Re: new BLD and the framework (FLD)

> I read FLD and I did not find nothing unclear. Maybe some examples 
> (together with a semantic structure) are useful but I believe they don't 
> have to be  part of the document.
> Anyway the document is hard to be read by potential simple users of the 
> standard. I will try to build an example use case.

Thanks Adrian,
This was the whole idea: to split the framework (which is more
mathematical) from the dialect. "Simple" users just need to read the dialect.
The framework really needs to be read only by **dialect designers**.

We will need to write a cover document, which will be a roadmap for all the
RIF documents, incl. requirements, etc. This would be a place to explain
who should read what.

> Also, seems that you cannot build a semantic structure without assuming 
> the existence of signatures. Otherwise the I_F mapping cannot be 
> finitely defined.Without signatures if I have a functional symbol p then 
> I have to define I_F(p) for all finite length sequences of arguments 
> i.e. I_F(p):D -> D, I_F(p):D^2 -> D, ...
> In the presence of the signature, just the corresponding functions have 
> to be part of the semantic structure.

Implementations never build semantic structures. They use algorithms that
implement the semantics in equivalent ways.

If you have function symbols then semantic structures cannot be built
whether you have signatures or not, because they are infinite.


> -Adrian
> Michael Kifer wrote:
> > Both drafts are now finished (except for XML) and ready for review:
> >
> > Framework:
> >
> >
> > BLD:
> >
> >
> >
> > Shake it up!
> >
> > -m
> >
> >   

Received on Tuesday, 12 February 2008 16:48:28 UTC