[SCL] Re: some explanation

[Murray Altheim]

Pat Hayes wrote:
>Murray Altheim wrote:
>>John F. Sowa wrote:
>>
>>>Ed,
>>>
>>>Thanks for the comments.  I basically agree.
>>>
>>>And while we're talking about translating logic
>>>to XML form, I would like to quote a passage
>>>by Tim Bray, one of the chief developers and
>>>proponents of XML who also wrote one of the
>>>early tutorials on RDF back in 1998.
>>>
>>>Although Tim strongly supports XML, he is decidedly
>>>lukewarm about RDF syntax, and I strongly agree
>>>with him.  As I said before, there are many good
>>>applications for XML, but please note what Tim has
>>>to say at the end of the excerpt below:
>>[...]
>>
>>John,
>>
>>I might note that so far as I understand from talking with Pat,
>>the syntax we're proposing for XCL is XML-based but not RDF/XML-
>>based, "RDF/XML" meaning the XML serialization of RDF -- the one
>>most people think of when they think "RDF."
> 
> Absolutely. In spades.  I have never ever used 
> RDF/XML for RDF (If you read the RDF semantics, 
> there is no RDF/XML in it anywhere.) As I 
> regularly tell audiences, RDF/XML 'striped 
> syntax' is the 8-track tape of the XML world.
> 
>>The proposed syntax
>>for XCL is of the former variety:  XML-style computer markup, and
>>is the kind of XML markup that Tim Bray would consider readable.
> 
> Right. And that you, Murray, would consider extendable.

Extendable and very usable. Pretend I'm your prototype XML nerd
(as if you don't already).

>>I think Tim and I see pretty much eye-to-eye on what constitutes
>>readable XML. If someone wants to convert XCL to an RDF/XML
>>syntax, I suggest someone create an XSLT stylesheet to do that.
>>It would be relatively straightforward. It's not something I'm
>>interested in doing.
> 
> Me neither. I think even the W3C mavens are 
> drifting slowly but surely away from RDF/XML: 
> there are a variety of other RDF notations in 
> use, some of them using XML, and some of them 
> creeping into W3C recommendations.

I remember talking to several people quite a number of years ago
who shall remain nameless (since it might be impolitic to talk
about a private conversation) who were RDF people at the W3C and
really *hated* the RDF XML serialization (what you guys are calling
RDF/XML). And we've seen lots of N3 ever since, given one of them
was in a position to do something about that ugliness. Unfortunately,
the developer world does not recognize N3 parsers the way they do
XML parsers (which are now an inherent part of Java, and standard
modules for perl, python, etc.).

>>Some people have criticized my particular design style as being
>>verbose; the reason for the verbosity is partly readability.
>>With no criticism intended, you can see the difference between
>>my style and say, Steve Newcomb's SGML style in looking at the
>>SGML DTDs for ISO 13250 Topic Maps and the XTM 1.0 syntax. I
>>can't claim sole credit for the latter (it was a team effort),
>>but the element naming conventions are my style, i.e., no
>>short little element names, no abbreviations. XCL is actually
>>shorter than I'd normally make it, using 'pred' instead of
>>'predicate', but it's pretty readable. You can see a very early,
>>strawman draft at:
>>
>>     http://purl.org/xcl/
>>
>>which (Pat correct me if I'm wrong) is being used as a rough
>>basis for the current XCL syntax.
> 
> Right. Maybe 'seed' is more like it.

Hey, I don't really care. My name isn't going to be on any of
this anyway since it's an ISO spec. My name got taken off the
XML Topic Maps spec (and I got over it, kinda), so I've been
learning to live underground. I only meant to suggest that
I've been harping on you guys for several years now and I'm
seeing at least one of you not vomit at the thought of XML.

>>(And yes, I've got the PURL
>>domain for XCL so at some point we can point it to the final
>>online specification...)
> 
> We are in your debt.

I'll bill you later. But this does mean that the XML Namespace
for XCL can be

     http://purl.org/xcl/1.0/

which becomes the base URL for everything XCL-based. And this is
how all the XML-heads will use XCL. As according to the W3C,
*everything* must have an XML Namespace.

