--- a/model/comments/issue-437-satya.txt	Mon Jul 09 09:10:19 2012 +0100
+++ b/model/comments/issue-437-satya.txt	Mon Jul 09 09:28:32 2012 +0100
@@ -39,6 +39,15 @@
    > then it is more appropriate for the user or application to define/use
    > a more relevant construct wasStartedAt (for time), wasInformedBy (if
    > "starter" activity is known).
+
+There is confusion between prov-dm and prov-o. In prov-o, there is wasStartedAt.
+It doesn't exist in prov-dm.
+
+wasInformedBy(a2,a1) is not equivalent to wasStartedBy(a2,-,a1,-)
+
+
+It is reasonable to keep the name wasStartedBy because the trigger is implied when not specified.
+
    > 
    > Section 5.1.7
    > 1. Similar to above comment about wasStartedBy - "trigger" should not
@@ -46,6 +55,14 @@
    > the trigger then the wasEndedBy construct should not be used and
    > alternatively wasEndedAt, wasInformedBy should be used.
    > 
+
+Likewise.
+
+wasInformedBy(a2,a1) is not equivalent to wasEndedBy(a2,-,a1,-)
+
+It is reasonable to keep the name wasEndedBy because the trigger is implied when not specified..
+
+
    > Section 5.1.8
    > 
    > The two example assertions in Example 27 are also confusing:
@@ -60,33 +77,60 @@
    > 
    > wasInvalidatedBy(buy_one_beer_get_one_free_offer_during_happy_hour,
    > -,2012-03-10T18:00:00) should be wasInvalidatedAt
+
+Same as above.
+
    > 
    > Section 5.3.5
    > 1. The document should give a clear/concrete example of the use of
    > Influence (in querying?). Since the document recommends use of
    > specific "sub-types" of influence, the motivation or use of Influence
    > is not clear in current description.
+
+Querying is beyond the scope of prov-dm. We can't give an example.
+
+Suggestions for Influence are welcome!
+
    > 
    > Section 5.4.1
    > 1. The requirement for defining "Bundle Constructor" is not clear - is
    > it similar to a class constructor used in object oriented programming?
    > The relevance of such a construct in DM is not clear.
+
+
+TODO
+
+
    > 
    > Section 5.4.2
    > 1. Suggest that in example 39, the assertion
    > wasGeneratedBy(ex:report2, -, 2012-05-25T11:00:01) should either have
    > an "activity" or renamed to wasGeneratedAt
    > 
+
+Same as above.
+
    > Section 5.5.1
    > 1. The use of the statement "In particular, the lifetime of the entity
    > being specialized contains that of any specialization." is not clear?
    > Is it used in any constraint?
    > 
+
+Definition agreed and voted on by the WG.
+
+Constraints should make this explicit.
+
+
    > 2. Suggest using an alternate example for specialization, since
    > "bbc:news/" will be a folder and "ex:bbcNews2012-03-23" will be a file
    > within that folder (as web directories are structured)?
    > Alternative example: 
    > specializationOf(ex:W3CLastCallWorkingDraft, ex:W3CWorkingDraft)  
+
+
+Why? http://www.bbc.co.uk/news/ is a URI.
+
+
    > 
    > Section 5.5.2
    > 1. It is a bit difficult to understand Example 44:
@@ -96,6 +140,9 @@
    > 
    > Since, there is not correlation between wasDerivedFrom and alternateOf
    > (as far as I know), suggest removing Example 44.
+
+Checking with James if this is OK.
+
    > 
    > Section 5.6
    > 
@@ -106,6 +153,15 @@
    > Collection seems to be a specialization Bundle (for describing
    > provenance of provenance).
    > 
+
+bundles are sets of descriptions. Descriptions are not
+entities. Satya, you were one of those opposed to the idea of
+identifiable records ....
+
+Collections contains entities.
+
+
+
    > ================
    >  
    > 

--- a/model/prov-dm.html	Mon Jul 09 09:10:19 2012 +0100
+++ b/model/prov-dm.html	Mon Jul 09 09:28:32 2012 +0100
@@ -3161,7 +3161,7 @@
 </section>
 
 <section id="term-collection-membership">
-<h3>Collection Memberhsip</h3>
+<h3>Collection Membership</h3>
 
 <p>A <strong>collection membership</strong> relation is defined, to allow stating the  members of a Collection. </p>