Re: PROV comments from Clark&Parsia

On Jan 16, 2013, at 12:07 AM, Héctor Pérez-Urbina <hector@clarkparsia.com> wrote:

> 
> 
> 
> On Tue, Jan 15, 2013 at 5:56 PM, James Cheney <jcheney@inf.ed.ac.uk> wrote:
> >
> >>
> >> Does this help?  In any case, as noted below, we're happy to revisit
> >> the terminology - to me, though, "uniqueness inferences" does not
> >> sound right.
> >
> > Having said all this, I must also say this is a rather minor point. We
> > don't have any objections if you would like to keep using the term
> > "uniqueness constraints" but then adding a paragraph clarifying this
> > point would be good.
> 
> 
> Hi,
> 
> Yes, we are only debating nomenclature here; of course I agree that one can view a functional dependency/uniqueness constraint as an inference whose conclusion is an equality formula.  I just wanted to explain my rationale for grouping the different kinds of formulas into "inferences" and "constraints".  From a formal point of view, they could all be called formulas or they could all be called constraints.
> 
> I am happy with your proposed resolution of keeping things as is and clarifying the point (i.e. explaining that the names "constraint" and "inference" are semi-arbitrary and may or may not map to what seems natural in other settings).
> 
> I wanted to respond to Hector's comments in detail because some of Hector's examples gave me the impression that there was some confusion about how existential variables behave, which in turn may mean that there is something that needs to be further clarified in the document.  I will respond separately.  Thanks again for your detailed comments.
> 
> These existential variables are labelled nulls (in DB nomenclature), right? If so, I understand how they work, my point is that this should be made clear to avoid confusion :).
>  

In a nutshell yes; thanks for saving me from spending a lot of time explaining all this!  

At some point we were explicitly saying that existential variables are analogous to RDF blank nodes and to labeled nulls in databases, but this text seems to have evolved to a state where it is ambiguous what we mean, and that should definitely be fixed.


Thanks again,
--James