- From: Myers, Jim <MYERSJ4@rpi.edu>
- Date: Mon, 27 Jun 2011 12:49:03 -0400
- To: Paolo Missier <Paolo.Missier@ncl.ac.uk>
- CC: <public-prov-wg@w3.org>
> Creating names for things in a given state is fine. However, I feel that at > some point you will have to contend with the issue of "what has changed" > between two things. It may not be the values of some properties, but surely > somehow you have to account for the nature of a change? Absolutely, though I'm not sure how far PIL can go there in the sense that I'm not sure what else to add to PIL that would not be domain specific - beyond the thing-process-execution-thing framework and IVP to shift between levels (and handling agency, etc.). If PIL allows you to identify things and state their properties (including perhaps a link to their external object type), someone with domain knowledge would be able to understand the nature of the change. You probably need to understand the nature of the classes of things you're talking about, as well as the nature of the process (understanding its recipe, which we've also tried to keep out of scope (just the link to recipe is in)) to understand the nature of the change. At the PIL level, is it enough if the answer is that the "what has changed" is the aspect/state of stuff represented by the invariant properties? Jim
Received on Monday, 27 June 2011 16:50:16 UTC