[Geodata] evaluation requirements for modern metadata catalogs

[Norman Barker]

Hi,

having been trying to do this catalog thing for a while now (both the
minimal metadata and the ebRIM model) and I am slightly concerned that
as a geodata group we don't know what we want, and perhaps there is a
certain amount of Java/OGC/ISO/Python/REST/SOAP/OAI bashing (pick your
camp depending on your persuasion!)

We have requirements for minimal metadata (dclite4g) which is great,
we also have the concept that we want this catalog to be 'out there' -
also great.  I like the general discussion as well.

However I don't see use cases, the kind of thing I am looking for is
(simplistic example)

I want to use a catalog and I want to do the following

1) search on bbox
2) search on keywords
3) discover services offering data matching (1) and (2)

I think if we could come up with evaluation criteria / use cases we
could get past some of the implementation specifics when it comes to
saying if something is right or not or whether it matches needs (since
'right or not' is a bit stark).  If these could go on the wiki then
great; I am a developer I shouldn't be providing use cases, these
should come from users.  If I provide a use case then it is going to
centred around my expertise bias.  I am sitting here now thinking my
catalog is kind of cool, geonetwork is also kind of cool, however
others (rightly) are saying I don't think so, and there is no use case
benchmark.

Just my thoughts, flame me if you want, it's gmail :-)

Norman