[Zope3-dev] Re: RDFLib and Zope 3

[Daniel Krech]

On Aug 29, 2005, at 11:24 PM, Gary Poster wrote:

>
> On Aug 26, 2005, at 3:03 AM, Daniel Krech wrote:
>
>> On Aug 25, 2005, at 3:10 PM, Michel Pelletier wrote:
>>
>>> On Thu, 2005-08-25 at 14:32 -0400, Gary Poster wrote:
>>>
>>>>> see what he thinks.  I wonder how "lite" the component kernel  
>>>>> can go.
>>>>>
>>>>
>>>> The only thing I have in mind is the interface package, which is  
>>>> what
>>>> Twisted uses.  That's all we would need.  zope.component needs
>>>> zope.interface, zope.testing, and zope.exceptions, according to its
>>>> DEPENDENCIES.cfg.
>>>>
>>>
>>> Right.  Well in this case we would provide just a very simple  
>>> interface
>>> facade that had no effect when run in an environment with no
>>> zope.interface (ie, catch the ImportError, null-out the facade)  
>>> or hook
>>> into zope.interface if it is available.  This way rdflib can be  
>>> still be
>>> used with or without zope.interface.
>>>
>>
>> Sounds good.
>>
>
> OK, cool.

We've started a branch where we'll be depending on zope.interface and  
in the end may just choose to grow a dependency on zope.interface or  
provide some kind of fallback if possible. We're just going to  
concentrate on getting going with zope.interface for now.

>
>
>>>>> In the mean time the adapters can live inside Zemantic, which  
>>>>> is an
>>>>> rdflib to zope bridge anyway.  Let me know if you want to send
>>>>> patches,
>>>>> otherwise I'll probably get around to adding functionality like  
>>>>> this
>>>>> soon.
>>>>>
>>>>
>>>> I'm actually interested in trying to hook this up, but have very
>>>> limited time.  I might play with it just within RDFLib alone during
>>>> some hobby time tonight, but otherwise may  need to toss this  
>>>> off to
>>>> you if you'll catch it.
>>>>
>>>> I also kind of want to hear Dan's reaction before I spend too  
>>>> much time.
>>>>
>>>
>>> #redfoot on freenode is a good place to catch him, and me.
>>>
>>
>> Yep, feel free to stop by anytime.
>>
>
> OK, cool, I plan to again. :-)
>
>
>>>> I thought I read that an RDF triad was itself something that  
>>>> could be
>>>> a node in another RDF triad, but I can't find that anywhere  
>>>> now.  Can
>>>> you confirm or deny? :-)
>>>>
>>>
>>> Yes, it's called reification, making a statement about a statement.
>>>
>>> http://www.w3.org/TR/rdf-primer/#reification
>>>
>>> http://en.wikipedia.org/wiki/ 
>>> Resource_Description_Framework#Statement_reification_and_context
>>>
>>> http://www.w3.org/TR/rdf-mt/#Reif
>>>
>>
>> Reification is probably best avoided. I'd recommend seeing if you  
>> can make use of a context or quoted graph instead.
>>
>
> I have read that 'context' actually is still a word looking for a  
> firm definition in RDF, but I like the interpretation that it is  
> the source of the assertion.  That's not a general answer for the  
> sorts of things that reification can provide, though.

Correct, 'context' was not a term defined in the set of RDF  
specifications, but all the RDF toolkits end up implementing them  
because pretty much all applications need them for something or  
another. RDFLib's support for contexts does not constrain the  
interpretation of what a context represents... it only provides a way  
to identify a set of triples. The identifier can be either a URIRef  
or BNode and can be used to make assertion such as the source of the  
triples, creation date, or whatever assertions your application wants  
to make about the context.

> The quoted graph does seem interesting: I said before that it seems  
> very similar to reification, and it does, but I guess it subsumes  
> the data structure that reification can provide...I also said that  
> I wonder about efficient indexing, and I still do.
>
> On the plus side for context, you've implemented it. ;-)

I hope to be implementing support for quoting soon... not sure of an  
ETA yet.

> I'm interested in contemplating RDF as a full catalog solution for  
> Zope, at least as a thought experiment.  The SPARQL work seems  
> interesting, in regards to some of the recent discussion on the  
> Zope 3 list; and the ability to seamlessly and simultaneously query  
> both relationship information between objects and more traditional  
> catalog data seems compelling.
>
> It seems to me that allowing a back end to index particular  
> relationships--and joins, particularly through blank nodes--would  
> be a big step in letting RDF actually be a front end to a full  
> catalog.  Another step would be having a front end that honored  
> rdfs range and domain constraints.
>
> I plan to get on IRC and bother you all again as soon as I have  
> time to do so. :-)

Hope to see you soon,

> Gary
>

Daniel Krech, http://eikeon.com/