Alexander Johannesen wrote:
> And *seriously*, no
> one should need to know how meta data is recorded in a MARC record,
> not even catalogers. Seriously.
So goes your theory, and outside the real world, I agree.
But the real world needs to change first to allow this to happen.
Don't get me wrong, I agree with most anything you are saying,
but lots of it is pipedreams right now, to a much larger extent
than you seem to realize.
(OTOH, sure, this forum _is_ a place for pipedreams.)
But what communicative device do you suggest for catalogers?
I mean, they do need to talk about what they are inputting
when and how and in which fields or elements or attributes
or what have you. These things need to have names, and they
need to be brief and precisely defined and internationally
understood. Present your concrete idea for this, something to
look at, present an alpha version of a functioning model for a start
or to pluck apart.
> This is about models and semantics,
> *not* about formatting.
>
For now and the foreseeable future, there's not the option to just
switch to new models and semantics, in the world as it is. We need
to get there, that's right. This is the NGC forum, and that
means we are not supposed to perpetuate past errors. The models
and semantics, however, that have been presented so far are
obviously either not convincing enough in themselves or couldn't
make their way into convincingly operable designs. Or else we'd
already have the overwhelming success of an XML based bibliographic
engine that blasts all legacy designs aside. I'm not saying it is
impossible to construct one based on XML, but I'm not enthusiastic
either.
B.Eversberg
Received on Tue Aug 28 2007 - 03:56:22 EDT