Re: Usefulness and Limitations of XML
Matthew West wrote:
> [Chris Menzel wrote:]
> > Have a decaf, Jim! :-) Mike's point, I think, is that current XML
> > does not natively contain the constructs needed for an SUO (or for an
> > interlingua) -- in particular, the standard apparatus of first-order
> > logic. Hence, we need a language *anyway* with that
> > expressive power. KIF appears to be the one, at least for an
> > interlingua.
>
> I think you are missing the point. XML is an alternative to BNF, not
> KIF.
Well, whether that *was* the point was exactly my question to Jim in the
following paragraph of my note, Matthew. And if you'll reread the post
from Mike U that Jim was responding to, you'll see that Mike was talking
about XML's limitations as a language for an SUO, not about its being an
alternative to BNF. Let me quote Mike:
XML has no formal semantics, and is unlikely to have one in the
forseeable future. XML Schema are coming along, but fall far short of
the expressive needs that have received much discussion on the SUO list.
...
XML schema are not scheduled for realease for many moonths to come, we
cannot hold up our effort by waiting around. I stand by my view that we
need to develop our own SUO language (or adopt or adapt an existing
one).
Doesn't my point above look like a restatement of this? My confusion
arose from the fact that Jim appeared to be objecting to this point, but
then shortly afterward seemed to suggest that he was simply talking
about the use of XML as an alternative to BNF, which is apparently how
you were reading him.
Let's try to step back and summarize. It appears that there are two
issues here:
1. XML is inadequate as an SUO, as it lacks requisite expressive power.
2. XML as it stands is a standardized alternative to BNF, so we needn't
waste time developing a standard for BNF (or EBNF).
Re 1, from the little I've seen, Mike U appears to be correct about
XML's expressive inadequacies. Hence, a standard language with at least
KIF's expressive power for the SUO is still needed.
Re 2, I will defer to the XML mavens about XML's adequacy as a medium
for expressing grammars, but it is essential that at least that portion
of XML that would be used for this have a formal semantics as clear as
the one that can be given for BNF. However, I think the brouhaha over
BNF has been a tempest in a teapot. All we're talking about here is the
medium by which a simple, extremely familiar and well-understood type of
grammar is to be defined. This is really trivial stuff and not worth
any more energy than is necessary. BNF, even in its homegrown forms, is
perfectly understandable and, once again, it would be easy to bring its
syntax and semantics up to explicitly standardizable quality. Indeed,
it may even already *be* standardized in ISO 14977 if Chris Angus is
correct. (How do we find out? Frank?) Meanwhile, even if it isn't,
because it is so widely used and well-understood already, there seems to
be no reason (??) not to continue to use it and move on with KIF
development etc, with a couple of people shepherding along whatever
niggling BNF standardization work needs to be done.
That said, if XML can express a recursive grammar as simply and
effectively as BNF, and the syntax for this and its corresponding
semantics are worked out, standardiz[ed|able], and ready to go, and
folks think that that's the mechanism we ought to use for defining KIF's
grammar (and the grammars of any of the more specialized languages that
are likely to arise), that's fine and dandy with me. Who wants to do
it?
-chris
--
Christopher Menzel # web: philebus.tamu.edu/~cmenzel
Philosophy, Texas A&M University # net: chris.menzel@tamu.edu
College Station, TX 77843-4237 # vox: (979) 845-8764