Re: A set of proposals to change the schemas - please vote

[Tara Athan <taraathan@gmail.com>]

+1 for ( A ), ( B ), ( C ) and the Refinements C.1, C.2, C.3

Tara

On 10/6/15 4:49 PM, Tara Athan wrote:
> Please vote on ( A ), ( B ), ( C ) and the Refinements C.1, C.2, C.3 
> individually, by email or skype chat. The procedure is approved by the 
> co-chairs.
>
> PROPOSED ( A ): All skippable edges that currently are allowed to 
> occur in leaf form (in the compact serialization) and have a unique 
> child element in the non-empty version of their content model (in the 
> normalized serialization) should be eliminated in the leaf form from 
> the compact serialization (leaf form is already not allowed in the 
> normalized serialization). These include (exact list may be different 
> once the implementation is complete):
>
> appliesAlternatives
> appliesAssociations
> appliesAssociation
> appliesAuthority
> appliesJurisdiction
> appliesPenalty
> toPrescriptiveStatement
> appliesTemporalCharacteristics
> appliesTemporalCharacteristic
> fromLegalSources
> hasComment
> hasParaphrase
> atTime
> hasLegalSources
> hasLegalSource
> hasTemplate (in Constitutive, Override, Penalty, Prescriptive, 
> Reparation Statement)
> hasTemporalCharacteristics
> hasTemporalCharacteristic
>
>
> PROPOSED ( B ): The normalized serialization be modified to allow 
> "leaf" edges in the same way they occur in the compact serialization, 
> so there is no need to transform these leaf edges when converting 
> between the compact and normalized serializations. For example:
>
>  <Role>
>      ...
>       <filledBy keyref="#c1"/>
>  </Role>
>
>  is allowed in the compact serialization. As currently defined, this 
> must be changed to
>
>  <Role>
>      ...
>    <filledBy>
>       <Actor keyref="#c1"/>
>    </filledBy>
>  </Role>
>
>  in the normalized serialization. The proposal would mean the original 
> would not need to be changed in the conversion to the normalized 
> serialization. On the other hand,
>
>   <Role>
>      ...
>       <Agent key="c1" keyref="#c0">...</Agent>
>   </Role>
>
>   in the compact serialization would still need to be converted to
>   <Role>
>      ...
>    <filledBy>
>       <Agent key="c1" keyref="#c0">...</Agent>
>    </filledBy>
>  </Role>
>
>  Note: if proposal ( A ) is approved, then proposal ( B ) will apply 
> to (much) fewer edges. This includes (exact list may be different once 
> the implementation is complete)
>
> hasActor
> filledBy
> hasStrength
> appliesStrength
> hasTemplate (in Factual Statement)
>
> Tara Athan: PROPOSED ( C )
> Add two new elements <LegalReference> and <Source>
>
> The usage would be as follows:
> <LegalReference> and <LegalSource> are used only for the legal 
> documents and fragments thereof, while <Reference> and <Source> are 
> used for everything else.
>
> <...Reference> is used for non-IRI identifiers and <...Source> for IRI 
> identifiers.
>
> There is a corresponding change required to the Vocabulary.
>
> Refinements. (Dependent on Proposal ( C ).
> 1. restrict xml:base to the root (keep it optional there)
> 2. restrict values of keyref in lrml elements to the form "#..."
> 3. restrict values of keyref in ruleml elements to the form ":..."
>
> Note that these latter two refinements mean that users *must* use the 
> <...Reference> and <...Source> elements to define local identifiers 
> for external identifiers if they want to refer to them using keyref. 
> It also enforces the use of the default prefix for local CURIEs.
>
> Tara