Re: PROV-ISSUE-410 (prov-primer-review): Feedback on Primer document [Primer]

Hi Simon, Yolanda

the primer reads well, as we knew already.
I have one more general comment and then specific notes, below.

General comment: I know there was a group decision to omit the PROV-N version of the examples. I still believe it was not a good 
idea. None of the people I can send this primer to are interested in the turtle syntax, in fact some won't even be able to parse it. 
I understand it is the preferred format, but can someone remind me why PROV-N was omitted altogether?
   Either way: I also don't understand why I am listed as author/contributor, as that was my only contribution.

specific comments follow. I hope they help

Regards,
   -Paolo

2.3
Activities /may/ generate ...?
Activities /may/ also use...?

should invalidation also be mentioned in this context?

2.4
"the agent must be declared explicitly both as an agent and as an entity."
   term "declared" may be open to interpretations. Consider:
the agent can be viewed both as an agent and as an entity.

2.5
"Roles are application specific, so PROV does not define any particular roles."

this must be puzzling: if this is the case, why mention roles at all in this document?
need to somehow explain that a compliant implementation is expected to recognize that there is a recognized concept of role, 
although it may not understand its specific values.

2.9

these examples may be confusing:

The same entity can evolve over time into different versions, e.g. a document that is repeatedly updated and has subsequent releases 
over time.
The same entity can be copied or replicated, e.g. a document may be copied to several directories.

because very similar scenarios were referred to as "revision" and "derivation" in 2.8  just above

sec 3
one may wonder whether stating that individuals be instances of specific classes is redundant when they are used as subjects of 
properties (i.e., in some cases where the domain of the properties is set, which is the case in PROV-O, but this is never stated).


3.4 and subsequent examples

possibly point out that the RDF statements in these examples may include non-provenance statements, i.e., ex:chartgen foaf:name ...
so in general one can mix provenance and non-provenance aspects of entities and agents.


3.10 title:  data -> records ?


typo:

sec 1 how to how

On 6/14/12 4:24 PM, Provenance Working Group Issue Tracker wrote:
> PROV-ISSUE-410 (prov-primer-review): Feedback on Primer document  [Primer]
>
> http://www.w3.org/2011/prov/track/issues/410
>
> Raised by: Simon Miles
> On product: Primer
>
> This is the issue to collect feedback on the primer document.
>
> Document to review is available from:
>
> http://dvcs.w3.org/hg/prov/raw-file/default/primer/Primer.html
>
> Questions:
>   - Is it intuitive, readable, and an appropriate introduction to the other documents?
>   - Do you judge it to be comprehensible to the range of communities that might use PROV?
>   - Is the new way of presenting examples, with choice of format, helpful?
>   - Are the examples up to date with regard to PROV-O and PROV-N?
>
> Thanks,
> Simon
>
>
>


-- 
-----------  ~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

Received on Wednesday, 4 July 2012 14:55:18 UTC