for our centralized 78 discography system:

*****

We have to remember that discographical data is "data", not "text",
and it is important that it be normalized, properly structured, etc.

If we make our discography strictly text-based (like what Brian Rust
did), that's worse than no data at all, even if we try to normalize
the text structure.

We also have to properly deal with "exceptions" to the dominant model
of how 78 discographical data is "structured."

For example, we have the Benny Goodman recording of "Popcorn Man"
for RCA Victor, which was issued as a side on a 78. It was quickly
withdrawn (only a few of the records made it to the public), and a
new 78 with the same record number was issued but with a different
side. So a discographical database model has to handle certain
exceptions that deviate from the well-known basic model.

(Another exception is how to handle the Fats Waller record, Vic
V-38050, where the labels for "The Minor Drag" and "Harlem Fuss" were
reversed on all discs. If one is to tie the discographical data to a
song/composition database, how does one handle this? Fortunately the
vast majority of 78s follow the basic model...)

Anyway, "wiki" itself, which is text-based, is not the approach, but
the wiki *principle* does suggest an open database structure allowing
entry and correction of data by those we allow, and to keep a running
history of changes so bad changes can be trivially reversed.

We also need to consider how to export the data in a standardized and
open form, and I recommend we develop an XML schema for this purpose.

This would take some design effort, and that's why getting a few
database "gurus" who are also familiar with 78 discography to get
involved. I actually have ideas for the XML schema which needs to
codify the "ontology" we decide upon (and I have ideas for the
ontology underlying this all -- basically a two-part structure, one
focused on the metadata associated with a recording session, and the
other on each "recorded artifact" the session produces)

Anyway, learning about wiki is a good thing, if for no other reason to
understand the principles of what wiki enables.

*****


Jon Noring


[Subject index] [Index for current month] [Table of Contents]