Hi Bruce,
I agree that ISO TC 211 do not have the resources to host a profile registry. I also agree that an organisation, willing to host a profile registry, is likely to be the solution. However, this organisation must be endorsed by TC 211 for it to have any authority to host such a registry.
Ideally a class A TC 211 liaison organisation, such as OGC (hint, hint, Mark? Carl?), should host a profile registry. This would make the endorsement by TC 211 easier or more relevant.
Alternatively, each country could host their own profile registry and then an official TC 211 endorsed web site (or preferably a registry) could list these endorsed profile registry web sites. This would require the hierarchical registry model. Again, a class A TC 211 liaison organisation, such as OGC (are these hints subtle enough ;--) ), could provide the central registry that registers the profile registries and passes searches onto them. Hence OGC, on behalf of TC 211, would provide the resource to search all registered profiles and OGC would only need to register the national registries and not each and every profile register. Less work for OGC but still a lot of recognition. ;--)
I think all this is ethnically doable. The only restriction is the incentive and resources for organisations to create and register these profile registries.
An interim solution may be to register the metadata for profiles on Catalogue Services - Web (CSW), eg. GeoNetwork, apply a hierarchyLevel of "profile" and then link from the metadata to the actual profiles. These CSWs could then be registered on an officially endorsed by TC 211 CSW and then distributed searches would allow people to find the metadata - and hence the data - of the world profiles by using a search where hierarchyLevel='profile'.
This does not address how profiles are endorsed by TC 211. I have seen some basically wrong profiles because they do not follow the TC 211 standards and specifications. Someone, maybe a TC 211 volunteer group, will need to check that the profiles meet the TC 211 requirements before they can be accepted.
Anyway, this is not likely to happen because of a few emails. We need to lobby the appropriate organisations etc.
Thanks.
John
-----Original Message-----
From: Westcott, Bruce [mailto:bruce.westcott@anonymised.com]
Sent: Wednesday, 3 February 2010 7:22 AM
To: Hockaday John
Subject: RE: [metadata] [GeoNetwork-devel] identifying ISO
metadata profiles [SEC=UNCLASSIFIED]John,
The idea of an ISO-sanctioned registry is a complete
no-brainer, but my
take on it is that they don't have the mechanisms and scope to put an
actual service on the ground. I suppose they are just awaiting some
public-spirited member/country coming forward to volunteer to do so.
The need for a registry of profiles and extended elements is something
perhaps we should push some authorities do put some resources
into.
<snip>
Bruce Westcott
Geospatial Metadata Consultant
07536-525307 -- mobile
01494-882180 -- office
High Wycombe, Buckinghamshire - UK-----Original Message-----
From: John.Hockaday@anonymised.com [mailto:John.Hockaday@anonymised.com]
Sent: 02 February 2010 00:09
To: steve.richard@anonymised.com
Cc: metadata@anonymised.com;
GeoNetwork-devel@lists.sourceforge.net
Subject: Re: [metadata] [GeoNetwork-devel] identifying ISO metadata
profiles [SEC=UNCLASSIFIED]On Monday, 1 February 2010 5:16 PM Stephen Richard wrote:
> On 1/31/2010 4:05 PM, John.Hockaday@anonymised.com wrote:
> > Hi All,
> >
> > The ANZLIC Metadata Profile version 1.1 uses the
> MD_Metadata namespace.
> I assume you mean http://www.isotc211.org/2005/gmd?Yes.
>
> > This is because the ANZLIC Metadata Profile does not add
> nor exclude any ISO 19115 metadata elements. What it does do
> is make the MD_Metadata/fileIdentifier mandatory. This is
> checked in the ISO 19139 XML using Schematron rules. Hence,
> in the matter of compliance, an ANZLIC Metadata Profile XML
> metadata record is the same as an ISO 19139 XML metadata record.
> >
> They both conform to the same xml schema, but I'm assuming there are
> conventions in ANZLIC that are useful for a metadata consuming
> application to know. How does an ANCLIC metadata document
> tell the user
> that the metadata conforms to that profile?As mentioned, by validation using the ANLZIC Schematron.
> > Regarding the metadataStandardName. I believe that ISO
> 19139 should add an attribute to this element of type URL
> that allows a hypertext link to the actual XML implementation
> of the profile identified by the metadataStandardName and the
> metadataStandardVersion. This will allow users to find the
> validation XML - XSDs, Schematron etc. - and the profile
> documentation for the profile.
> >
> like all controlled vocabulary used in xml, I'd like to see it
> implemented as a scoped name with an identifier for the containing
> vocabulary, an identifier for the concept in the vocabulary, a
> language-localized label for the concept, and a URL for a definition
> service or other resources... (interoperability requires a
controlled
> vocabulary of metadata standards....)I believe a nicer solution would be an official ISO TC 211 endorsed
registry of registered profiles. This will allow users to find all
approved existing profiles to see if something exists that meets their
needs before defining yet another profile. If an existing profile only
partially meets their needs then they could extend the
existing profile
and register that extension at the official ISO TC 211
endorsed registry
of profiles.John
>
> steve
>
> > My two cents worth. ;--)
> >
> >
> > John
> >
> >
> >
> snip
>