On Sat, May 25, 2013 at 1:21 AM, Christian Mollekopf <[email protected]>wrote:
> On Saturday 25 May 2013 00.04:20 Vishesh Handa wrote: > > On Fri, May 24, 2013 at 5:11 PM, Christian Mollekopf > > > > <[email protected]>wrote: > > > On Thursday 23 May 2013 19.44:04 Vishesh Handa wrote: > > > > Hey guys > > > > > > > > We seem to have run into a situation where we need to fetch some > > > > > > identifier > > > > > > > of a nco:ContactMedium, and we noticed that nco:emailAddress does not > > > > derive from nao:identifier. > > > > > > > > whereas nco:imID does. > > > > > > > > If no one object, I can add the required super property. > > > > > > Note that an email address is not actually a unique identifier. Many > > > people can > > > share email addresses (family account, company account, ...), so this > > > breaks > > > the "Inverse Functional" [0] property this property is supposed to have > > > according to [1]. Maybe there are other options? > > > > It is still inverse functional. > > > > I have confuse nco:emailAddress for nco:EmailAddress... > nco:emailAddress is of course inverse functional for nco:EmailAddress but > not > for nco:Contact (which I confused with nco:ContactMedium). > > +1 for nao:identifier for nco:emailAddress. > > You can ignore the rest of the mail if everything is clear. > > Cheers, > Christian > > > > As I understand [1], inverse functional means for: > > <nepomuk:/resA> a nco:Contact ; > nco:hasEmailAddress <nepomuk:/res/email> . > > that whenever I see > > X a nco:Contact; > nco:hasEmailAddress <nepomuk:/res/email> . > > X MUST be the same as <nepomuk:/resA>. > > So i.e. a UID would be an inverse functional property which would qualify > as > nao:identifier. Also, the nco:emailAddress (the string), is an inverse > functional property of nco:EmailAddress (the resource). > > > If multiple contacts want to share an email address then all of them can > > link to the same nco:EmailAddress. I do not think we need a separate copy > > of the nco:EmailAddress for each contact. > > > > Agreed, we don't need nor want a separate copy for each contact. But the > email > address doesn't identify the contact. > Yeah. It does not. Trueg and I had a long discussion about this. I don't like the fact that it doesn't. > > > Something like this - > > > > <nepomuk:/resA> a nco:Contact ; > > nco:hasEmailAddress <nepomuk:/res/email> . > > > > <nepomuk:/resB> a nco:Contact ; > > nco:hasEmailAddress <nepomuk:/res/email> . > > > > <nepomuk:/resC> a nco:Contact ; > > nco:hasEmailAddress <nepomuk:/res/email> . > > > > > > <nepomuk:/res/email> a nco:EmailAddress ; > > nco:emailAddress "[email protected]" . > > > > --- > > > > If the Nepomuk identification works perfectly then this is what we > > currently have. > > > > > Cheers, > > > Christian > > > > > > [0] http://www.w3.org/wiki/InverseFunctionalProperty > > > [1] http://www.semanticdesktop.org/ontologies/nao/#mozTocId802441 > > > _______________________________________________ > > > Nepomuk mailing list > > > [email protected] > > > https://mail.kde.org/mailman/listinfo/nepomuk > -- Vishesh Handa
_______________________________________________ Nepomuk mailing list [email protected] https://mail.kde.org/mailman/listinfo/nepomuk
