W3C home > Mailing lists > Public > public-rif-wg@w3.org > November 2007

Re: extensibility

From: Michael Kifer <kifer@cs.sunysb.edu>
Date: Sun, 04 Nov 2007 11:39:41 -0500
To: "Sandro Hawke" <sandro@w3.org>
Cc: public-rif-wg@w3.org
Message-ID: <9720.1194194381@cs.sunysb.edu>

> kifer@cs.sunysb.edu (Michael Kifer) writes:
> > > I've fleshed out the Extensibility Design Choices page a lot more.
> > > Feedback ASAP would be helpful.
> > > 
> > >     -- Sandro
> > > 
> > > [1] http://www.w3.org/2005/rules/wg/wiki/Arch/Extensibility_Design_Choices
> > 
> > Related to extensibility (in fact, a big part of it) is what I was
> > calling the "Framework for Logic Dialects" and was repeatedly 
> > threatening to write down :-)
> > 
> > (Well, we decided to split BLD into BLD proper and the framework. So, it is
> > no longer a threat but my action item. The current BLD draft is too complex
> > because the framework and the actual BLD are mixed in one document, so it
> > was hard to see the forest for the trees.)
> > 
> > Anyway, I started this document and wrote an overview of the framework in
> > http://www.w3.org/2005/rules/wg/wiki/FLD/
> > The rest of this document will be mostly cut and paste from the current BLD
> > draft and the actual BLD will become a small part of the current draft.
> > 
> > This overview should clarify the overall idea and will hopefully help the
> > discussion of the extensibility issues.
> Excellent, thank you.  As far as I can tell, what you're writing about
> here is orthogonal to design questions I'm working on.  Do you see
> places where they interact?  It seems like FLD languages will be a
> family of RIF dialects where extensibility is much more constrained than
> in RIF in general -- that seems good.

Yes, it is orthogonal and only applies to logic dialects. But some elements
of it could be also applied to PRD.

> As a side note, I don't understand your definition of Data Types in this
> document.  "... and no pair of distinct symbols in that symbol space can
> be interpreted by the same string.  Symbol spaces which such special
> semantics are call data types."  I'm not sure what "interpreted *by* the
> same string" means, but it sounds like this idea would rule out xsd:int,
> since 01^^xsd:int and 1^^xsd:int are the same thing, as I understand it.
> Maybe you can clarify that text at some point....

This paragraph talks about the symbol space of strings (xsd:string).
But the phrase "such special semantics" might, indeed, be a confusing,
so I made it better, I hope.

Meanwhile, I also wrote up the syntactic framework for RIF-FLD.
This should make things more concrete. It is 90% based on what we have in
RIF-BLD, but cleaned up further. The new RIF-BLD will not have all that
foundational crap and will be much easier to read for normal people. Only
at the end there will be a section that will explain how BLD is derived
from RIF-FLD. So, readers will be able to just omit that, if they are not
interested in the grand schema of things.

Received on Sunday, 4 November 2007 16:39:51 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 21:47:48 UTC