Re: proposed changes to the EARL schema

Hi Shadi,

tor, 21,.04.2005 kl. 22.54 +0200, skrev Shadi Abou-Zahra:
> > Would it be an idea with an URI, ID or version tag that 
> > could be used to identify each EARL report uniquely? 
> > The assessment tool does not have to create that, but 
> > it would be nice to have some standardised way of 
> > assigning ID's and possibly versions for EARL reports 
> > that are collected from tools when the RDF data is 
> > going to be stored in a larger triplestore. Another way 
> > might be to assign URI's like Annotea does.
> > 
> > E.g. using <rdf:ID> and <dcterms:hasVersion> 
> 
> You get that for free from the uniqueness of URIs. I added rdf:ID to
> each earl:assertion so that each assertion is unique *within* a report.
> This is in the control of the tool producing the EARL. After that, each
> document has a unique URI and the assertion is only a fragment therein.
> Examples:
> 
> http://example.org/my-reports/report123#assertion666
> file://c:/documents/my-reports/report123#assertion666

I was thinking of a slightly different use case. I was considering the
case of using a persistent triplestore, like 3store, Jena or Kowari for
storing testsuites or test runs that keeps a run number and several EARL
test reports contained within the test suite/test run. Each test
suite/test run would have to be unique and versioned. I was not thinking
about storing files, but an RDF tree or grove of tests and tests...

I do not see how the EARL report can identify itself in this scenario?

Regards,
-- 
Nils Ulltveit-Moe <nils@u-moe.no>

Received on Thursday, 21 April 2005 21:55:36 UTC