Re: [BLD] NAU as external function?

NAU's, frames, positional - these are all the same. It depends on how the
external source tells the world to contact it.

Flora-2 allows external sources to be frames. It also allows them to be NUA's
(almost: the order matters, but axioms can be added by the user to make those
behave exactly like NAUs).

The whole point here is extensibility. If your system does not support certain
kind of externals, data types, etc., then it is supposed to reject a document
that has those. This does not impose any requirements on any existing system or
on any future system, if it chooses to not allow something.
On the other hand, it *permits* systems to support NAUs, frames, etc., as
externally defined functions, and allows them to support additional data types.


	--michael  


On Wed, 02 Jul 2008 22:57:46 +0200
Christian de Sainte Marie <csma@ilog.fr> wrote:

> Harold, Michael,
> 
> I included NAU as external function in PRD, because they are in BLD and 
> on Gary's insistance. But I still have a doubt, even in BLD: what are 
> they? Is there any language (LP or else) that support them?
> 
> Cheers,
> 
> Christian
> 
> Gary Hallmark wrote:
> 
> > NAU for external functions means that you give each arg a name and pass 
> > arguments using (arg-name, arg-value) pairs.
> > 
> > I guess we could have done that for DTB but we didn't; we use positional 
> > arguments instead.  (It might be kind of nice to have meaningful 
> > argument names in the XML RIF document that a consumer receives)
> > 
> > I see no reason to deviate from BLD on this point. Because I expect many 
> > implementations to restrict their usage of external functions to those 
> > defined in DTB, and because DTB does not use NAU, I expect this will be 
> > a rarely used feature (in BLD and PRD)
> > 
> > Christian de Sainte Marie wrote:
> > 
> >> Gary,
> >>
> >> Gary Hallmark wrote:
> >>
> >>>>
> >>>> #4: include NAU in PRD FPWD, add an editor's note and raise an issue 
> >>>> (was: option 2 in [1]).
> >>
> >>
> >> I was implementing that when I noticed that, according to the PS, an 
> >> external TERM (that is, an external function call) could also have 
> >> named arguments.
> >>
> >> Is this really what you want? It does not make sense to me (in the 
> >> proper sense: I do not understand what it means), but if this is what 
> >> you want, I will put it in PRD FPWD (and I will assume that you know 
> >> what it means, of course :-)
> >>
> >> Cheers,
> >>
> >> Christian
> >>
> > 
> > 
> 
> 
>

Received on Wednesday, 2 July 2008 21:06:47 UTC