Re: PROV-ISSUE-533: Notation, of attributes [prov-n]

Dear all,

Find below a proposed response to ISSUE-533. Feedback appreciated.

Regards,
Luc


>       ISSUE-533 (Named Attributes)
>
>   * Original
>     email:http://lists.w3.org/Archives/Public/public-prov-wg/2012Sep/0127.html
>   * Tracker:http://www.w3.org/2011/prov/track/issues/533
>   * Group Response:
>       o There is no right or wrong approach, there are essentially two
>         different philosophies. Either we adopt a named attribute
>         approach as suggested in the feedback, or we go for a
>         positional attribute solution.
>       o As suggested, by the author, it become a choice between:
>           + wasDerivedFrom( derivation = $d, drv_entity = $e2,
>             src_entity = $e1, activity = $a, generation = $g2, usage =
>             $u1, [ optional_attributes] )
>           + wasDerivedFrom(d; e2, e1, a, g2, u1, attrs)
>       o The Working Group opted for the positional argument approach
>         for the following reasons:
>           + It is commonly used in programming languages and logic; it
>             is also the approach used in OWL functional syntax
>           + It is more concise as the above example illustrates
>           + This latter point is particularly important when we write
>             inferences (see prov-constraints). For example, the
>             following inference is much more readable using positional
>             notation.
>               # IF wasGeneratedBy(id; e,a,_t,attrs) THEN
>                 wasInfluencedBy(id; e, a, attrs).
>           + Other serializations produced by the Working Group and
>             elsewhere adopt a named attribute approach (e.g. PROV-XML
>             and PROV-JSON).
>       o As far as the optional attributes were concerned, the
>         requirements were different:
>           + They are optional;
>           + A given attribute may occur multiple times with different
>             values;
>           + They can be application specific.
>       o For these, the positional solution was not suitable, but the
>         named attribute solution was good.
>
>   * References:
>       o PROV
>         constraints:http://www.w3.org/TR/2012/WD-prov-constraints-20120911/#influence-inference
>       o PROV-XML
>         Schema:http://dvcs.w3.org/hg/prov/raw-file/default/xml/schema/prov.xsd
>       o PROV JSON:http://users.ecs.soton.ac.uk/tdh/json/
>   * Changes to the document: none
>   * Original author's acknowledgement:
>
>
>     [edit
>     <http://www.w3.org/2011/prov/wiki/index.php?title=ResponsesToPublicComments&action=edit&section=52>]
>



On 10/09/12 10:50, Provenance Working Group Issue Tracker wrote:
> PROV-ISSUE-533: Notation, of attributes [prov-n]
>
> http://www.w3.org/2011/prov/track/issues/533
>
> Raised by: Paolo Missier
> On product: prov-n
>
> Would strongly prefer to see all attributes in the form of "name = value" pairs rather than relying on positional references. If the purpose of this syntax is to facilitate human readability ("aimed at human consumption", as stated in the doc), that will be achieved more easily when named attributes are used rather than expecting people to remember the order of attributes for each expression. For example:
>
> wasDerivedFrom( derivation = $d, drv_entity = $e2, src_entity = $e1, activity = $a, generation = $g2, usage = $u1, [ optional_attributes] )
>
> This would also be consistent with how the optional attributes are specified.
>
>
>
>

-- 
Professor Luc Moreau
Electronics and Computer Science   tel:   +44 23 8059 4487
University of Southampton          fax:   +44 23 8059 2865
Southampton SO17 1BJ               email: l.moreau@ecs.soton.ac.uk
United Kingdom                     http://www.ecs.soton.ac.uk/~lavm