Re: a question on PROV ontology versioning

On 12/15/2014 03:27 AM, Paolo Missier wrote:
> Dear all,
>
> I am helping the dev team on the DataONE project (dataone.org) understand
> and make the best of PROV-O in their production environment.
> One question that came up concerns versioning.  Below are snippets from
> the recent conversation, for context.
> Question: how do you refer to a specific version of PROV-O?  This can me a
> moot point as there may not be a next versionŠ but I hope you see the
> general point.

The practice at w3.org (and I hope everywhere) is to never modify the 
meaning of an identifier once people are told they may rely on it, such 
as by the specification becoming a Recommendation.   So if there's ever 
a new version of PROV-O, it will simply add new terms, and possibly 
deprecate existing ones.  It wont change existing ones.

Does that address your concern?

       -- Sandro (previously W3C staff contact for Provenance WG)



> Thanks for any insight
>
> ‹Paolo
>
>
>
>> Išm thinking wešre going to have conflicts if we use the
>> http://www.w3.org/ns/prov# namespace to refer to different versions over
>> time.  I think this needs to be a discussion item on Mondayšs call.
>>
>>
>>
>>> That's all well and good, but how does the version IRI help us figure
>>> out that the prov:someConcept class that I used in 2013 is different
>> >from the one I am using today (still called prov:someConcept with exact
>>> same namespace and fragment)?
>>>> Just to follow up on this thread, I now understand that the PROV
>>>> Ontology *is* in fact versioned using the OWL conventions described
>>>> here:
>>>>
>>>>
>>>> http://www.w3.org/TR/2009/REC-owl2-syntax-20091027/#Ontology_IRI_and_Ver
>>>> sion_IRI
>>>>
>>>> So, the PROV ontology found at http://www.w3.org/ns/prov-o contains
>>>> the following property:
>>>>
>>>> owl:versionIRI <http://www.w3.org/ns/prov-o-20130430> ;
>>>>
>>>> Išve used the same convention when modifying the ProvONE ontology, so
>>>> I think wešre good with respect to future versions now, and PROV is
>>>> already strongly versioned because of this property assertion.
>>>>
>>>>> Out of curiosity, how do you make revisions in the future?
>>>>>
>>>>> We definitely need to have a robust strategy for our [extension]
>>>>> ontologies and I'd like to hear ideas on that.
>>>>>
>>>>>
>
>

Received on Monday, 15 December 2014 12:34:56 UTC