This will probably bring some heat, but a guy's got to do what the guy's
got to do...
I've noticed a lot of qualifier sets incorporating sub elements that appear
to be designed to encompass metadata that does not belong to the resource
in hand. For example, email addresses for agents. In a perfect world we
would be strongly discouraged from putting email address, which really
belongs to the agent, in the metadata of the resource. It will create
havoc when email addresses change. (For example, multiple occurrences of a
now outdated email address. For you fellow former and/or peripheral
techies, it's the old data management saw about redundant data.)
In that perfect world (the one in which we don't live), the email address
would be stored (along with the rest of the metadata that describes the
agent) in a separate record creating a single central point for reference
and maintenance.
Just some fodder [sic] for discussion...
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|