responses to Paolo's notes
authorLuc Moreau <l.moreau@ecs.soton.ac.uk>
Tue, 19 Jul 2011 00:02:28 +0100
changeset 4 78350c122eec
parent 3 ca92de105c08
child 6 d13b6804ac74
responses to Paolo's notes
model/ProvenanceModel.html
--- a/model/ProvenanceModel.html	Mon Jul 18 23:39:51 2011 +0100
+++ b/model/ProvenanceModel.html	Tue Jul 19 00:02:28 2011 +0100
@@ -99,6 +99,11 @@
 
 <div class="note">PM: propose to add a sentence on annotations to assertions, as follows. This WG does not prescribe the means by which assertions are made, for example on the basis of observations, or inferences, or other means. However, it should be possible for asserters to annotate any assertion with a description of the justification for that assertion, as well as additional meta-information, such as authorship of the assertion.</div>
 </p>
+
+<div class="note">Luc: For now, I would drop this sentence. We will populate this section later.</div>
+</p>
+
+
     </section> 
 
     <section> 
@@ -124,6 +129,10 @@
 
 <div class="note">This def. of situated entities is not clear to me. It relies on <em>state</em> and <em>situation</em> both of which <strong>do</strong> need a def. IMO. What is it trying to replace?</div>
 
+<div class="note"> Luc: the definition of old:thing referred to entity state. That's what it corresponds to. </div>
+
+
+
 <p>PIL is a language by which   representations of the world can be expressed using terms that are drawn from a controlled
  vocabulary.
 
@@ -132,16 +141,26 @@
 
 <div class="note">I have rephrased your sentence above. Both this version and yours still beg the question, of how one can distinguish amongst assertions made by different asserters. This may eventually boil down to some named graph language construct in the implementation, but for the time being the ability to associate an attribution to each assertion is crucial. I would make this non-optional (<code>must</code>).</div>
 
+
+<div class="note">I don't think we want the provenance of provenance directly attached to assertions. That's what provenance containers and/or account are for.  That's where attribution should/must be inserted.</div>
+
+
+
 <p>All assertions in PIL are interpreted as a record of something that has happened.</p>
 
 <div class="note">this sounds exactly like an <strong>event</strong>. Can we just use the term?</div>
 
-<p> No assumption is made on how PIL Assertions are asserted. They can
-be can be made by any means: observing, guessing, lying, or any other
-way.</p>
+
+<div class="note">Sorry, don't understand, can you write the sentence? I would possibly like to use word event for generation/use/start/end.</div>
+
+<p> 
+This specification does not prescribe the means by which assertions are made, for example on the basis of observations, inferences, or any other means. 
+</p>
 
 <div class="note">I find this a bit too informal. I have proposed a rephrasing above</div>.
 
+<div class="note">Updated</div>.
+
     </section> 
 
 
@@ -161,14 +180,20 @@
 <p>A  <dfn id="dfn-bob" title="BOB">BOB</dfn> represents an identifiable situated entity.</p>
 
 <p>A BOB assertion
- is about a situated entity, whose state and in the world are variant;
+ is about a situated entity, whose state and situation in the world are variant;
  a BOB assertion is made at a particular point and  is invariant.
  It includes a set of characteristics, predicate-value pairs, which hold about the situated entity;
 </p>
 
 <div  class="note">"whose state and in the world"  is that what you wanted to say?</div>
+
+
+<div  class="note">"whose state and SITUATION in the world"</div>
+
 <div class="note">PLEASE define (a) variant, (b) invariant. This really trips me up as a reader. I think we have agreed that variance/invariance is </em>relative</em> to a scope (time interval, interval between two events?. In any case, these terms must be defined.</div>
 
+<div  class="note">There was strong support for the above at F2F. Invariance is about an assertion ... our assertions are not mutable.  Variant ... meaning, which changes, does it need definition?</div>
+
 <p>A BOB assertion:
 <ul>
 <li> MAY contain an identity;</li>
@@ -194,6 +219,9 @@
 <p>A <dfn id="dfn-ProcessExecution">process execution</dfn> represents an identifiable activity that performs a piece of work.
 
 <div class="note">" that performs a piece of work"  is redundant as at the top we write that "activity" is a primitive concept to be interpreted in its "informal meaning"</div>
+
+
+<div class="note">" I am not very supportive of it, but was agreed at f2f1.  </div>
 </p>
 
 <p>The activity that a process execution represents has a duration, delimited by its start and its end; hence, it occurs over a continuous time interval.</p>
@@ -225,14 +253,21 @@
 
 <div class="note">I think this needs to be qualified: "according to the asserter that makes a generation assertion..."</div>
 
+<div class="note">I don't think so, since preamble already says it. </div>
+
+
+
 
 <p>A Generation assertion:
 <ul>
 <li> refers to a process execution;
 <li> refers to a BOB, which represents the situated entity that is consumed;
   <div class="note">Consumed?? </div>
-<li> contains a role, distinguishing all possible generations of a same BOB by a same process execution;
+<li> contains a role, distinguishing this generation from all other generations by the same process execution;
   <div class="note"> role? bullet item above not clear at all to me. All of a sudden we have multiple possible generations by the same process? this is going to be confusing.</div>
+
+  <div class="note"> sentence was wrong. </div>
+
 <li> MAY contain a "generation time", the time at which the situated entity was consumed.</p>
 </ul>
 </p>
@@ -242,6 +277,8 @@
 
 <div class="note"> I hope we can discuss the two statements below. Suddenly we have a relational notation for Generation. For consistency with the list above, I would put pe before x.</div>
 
+<div class="note"> Order of arguments is: effect first, cause second.</div>
+
 <div class="note"> Also: characteristics -> properties?</div>
 
 <p>Given an assertion <b>Generated(x,pe,r)</b> or <b>Generated(x,pe,r,t)</b>, the
@@ -342,6 +379,8 @@
 
 <div class="note"> What follows below definitely needs revision</div>
 
+<div class="note"> I didn't reall attempt any serious rewriting here</div>
+
 <p>Let A and B be two entity states. An assertion "B is an IVP of A" 
 indicates that, for its asserter, A and B represent the same entity in 
 the world, and the entity states modelled by A and B are consistent.</p>