RE: synonym anchors, in RDFa

This is what terrifies me. It's why I want the spec (or an appendix) to give certain "model cards", all laid out (ready for cut and paste). 

 

An example with several keys, for one #fragment-id. 

AN example with several fragment-ids for 1 modulus. 

An example in which different fragment-ids (all onto 1 mod) each bind to a different object class.

 

 

When I look at http://morph.talis.com/?data-uri%5B%5D=http%3A%2F%2Fyorkporc.blogspot.com%2F&input=&output=turtle&callback=, Im even more terrified. I thought I was making 3 or 4 statements. But, look at what someone thinks Im saying.Ive no idea what impact all that will have on the poor webid validator. As someone said, earlier, perhaps i just been damn lucky, so far.

 

I have not dared do the same turtle rendiition of my proxy profile, at ODS. But, I think we are ending up with something I understood (once): the notion of subobjects in X.500. Different managament entities would manage a different set of classes and the instances and their properties, and the containing-object would be in effect a mashup of the several files. The webid proxy profile is similar, and a bit more flexible, since the subobjects can be in entirely different namespaces and on different servers.They dont have to be in the same x.500 naming domain == ActiveDireectory partition, or under the same Adminsitrative management domain.

Received on Friday, 23 December 2011 20:32:08 UTC