[Scl] Re: Report on Common Logic

[andersen at ontologyworks.com]

> Welcome back.

Thanks, John!  Still unwinding, but it's nevertheless good to be back, as
they say, in the "Big PX".

> Re Afghanistan:  Is it true that you were no longer getting
> combat pay because it was now considered an R & R spot rather
> than a war zone?  I can understand that our fearless leader
> needs to reduce expenses so he can give more tax breaks
> to his campaign donors.

Hee!  Bush may in fact want to give tax breaks to his campaign donors, but
he's not doing it by cutting combat pay.  It was on but is being turned off
for regions surrounding Afghanistan (e.g., Uzbekistan) but not in the
country itself.  Luckily, I am no longer collecting it :-D  Plus, the break
isn't that big even if it were there - out of the approximately $1.4B per
month being spent on OEF, saving about $2M on combat pay wouldn't come to
much.

> Re seqvars:  I'm not against them, as long as the formal
> definition doesn't obscure everything else.  I don't want
> Ian or anybody else coming up with an alternative semantics
> that is substantially simpler just because it lacks some
> of the goodies of SCL.

Isn't the "alternative semantics" just classical FOL semantics modified for
the abstract syntax?

> I agree with Chris's point:

> > The trick is not to clobber seqvars but to clearly indicate
> > that they are an *optional* feature of the framework.  Most
> > people are smart enough not to be turned off by features
> > they can easily ignore for their own purposes.  We are
> > striving for both power *and* flexibility.

I also agree.

> I have never heard a clear argument for putting seqvars
> in the core instead of an extension that also includes an
> ontology for sequences.

Well, maybe it's been that I have been away too long, but I do remember the
semantics for seq vars being somewhat, but not too, complicated.  Correct
me if I'm wrong, but we *use* FOL all the time without having to include
the axioms of set theory in every ontology we write with it.  Why are we
forced to include a theory (ontology) of sequences explicitly when we wish
to *use* seqvars.  Is not such a theory of sequences simply part of the
background theory of the semantics?

  .bill


--------------------------------------------------------------------
mail2web - Check your email from the web at
http://mail2web.com/ .