- From: Paul Groth <p.t.groth@vu.nl>
- Date: Thu, 27 Sep 2012 21:38:09 +0200
- To: "Freimuth, Robert, Ph.D." <Freimuth.Robert@mayo.edu>
- Cc: public-prov-comments@w3.org
Dear Robert Thank you for your comment. Below is the suggested resolution. Please let us know if you are fine with it. You can find any suggested changes in the latest editor's draft at http://dvcs.w3.org/hg/prov/raw-file/default/model/prov-dm.html Thanks, Paul ISSUE-504 (collection/bundle) "A bundle (2.2.2) is a set of provenance descriptions, and a set is a type of collection (2.2.3). Therefore, it follows that a bundle is a specific type of collection (where the members of the collection are called "descriptions", as in section 5.4). This should be stated explicitly. Furthermore, since a bundle is a specialized collection, the spec would read better if their order in the document was reversed." Original email: http://lists.w3.org/Archives/Public/public-prov-wg/2012Sep/0094.html Tracker: http://www.w3.org/2011/prov/track/issues/504 Group Response - It is correct that: A bundle is a named set of provenance descriptions (2.2.2). It is also correct that section 2.2.3 indicates that many types of collections exist, including sets. However, section 2.2.3 states: A collection is an entity that provides a structure to some constituents, which are themselves entities. - In PROV, provenance descriptions are not given identifiers and are not regarded as entities. Identifiers occurring in provenance descriptions denote "things in the world" (resources). - To be able to talk about the provenance of PROV descriptions, the bundle construct allows a set of descriptions to be named, and become a "thing in the world". Such a bundle is an entity whose provenance can then be described. - In conclusion, a PROV bundle is not a PROV collection. References: - Issues on identifiers: http://www.w3.org/2011/prov/track/issues/183 http://www.w3.org/2011/prov/track/issues/215 Suggested change: update the definition of collection as follows: A collection is an entity that provides a structure to some constituents that MUST themselves be entities.
Received on Thursday, 27 September 2012 19:38:41 UTC