Reporting Specification Issue Tracking

Add new issues to this page.

I was expecting to see a notion of identity required of the artifacts which are being reported upon. My motivating example is a capability to perform trend analysis over time - with repeated artifacts identified - for example, 10% open defects fixed since last time.

When paging over large datasets, is it a requirement that no artifact be included more than once (I think "same" touches on identity.)

Is there any guidance or specification around consistency across pages?

-- IanGreen - 13 Aug 2009

The Insight Reportable REST specification has the notion of metadata=schema query filter string that will return the XML Schema for a resource URL. One thing I believe we should consider is using RDF Schema instead of XML Schema. XML Schema tends to be closed in, in the sense that it is possible but not easy to specify that XML elements not recognized by a particular processing agent are legal and should just be ignored if not understood. RDF Schema has it's own multi-inheritance like facility that associates properties with classes, but does not restrict the properties to a specfic required and/or limited set.

The advantage of such an RDF Schema approach is that if there is a general RDF class of resource, say BugReport? , that has a general RDFSchema associated with it, individual bug reporting systems can freely associate new attributes with their implementation of BugReportForMySystem? . Any processing agent that can deal with BugReport? should be able to deal with either BugReport? or BugReportForMySystem? instances.

-- StevePitschke 8 Oct 2009.