From W3C Wiki
Revision as of 22:08, 20 January 2009 by DannyAyers (Talk)

(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Jump to: navigation, search


AWWSW Questions / Use Cases

  1. An application consumes an RDF statement, written by some other statement author. How should the application determine what other RDF statements should to be assumed in order to determine what the original statement "means"? -- DBooth [And "should" according to whom? -JAR]
  2. If a URI is dereferenced yielding a 200 response, what RDF assertions are implied by the 200 response? -- DBooth [JAR: This may relate to the "trust layer", in which case we need to identify who are the principals, what are they asserting, and are they authorized to have those assertions be made?]
  3. If a URI is dereferenced yielding a 303 redirect to a new URI, and dereferencing the new URI yields a 200 response, what RDF assertions are implied regarding the first URI? Similarly, if a URI contains a fragment identifier, but dereferencing the root (part before the fragment identifier) yields a 200 response, what RDF assertions are implied regarding the first URI? -- DBooth
  4. An application consumes an RDF document, written by some other statement author, containing a set of RDF assertions. The application dereferences one of the URIs in the statement, which yields a 303-redirect to a document containing a set of RDF assertions, but the assertions seem suspicious. Must the application use them? -- DBooth
  5. An RDF statement author is writing some RDF assertions, and wishes to use a URI u (minted by someone else) to denote a particular resource. URI u has been used in the past by other statement authors to denote this resource, but now, when u is dereferenced through a 303-redirect to its declaration, the assertions in its declaration clearly appear to be different than they were before, and would now be incompatible with the assertions that the statement author is writing. What should the statement author do? -- DBooth
  6. An application combines RDF assertions from two documents A and B, written independently by other statement authors, that both make assertions involving a URI u, minted by a fourth party who has published a third set of RDF assertions that are accessible indirectly via a 303-redirect through u. The assertions in A are consistent with the u's declaration, and the assertions in B are (separately) consistent with u's declaration, but the assertions in A are inconsistent with the assertions in B. What (if anything) is wrong? How can the application make use of both A and B? -- DBooth
  7. TimBL would like to know what triples to assert inside of Tabulator as a result of an HTTP interaction, in particular what to do about the various 30x status codes. These triples should allow tabulator code to deduce useful information for the user interface (such as what to offer an image view of, etc etc), and should allow one to discover what went wrong with access to an information resource, etc. -- TimBL
  8. People would like to store information about saved HTTP responses for persistent caches, so that information can be stored with a software installation to save the software looking well-known web pages; so that software looking up much-referenced pages can create a persietnet cache in real time and share it with other applications, etc. -- TimBL
  9. A theory of provenance of web artifacts (documents, parts of documents, RDF graphs, parts of RDF graphs) is of interest, in particular for scientific applications. One wants to be able to understand how text and triples come to be said, by whom or by what automated process. -- Stuart, JAR
  10. When a Semantic Web agent (a HTTP client or server) acts as a tool for a human agent in making assertions, how can one maximise the fidelity of the declared assertions against the intended assertions? -- DannyAyers