Simon Spero said:

>If Mac finds  this weaker definition  is still underinclusive, then it
>would be helpful if he could explain what he understands by the term
>"linked data"; if it is sufficiently inclusive, it would be helpful if he
>could describe how UTLAS implemented naming and dereferencing, and what
>communications network was used.

For me, "linked data" means a "link" replacing "data".

Utlas used their own software.  LC and LAC authorities, along with
client created authorities, were liked to MARC records.  Whether the
linked data is stored inhouse or remotely doesn't, it seems to me,
affect the situation.  Many have said that remote data would be stored
inhouse.

OCLC considered buying the Utlas software, but it was said they found
it too complex, and they were unwilling to offer the range of products
(which then included microform and print catalogues, as well as cards
and MARC records).

Earlier I'm told LC considered it, but rejected it on the grounds
that is was not US.

Canada was simply too small a client base for so ambitious a project.  
Other utilities and software have yet to reach their sophistication.  
For example, they had "nesting" ASNs for headings plus free floating
subdivisions for subject strings which had not been established as
strings.

When "auth"ed a heading, the whole string was checked; if no match,
subdivisions were exempted from the check one by one until a match was
found, and ASN substituted for the text.



   __       __   J. McRee (Mac) Elrod ([log in to unmask])
  {__  |   /     Special Libraries Cataloguing   HTTP://www.slc.bc.ca/
  ___} |__ \__________________________________________________________