[tdwg-tapir] tapir namespace

[Steven Perry]

I'd really like to see the namespace URIs act as URLs that point 
directly to the actual schema or ontology.  W3C is inconsistent with 
this, sometimes the namespace URL refers to an overview HTML page and 
sometimes to the schema itself which makes looking for W3C standards 
very difficult.

One reason I'd like to see namespaces point to schema is that it makes 
it easier to verify and diff versions.  Also, if we switch to 
ontologies, RDF does not require the schemalocation attribute, in fact 
it doesn't even require serialization to XML.  With RDF the accepted 
practice is to use the namespace URI as the location of the ontology.  
Many RDF software tools depend upon this.

One thing to keep in mind is that the TDWG webserver should be set up to 
serve things like ontology.rdf and schema.xsd as index pages and it 
should know to set the content type to application/xml for both.

If we go this route, then the revision information for a schema or 
ontology cannot be captured in it's filename.  Instead it must be in the 
directory name.

-Steve



Roger Hyam wrote:

>Hi All,
>
>Ricardo and I just discussed namespaces and schema locations without 
>coming to a conclusion. TDWG needs a namespace and schema location 
>policy and we need to reserve a space for these as we put the 
>infrastructure in place.
>
>There are a host of issues to resolve including how we do resources in 
>ontologies, what happens if you try and resolve a namespace URI - do you 
>get a page not found, a generic message saying it is a namespace or 
>specific message or a redirect etc. Should the schema location be the 
>same as the standard location or embedded within it. Also we need to 
>have 'permanent' URLs for standards themselves. Oh and what happens if 
>we change the name of the organisation as people bring up from time to time!
>
>Some of the issues are closely related to GUID stuff so I am hope that 
>we can have a very brief discussion about it at the GUID meeting over a 
>beer and just decide on a load of these things. Many things are more or 
>less arbitrary and a decision just needs to be taken - any decision.
>
>Can you wait till February for a final policy on what the namespace URIs 
>will look like?
>
>I suggest namespaces will be of the form
>
>http://tdwg.org/namespaces/tapir/what_you_like
>
>(note no www as it is not a web thing. We could go for namespaces.tdwg.org/tapir ...)
>
>and schema stuff might be like this:
>
>http://www.tdwg.org/standards/tdwg_std_tapir_20060130/schema.xsd
>
>My big question is: Will the schema change without the standard changing?
>
>There would also be other related to the standard at:
>
>http://www.tdwg.org/standards/tdwg_std_tapir_20060130/
>
>including the normative documentation and a cover page xml document with 
>metadata for the standard.
>
>Does this make sense? I'd appreciate feedback as I am actively working 
>on the specification for all this stuff.
>
>Thanks,
>
>Roger
>
>
>Döring, Markus wrote:
>  
>
>>Hi,
>>another thing is the namespace of TAPIR.
>>Currently its http://www.tdwg.org/schemas/tapir/datasource/1.0
>>
>>Are there guidelines from tdwg for this?
>>Do we still need the "datasource" being part of the url? Initially this was done to indicate that this protocol only specifies the communication to/from a datasource and not between portals, aggregators, proxies or relays, ...
>>But are we still thinking of creating a family of TAPIR protocols?
>>
>>Personally I am quite indifferent. I would be happy with just a short one such as:
>>http://www.tdwg.org/tapir/1.0
>>
>>Any opinions on this?
>>
>>Markus
>>
>>
>>_______________________________________________
>>tdwg-tapir mailing list
>>tdwg-tapir at lists.tdwg.org
>>http://lists.tdwg.org/mailman/listinfo/tdwg-tapir_lists.tdwg.org
>>
>>  
>>    
>>
>
>
>  
>