Cross-posting Dave's response from oslc-core group to the oslc-cm list Thanks, Steve Speicher | IBM Rational Software | (919) 254-0645
Dave wrote on 03/05/2010 02:03:31 PM: > > What are we trying to accomplish with the dc:creator and owner fields? > > Are we trying to solve a distributed identify problem or simply trying > to use the right model for givenName, familyName, emailAddress type > information? I believe we should do the latter in the 1st version of > the Core spec, but I agree that we may also need to do the former at > some point. At this point we have the notion of a user account, but no > notion of a "live person identifier" and I think we have to live with > that at least for v1. > > For simple username information the FOAF fields are sufficient and I > believe a foaf:Person resource should be in the Core spec along with a > set of person properties. Actually, there are in the draft now. > > In my opinion the Core spec should define a set of "core properties" > and the datatypes for each (e.g. foaf:Person for dc:creator), but that > individual specs should be free to override these on a case-by-case > basis. > > - Dave > > > > On Fri, Mar 5, 2010 at 8:12 AM, Steve K Speicher <[email protected]> wrote: > > Thought the core WG should weigh in on this one. Olivier has a fair > > amount of experience with this and some good points to consider. > > > > Comments? > > > > Thanks, > > Steve Speicher | IBM Rational Software | (919) 254-0645 > > > > ----- Forwarded by Steve K Speicher/Raleigh/IBM on 03/05/2010 08:10 AM > > From: > > Olivier Berger <[email protected]> > > To: > > "[email protected]" <[email protected]> > > Date: > > 03/05/2010 05:43 AM > > Subject: > > [oslc-cm] foaf:Person vs sioc:User for dc:creator or oslc_cm:owner ? > > > > I've had a look at > > http://open-services.net/bin/view/Main/CmResourceDefinitionsV2 and it > > seems that foaf:Person is preferred at the moment for describing > > reporters/owners of CRs. > > > > I'm not sure this is the best choice, and would propose to use sioc:User > > instead. > > > > When reading : > > "General agreement on foaf:Person is enough as there are not too many > > systems that have multiple accounts per user. Can be managed by having > > separate person instances (exception, instead of the rule)" > > I'm puzzled... > > > > sioc:User is already there to manage the situation of several accounts > > (sioc:Users) of the same physical person (foaf:Person). > > > > Why chose the least suitable one when there's one more flexible ? ;) > > > > Is this to avoid dependency on sioc and limit to foaf ? > > > > foaf:Person seems well suited when dealing with real physical persons, > > whereas sioc:User is more about accounts/aliases/nicks of potentially > > anonymous personalities of these real people. > > > > Also, when thinking about automation, I believe dc:creators couls be > > robots/apps, where sioc:User may be preferred to foaf:Persons too ;) > > > > In any case, sioc seems more actively maintained than foaf, IMHO, so it > > may be preferred maybe. > > > > Any comments ? > > > > Best regards, > > > > P.S.: I'm in the process of drafting a forge ontology : > > https://forge.projet-coclico.org/plugins/mediawiki/wiki/wp2/ > index.php/Forge_Ontology_Proposal > > which should be interoperable with OSLC-CM, hence my recent investigatiosn > > of sioc vs foaf. > > -- > > Olivier BERGER <[email protected]> > > http://www-public.it-sudparis.eu/~berger_o/ - OpenPGP-Id: 2048R/5819D7E8 > > Ingénieur Recherche - Dept INF > > Institut TELECOM, SudParis (http://www.it-sudparis.eu/), Evry (France)
