W3C home > Mailing lists > Public > www-rdf-logic@w3.org > August 2001

Re: XML Serialization

From: pat hayes <phayes@ai.uwf.edu>
Date: Fri, 31 Aug 2001 13:55:36 -0700
Message-Id: <v0421010cb7b59e5e1a56@[130.107.66.237]>
To: Dan Brickley <danbri@w3.org>
Cc: www-rdf-logic@w3.org
>On Fri, 31 Aug 2001, pat hayes wrote:
>
> > >On Thu, 30 Aug 2001, pat hayes wrote:
> > >
> > > > >Now, why did the RDF WG chose XML instead of s-expressions or
> > > > >something else elegant?  I wasn't there, but I love rumor mongering
> > > > >and wild speculation.  Maybe they figured in the mood of the day, it
> > > > >would give RDF a leg up.  And it probably did, with the librarians.
> > > > >Perhaps it wasn more of a leg iron to the computer scientists, though.
> > > >
> > > > Quite, and elegantly put.
> > >
> > >
> > >Some of the work that fed into the RDF design didn't use XML.
> > >
> > >eg: http://www.w3.org/TR/NOTE-pics-ng-metadata
> > >(RDF's origins as a pornography description framework aka PICS-NG)
> > >
> > >or Guha's MCF stuff, http://www.guha.com/mcf/wp.html which itself went
> > >through the XMLization process, http://www.w3.org/TR/NOTE-MCF-XML/
> > >
> > >Probably the main reason for XMLizing all this is so that RDF could be
> > >mixed freely with other content, eg. embedded in SVG graphics, XHTML etc.
> > >http://www.w3.org/TR/SVG-access/ and so it could embed fragments of other
> > >markup languages (eg. MathML).
> >
> > Just as a technical point, I fail to follow the reasoning here. Is
> > there an assumption that anything other than XML is inherently
> > incapable of being mixed freely with, er, other content? MathML can
> > be represented in almost any language capable of rendering labelled
> > directed graph structures, surely?
>
>It was a sociological point: in 1997 all the other W3C specs were
>migrating towards XML as a common syntax; and XML was designed to allow
>multiple namespaces to be mixed together in a single document. Sure, we
>could have mixed curly and pointy brackets, but when you have as many
>working groups producing Web content formats as W3C, picking a common
>live-able-with format starts to look attractive.

To be sure, the bracket shape isn't really important, though why in 
God's name even librarians would think that '(' and ')' couldn't be 
used as brackets is beyond me. (Maybe it would have seemed too, like, 
you know, *ordinary*, or something? I guess they chose '<' and '>' 
for markup on the grounds that you hardly ever find mathematical 
characters like less-than in actual *text*, right? The kind that real 
people, with humanities degrees, read.)

While we are talking sociology/fashion, I am amused to note the the 
latest 'buzz' hitting the XML airwaves is about this very issue: 
http://www.xml.com/pub/a/2001/08/29/anglebrackets.html
So maybe things will get better, who knows?.

>BTW the W3C home page still shows an example of the old style PICS labels
>embedded in (X)HTML:
>
>	<meta http-equiv="PICS-Label"
>	content='(PICS-1.1 "http://www.icra.org/ratingsv02.html" l gen true for
>	"http://www.w3.org/" r (cz 1 lz 1 nz 1 oz 1 vz 1)
>	"http://www.rsac.org/ratingsv01.html" l gen true for 
>"http://www.w3.org/"
>	r (n 0 s 0 v 0 l 0))' />
>
>The problem is that all this sub-structure is invisible from XML tools
>(the DOM APIs etc).
>
> >
> > >You might ask why those folks use pointy brackets instead of curvy ones,
> > >but that's not an argument worth having in 2001...
> >
> > The issue isn't the shape of the brackets (though it is kind of
> > brain-damaged to choose the 'less-than' symbol as a bracket; it
> > strongly suggests that none of the XML designers were mathematicians)
> > so much as the gratuitous and wasteful use of four brackets and two
> > labels and a slash, where two brackets and one label would do fine
> > (not to mention that often, probably most of the time, you don't even
> > need the label anyway.) And the fact that this point is so blindingly
> > obvious to anyone with a modicum - nay, an infinitesimal grain - of
> > experience with formal notations does give the XML hoopla a slightly
> > sour note to many of us, I suspect.
>
><shrugs/>

No, it really does matter. XML-RDF is unusable by the RDF working 
group, and its syntactic wierdnesses and unmotivated complexities are 
a large part of the reason why RDF standardisation is taking so long. 
(RDF is such a simple language that it could be completely defined 
and locked down in a few weeks, if it were not burdened by having to 
fit into XML.)  XML-RDF-DAML is so horrible that no human being can 
read or write it reliably. What the hell is going on, when the 
language developers can't even use their own syntax because they are 
obliged to force it into this ridiculously unreadable, inefficient, 
clumsy notation that is being sung to the heavens as the answer to 
the world's problems, when there have been better notations available 
for decades, many of them in widespread use before XML was invented? 
This is an insane situation.

Moreover, almost every professional I talk to about this says 
something like: yes, yes, but just keep quiet about it, because we 
can't win this; XML has taken over and one just has to learn to live 
with it.  There is something genuinely wierd here. *Why* has XML 
taken over? Because Microsoft want it? Or because there is a kind of 
expectation bubble, where millions of people are somehow convinced 
that it is going to work miracles in some as yet unspecified way, by 
importing 'intelligence' into the Web? I strongly suspect that this 
is the major factor, and that the sane thing to do is to keep one's 
distance so as not to be hurt when the sky falls in, because in fact 
XML is *not* going to work miracles, and eventually a lot of 
disappointed people are going to be looking for someone to blame. I 
look forward to the XML winter, which I reckon should hit around 2006.

>Wasn't RDF's fault... ;-)

Yeh, that's what they all say. I'm told it wasnt XMLs fault either, 
and we should blame it on SGML; but at least that provided a way to 
define a sensible syntax if you wanted one.

Look, I can see the utility of XML for markup and for metadata. But 
we are in this situation because 'XML-izing' has become a kind of 
religion, and is applied irrationally even when the advantages are 
elusive, or maybe even imaginary, and the disadvantages are so 
extreme that they are having economic consequences. This isn't a 
sensible way to act; it's herd thinking, not good design.

Much of the fervor for XML seems to come from contrasting it with 
HTML. If the only formal language you have ever seen is HTML, then 
indeed XML looks wonderfully elegant. If you have never tasted sugar, 
honey seems like heaven. But one doesnt get good cooking by just 
adding sugar to everything.

Pat Hayes

---------------------------------------------------------------------
(650)859 6569 w
(650)494 3973 h (until September)
phayes@ai.uwf.edu 
http://www.coginst.uwf.edu/~phayes
Received on Friday, 31 August 2001 16:54:28 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 7 December 2009 10:52:40 GMT