>>And to reiterate, it's not expected that people will hand-write
>>in XML. It's a serialization syntax for machine-to-machine
>>interchange. Being "readable" doesn't mean people have to
>>read it, or write it. But it's good when you *can* read it,
>>whereas I agree with Tim -- I look at RDF markup and both my
>>eyes and brain just glaze over. Some people religiously love
>>it. To each their own.
>>
>>I don't enter into this conversation with any intention of
>>defending XML against other forms of markup; that argument is
>>silly to continue. The market is currently XML. There is no
>>demonstrable "backlash" against it, and universities and
>>training centers are *adding* courses in XML markup, not
>>removing them. XML is still very much nearer the beginning of
>>its lifecycle than the end. If there is to be a serialization
>>syntax for XCL, it might include other markup forms, but if
>>anyone wants to be taken seriously by both the development and
>>Web communities (which the W3C represents a *portion* of), it's
>>waaaay past time to be arguing about syntax and just deliver
>>what is necessary to gain traction amongst the target audience.
>>XCL can then take its place amongst other SGML-based ISO
>>standards, something long overdue.
> 
> Murray, lets play politics here. Lets avoid words 
> like 'based'. I don't really care what counts as 
> base and what counts as teapot handles. All that 
> matters to developers is that there is an 
> adequate, well-designed XML specification. 

Well, actually, people do care if something is just some document
sitting on a website or if it bears some stability. "Well-designed"
doesn't necessarily go with "standard," as I note that there's a
lot of well-designed specs that are ignored, and a lot of horrible
ones that have religious adherents.

But your point is taken. Mine is simply that the development community
is, so far as I can gauge by the continued growth of XML software,
healthy. And that XML is ultimately an ISO standard can provide a
good argument that there is some longevity to the format, unlike,
say, PDF. Developers don't care about longevity, but governments and
those who contract to them do. But as to how we talk about it in the
standard, heck, I don't care. I leave that to others.

> Whether that is in appendix 19c, or in section 
> 1.0 and labelled "foundation", isn't really all 
> that important. Right?

Absolutely.

> So, the CL group just 
> loves the idea of an abstract syntax as a 
> foundation, and that's what they now have. 
> Developers aren't going to read that part of the 
> spec anyway, its just there to give theoreticians 
> wet dreams.

I guess I'm glad I'm not a theoretician then.

----

Pat, I note that the section on XCL is still pretty scant. Is
there something you're expecting me to do? Input of some sort?

The only two issues I have with what's there can be put in a
different thread, but pertain to XML character encoding and
"escaping" content in <XCL:text>. Also, in XML specifications
that have only one XML Namespace, there's no reason to namespace
prefix all the elements, as "semantically" (ugh) the prefix is
just a shorthand for the URI. So <XCL:text> can be shortened
to just <text>, which makes the spec a lot easier to read. You
need a namespace declaration in either case, e.g.,

   <clmodule xmlns="http://purl.org/xcl/1.0/">
      <text>
        (forall (x) (implies (Boy x) (exists (y)
          (and (Girl y) (Kissed x y) )) ))
      </text>
   </clmodule>

is identical (from an XML parsing perspective) to

   <XCL:clmodule xmlns:XCL="http://purl.org/xcl/1.0/">
      <XCL:text>
        (forall (x) (implies (Boy x) (exists (y)
          (and (Girl y) (Kissed x y) )) ))
      </XCL:text>
   </XCL:clmodule>

But I can rewrite this into a new thread if you want to talk about
details now (rather than in this one).

Murray

......................................................................
Murray Altheim                    http://kmi.open.ac.uk/people/murray/
Knowledge Media Institute
The Open University, Milton Keynes, Bucks, MK7 6AA, UK               .

     Bush picks Wolfowitz to head World Bank
     http://www.guardian.co.uk/usa/story/0,12271,1439191,00.html

     Paul Wolfowitz is a charter signatory on the
     Project for the New American Century (PNAC)
     http://www.newamericancentury.org/statementofprinciples.htm

    "This is the new Mein Kampf. Only Hitler did not have nuclear
     weapons. It's the scariest document I've ever read in my life."
           -- Dr. Helen Caldicott, referring to the Project for the
     New American Century report entitled "Rebuilding America's
     Defenses: Strategy, Forces and Resources For a New Century"
     http://home.earthlink.net/~platter/neo-conservatism/pnac.html

    "This is a blueprint for US world domination."
     http://www.guardian.co.uk/comment/story/0,3604,1036571,00.html