- From: Graham Klyne <GK@ninebynine.org>
- Date: Fri, 25 Nov 2011 12:20:23 +0000
- To: Paolo Missier <Paolo.Missier@ncl.ac.uk>
- CC: public-prov-wg@w3.org
I have a nagging worry that if the Primer becomes dominated by the journalism example, we'll lose the concise but readable summary that is (IMO) invaluable for developers as both introduction and handy reference. The PROV-DM document doesn't really provide this IMO because it's very concerned with the details of the ASN and model constraints. The primer currently could serve this role (e.g. as the OWL primer does for OWL), but I fear that if it is completely structured around the example, it will become more of a "painting-by-numbers" guide. I'm not necessarily saying that the journalism example should not be present. But I really like the focused and progressive introduction of example material which to my mind really backs up the Primer's role as a primer. #g -- On 24/11/2011 14:28, Paolo Missier wrote: > Hi, > > just to pick up on Khalid's comments, which I agree with: whatever happened to > the journalism example? I would indeed have imagined the primer would be the > right place for it. > > I also agree that derivation should be pushed up as it is so central for provenance > > Regards, -Paolo > > > On 11/24/11 10:09 AM, Khalid Belhajjame wrote: >> >> Hi, >> >> Here are some comments on the primer. >> >> My main concern is with respect to the structure. I find that Section 2 [1] >> delves into the definitions without giving the reader a chance to have a rough >> idea on what the elements of the models are and how they are related to each >> other. In this respect, PROV-DM, for instance, illustrates in Section 3 an >> overview of the model [2]. I am wondering if something similar can be done in >> the primer. Probably not using the names of the constructs in the model, but >> rather using the newspaper example elaborated in detail in Section 3, which by >> the way I think it is nicely articulated. >> >> In Section 3, turtle is used for encoding the examples. It looks fine, but I >> am wondering if it would be better to use instead the examples listed in >> Appendix A, i.e., using the abstract syntax. In that case, the turtle examples >> may be moved to he Appendix. >> >> I like the order in which the concepts and relationships were introduced. >> Derivation is however left till the end. I think that concept is quite >> important when talking about provenance. Would it be better to promote it by >> placing the definition after, let’s say, Use and Generation.? >> >> In Section 2.3 [3],it is said that “Every Entity is created by an activity, >> which is called the generation of the entity”. Does that statement always >> hold? In other words, are all entities the results of an activity? >> >> Thanks, khalid >> >> [1] >> http://dvcs.w3.org/hg/prov/raw-file/default/primer/Primer.html#intuitive-overview-of-prov-dm >> >> >> [2] http://dvcs.w3.org/hg/prov/raw-file/default/model/ProvenanceModel.html >> >> [3] >> http://dvcs.w3.org/hg/prov/raw-file/default/primer/Primer.html#use-and-generation >> >> > >
Received on Friday, 25 November 2011 12:37:36 UTC