- From: Roger Menday <roger.menday@uk.fujitsu.com>
- Date: Sun, 10 Feb 2013 22:55:26 +0000
- To: John Arwe <johnarwe@us.ibm.com>
- CC: "public-ldp-wg@w3.org" <public-ldp-wg@w3.org>
- Message-ID: <B62836B4-AF07-4EE2-89B2-E34F640112A3@uk.fujitsu.com>
hi John, > Roger I am unable to map from mine to yours. > Just too much ambiguity. Perhaps if you showed the differences between what you call by-value and by-reference in terms of HTTP flows for container interactions then all would become clear. > A representation of a resource includes links to other resources. Writable LD should allow for additional links (i.e. triples) to be added (and deleted). So, this is add by-ref (i.e. to an existing resource). Conversely, growth of the graph of resources takes place with contextualised creation; a resource creates new resources. The creating resource links to the newly created resource. This is by-value link addition, where link creation is preceded by resource creation. That the difference between by-ref and by-value, as I see it. For the moment, I have not showed this in terms of HTTP flows for container interactions, because, well, we seem unclear about the specific nature of our containers, and I'm not sure which container model you want me to use :) Actually I think we are having this long discussion about how composition and aggregation co-exists because we are forcing composition semantics on resources and sub-ordinate resources. But, I don't think LD works that way. However if our containers contain *links*, then these links are properly subordinate - a resource *composed* of its links - and if a resource is deleted, it's outgoing links are also deleted. For me this is a much cleaner model. regards, Roger > Best Regards, John > > Voice US 845-435-9470 BluePages > Tivoli OSLC Lead - Show me the Scenario > >
Attachments
- application/pkcs7-signature attachment: smime.p7s
Received on Sunday, 10 February 2013 22:56:44 UTC