On 1/27/09 10:57 AM, "Karen Coyle" <[log in to unmask]> wrote:

> There's also the fact that validation must be built into individual
> systems, since there is no schema or dtd against which everyone can
> validate the MARC record. Thus, every system validates somewhat
> differently (and some not at all). Although MARC is a machine-readable
> record, we don't have a machine-actionable standard for our data. This,
> I must admit, just boggles the mind in this day and age.

Can somebody say "MARCXML or MODS complete with a schema"? Such solutions
offer at least syntactic validation if not also semantic validation. Oh
well.

-- 
Eric Lease Morgan