[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [Elist Home]
Subject: Re: [xtm-wg] Issue with instanceOf
Oops. I wish to make a subtle but important disclaimer. *Only* for the sake of improving the understanding of the processing model, I wrote: > ...if a topic were to somehow lose one of its memberships > in some association, perhaps because the association is > deleted, its memberships in various scopes would probably > be completely unaffected. I say "probably" rather than > "certainly" because there is a possibility that the > deleted association might have the topic in its scope, and > it might be the only association in that scope, so the > s-node might go away, too. In that sense, then, the > question of the existence of an association between topics > might also be the question of the very existence of an > s-node. This sounds as if I was describing a processing requirement for the Spec itself, by implying that there are occasions when s-nodes have to be deleted. There is no such processing requirement. Indeed: * There is no requirement that nodes be deleted just because they aren't connected to other nodes in certain ways. * It is even possible that, as implemented, a graph can always contain all the s-nodes that are possible, given the number of different topics that exist, regardless of whether they are actually used as the scopes of any a-nodes. There is no reason why the Spec needs to constrain the behavior of implementations in such a way as to require that only the actually-used s-nodes are allowed to exist in a conforming topic map graph. This is an example of a trade-off that implementers, and only implementers, should be allowed to decide about: is it better (a) to limit the number of s-nodes in memory to those that actually serve as the scopes of a-nodes, or (b) to minimize the amount of work required to manage the graph's inventory of s-nodes by permitting unconnected s-nodes to hang around? In general, we must avoid placing any unnecessary constraints on implementations. I was not trying to express or create such a constraint in the passage quoted above. I was just trying to clarify part of the model, and, in process, I'm afraid I created muddiness elsewhere. -Steve -- Steven R. Newcomb, Consultant srn@coolheads.com voice: +1 972 359 8160 fax: +1 972 359 0270 405 Flagler Court Allen, Texas 75013-2821 USA To Post a message, send it to: xtm-wg@eGroups.com To Unsubscribe, send a blank message to: xtm-wg-unsubscribe@eGroups.com
[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [Elist Home]
Powered by eList eXpress LLC