On Thu, 20 Mar 1997, Stu Weibel wrote:
> The Canberra meeting resulted in extensive discussion of two alternative
> syntaxes:
>
> The Knight-Hamilton Syntax
>
> The Extended Dot Syntax
>
> The latter had many more adherents than the former, but certainly the
> former has strong points in its favor as well. Both are short term
> fixes that we will outgrow or adapt.
>
> I see no reason why these two forms cannot coexist in the current
> environment. Deploying either is an improvement over the current state
> (deploying neither).
At UKOLN we are currently planning to hold metadata in separate database
(probably using SOIF records, though the actual format is not particularly
important). HTML embedded metadata will be generated on the fly using
server-side includes as pages are served.
Using this approach we can reasonably easily accomodate META syntax
changes as they occur and will also be able to move to generating embedded
XML or PICS labels as and when the methods for doing that are agreed
without leaving ourselves with large numbers of static HTML pages that
need updating. (We could even consider generating different formats of
metadata or none at all depending on who or what the client is).
There are some interesting problems with this approach, not least how one
links together an object with it's metadata - we are considering the use
of true identifiers, DOIs for example, for this purpose.
Andy.
--
UK Office for Library and Information Networking
University of Bath, Bath, BA2 7AY, UK Voice: +44 1225 323933
http://www.ukoln.ac.uk/~lisap/ Fax: +44 1225 826838
|