[...]
Just a briefish comment on the 'email addresses change and are therefore
not much use' perspective. If the data stored makes clear the
time-period during which that address was supposed to hold, we can use
the combination of period and emailaddress to do lookups in third party
services. This could either be through adminstrative metainformation or
the semantics of the data element(s) storing the email address itself.
So, if I write some Web-accessible document while an undergraduate at
university of bristol, and we record {creators email address, creation
date} accordingly, this sets us up for future historians to figure out
who the person was who had that email address on that date. There would
be no requirement on maintainers of metadata for that document to try to
keep up with my various subsequent addresses. All they would be required
to do would be ensure unambiguity regarding perido when the address was
considered valid.
For example, I could now write into a publically accessible database the
fact that from 1992 1995 I had an email address of
[log in to unmask] and that from 1995 to 1999 I've been
[log in to unmask] A record describing a document created
in 1994 ascribing an email address of [log in to unmask] to the
creator could, with a little care, be telling us enough to use 3rd party
information to track down current email details.
Dan
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|