Re: PROV-ISSUE-630 (prov-sem-fpwd-review): PROV-SEM review for FPWD [Formal Semantics]

Hi James,

doc reads very well overall, below are answers to the specific questions, and some comments.  Hope you find them useful.

Regards,
 -Paolo

1.  Is the purpose of the document clear and consistent with the working group's consensus about the semantics?  If not, can you suggest clarifications or improvements?

yes.

 2.  Are there minor issues that can be corrected easily prior to FPWD release?

yes. please see below

 3.  Are there blocking issues that must be addressed prior to release as a first public working draft?  

no.

4.  Are there non-blocking, but important issues that should be discussed and resolved for future editions? (no need to list TODOs already reflected in the document itself, unless there is disagreement about how to resolve them).

no. all issues raised here are fairly minor.


====
Minor issues:
====

1.1. " It is intended as an exploration of a reference semantics for PROV, not a definitive specification of the only semantics of PROV."
   this may be too defensive?: it seems to imply that there may be other semantics of PROV  but in that case it is not clear where sand when they would come from, and thus whether using this document is "safe".

1.2 final note: not sure what this refers too -- assume it's for internal use only.

3. 

3.1
    "Things are things in the world."  some may find this a bit tautologic? since you give examples of Objects, why not give examples of Things too here, this would actually help appreciate the sharp distinction between the two that you make in 3.2 ("objects are not things...").
   Use of term "object" before 3.2 may be confusing

3.2
  "An Object is described by a time interval and attributes with unchanging values."  would adding "within that time interval" clarify the distinction between things and objects? (and would it be correct)

   "however, certain objects, namely entities, are linked to things."  "linked to" seems deliberately weak -- is it too strong to say that entities can be representations of things?

3.2.4.5 Derivations
  "chaining one or more generation and use steps."  this may lead to incorrect inference assumptions (that from use/gen you can infer derivations). Since this is clarified in 4.4.10, maybe put a ref to it?

4.4.10 (also remark in 3.3.1) the terms "precise" and "imprecise" are not from PROV-DM, indeed not in CONSTR either. But in CONSTR inference 11 effectively requires derivation to be preciseI. it may be worth unifying the terminology?

 6 I think the terms "normal form" and "valid" only appear here (and in  1.1 -- purpose). It may be useful to recall their meaning here from the constraints doc? 



typos / gremlins:

- mix of italics/ plain font:   
    "alternateOf and specializationOf relations" in 1.1
    "We write Identifiers for the set of identifiers" in 2.1  etc
     prints cramped in my rendering, and looks cramped too in the browser

- abstract: on use -> on the use

3.2.1 remark  "orthe associated thing"

3.2.4.1 "invalidation} if and only of evt?Events." (partially unreadable, of -> if)

3.3  bullet point 5:  Association -> Associations

3.3., 3.3.1 EventActivity/ EventActivities


5.1 onwards:  a few rho --> \rho gremlins in Proofs


== END ===

-- 
-----------  ~oo~  --------------
Paolo Missier - Paolo.Missier@newcastle.ac.uk, pmissier@acm.org 
School of Computing Science, Newcastle University,  UK
http://www.cs.ncl.ac.uk/people/Paolo.Missier
PGP Public key: 0x45596549  - key servers: pool.sks-keyservers.net