Monday, July 9, 2012

How much structural symmetry is there between the Affiliation and the Connection?

This question has been occupying my mind for some time, and it relates to the way some Elements in Clique Space(TM) appear themselves, to exhibit precisely the properties that Clique Space was conceived to model. How far can I go with this? Can a Clique Space Agent Device use its data model to model relationships between its own Clique Space component object instances as they are created and deleted?

This question is a rather silly reductio ad absurdum. However, it might be interesting to note that discussions throughout this blog have touched on something I call the serving Agent Device's Clique.

Let's have another look at the serving Agent Device's Clique. I am in the process of removing the "three stage delegate" device Connection mechanism from the implementation, and working toward a new relationship between two of the implementation's components: both the Connection to the other device and the serving Agent Device's Clique modelling the collaborative exchange between an Agent Device and the other device must exist simultaneously. Delegate capacity (a capacity for the Element to contain and project an Agent Device's delegate server RMI stub to the other device through which commands will be sent back to the Agent Device from the other device) is removed from the Identity because a singular Identity will now be correctly implemented to represent collections of Connections and Affiliations around a single Axle. The delegate capacity will be shared between the Connection and the Participant: the Connection contains the RMI stub, but any message sent from the other device identifies the serving Agent Device's owner Participant created from this other device's Connection when instructing the Agent Device to perform some action in relation to a change in the other device's state.

So anyway, what am I getting at here? Well, put the delegates and all the mechanical minutiae for the serving Agent Device's Clique aside (I feel I had to explain it in part just to get it off my chest), and concentrate merely on what the Agent Device's Clique represents: a collaborative exchange between two users; the operator of the serving Agent Device, and the operator of the device which is connected to a particular Clique Space through this Agent Device, if these are indeed different individuals. While it should be obvious to the reader that the serving Agent Device's Clique models the collaboration between an Agent Device and another device (because I have just stated this), the same reader might also be ask themselves whether, because of the dependency between the serving Agent Device's Clique and the other device's Connection, this Clique appears also to represent the instance of the Connection to the given device.

I think such an observation is very interesting, as it appears to promise a similar symmetry in the relationship that manifests an Affiliation. More thought is definitely required before I start to head off in a specific direction here,but one question is this: who is collaborating with whom? Clearly, in the serving Agent Device's Clique, the two Participants are representing the operator or operators of the devices which are collaborating. However, consider this: where the Connection represents an association between an Identity (more accurately, the Identity's Axle through a given Identity) with a Media Profile, then who, if this association might also be seen as a collaboration, represents the component Media Profile? Likewise, if we are going to do something similar with the Affiliation, then who represents the Account Profile?

When we entertain this "structural symmetry", might we in actuality require two more new Cliques (rather than one because the symmetry in the Affiliation bares no functional relationship to the serving Agent Device's Clique) to model the structural symmetry: 1. one Clique to model the specific relationship between the Connection's Axle and Media Profile and 2. another Clique to model the specific relationship between the Affiliation's Axle and Account Profile? Would these Cliques have any natural value? I think they may...


1 comment:

  1. Thank god those Russian spamming fuckers have given up!