[XML-DEV Mailing List Archive Home] [By Thread] [By Date] [Recent Entries] [Reply To This Message]

RE: [www-vrml] x3d-3.0.dtd validation errors

• To: 'Jeff Sonstein' <jeffs@a...>
• Subject: RE: [www-vrml] x3d-3.0.dtd validation errors
• From: "Bullard, Claude L (Len)" <clbullar@i...>
• Date: Tue, 25 Nov 2003 10:07:43 -0600
• Cc: 'X3D/VRML list' <www-vrml@w...>, xml-dev@l...

Yes.  The XSV validator returns links into the 
XSD Specification.  The XSD spec is yet another 
turgid document in need of annotation and examples. 
There are good reasons for the turgidity that 
we don't need to discuss here, but the results are 
one finally has to read the DTD and Schema for 
Schemas.  Even then, what it said to me was that 
the choice element couldn't have the occurrence 
indicators.  Turned out, 'not in that context' 
and the context clue was elsewhere.  English 
annotation, good or bad, is how XML apps obtain 
semantics.  The authors did the right thing; I 
didn't find it.  My bad.

1.  Multiple eyes on software don't help at 
all if they work from a mistaken shared premise. 
Superstition at the speed of light is the fundamental 
danger of all propagation networks.  Noise made loud.

2.  Linking of results does help if and only 
if the taxonomy of definitions is correct such that semantic 
drift doesn't creep into the results.  

HINT:  A hypermedia/hypertext system is NOT an 
information space.  It is an amplifier of the 
information semantic for data stored in that 
system, right or wrong.  That is why searching 
is the new hot technology; the need to correlate 
against noise and indemnify results.

3.  At the end, we often have to turn to experts 
and if they have a many handed answer, we are just 
as lost.  Not in this case, but sometimes in others.

There is a reason for reference implementations. It 
is primarily a legal reason because by reference, 
it indemnifies the user (in this case, ISO or 
the W3DC) against that noise in a shared technology 
or definition of such.  Noise is inevitable for 
reasons we know well enough not to describe, and 
that is why the reference implementation is primarily 
a legal device.

len


From: Jeff Sonstein [mailto:jeffs@a...]

On 25 Nov 2003, at 9:51 AM, Bullard, Claude L (Len) wrote:

> That's fine until we get to one where there is
> some disagreement that we can't find a source
> for.

one of the things I like about
the way the oXygen app uses the apache XML jars
is that the validation process results in
actual hyperlinks into the relevant section of
the W3C documentation online
[see my earlier post
  entitled "Chapter and Verse"
  for examples]

do not get me wrong...
I am not claiming that this
*always* makes things "crystal clear"
for the W3C documentation
sometimes reads like ti was
originally written by Martians
then translated into Serbo-Croatian
then into Burmese
and finally into Bureaucratese
<grin/>

• Prev by Date: Re: date type validation with XML Schema
• Next by Date: RE: Is there a use for standardized binary XML (was RE: Microsoft FUD on binary XML...)
• Previous by thread: date type validation with XML Schema
• Next by thread: RE: [www-vrml] x3d-3.0.dtd validation errors
• Index(es):
  • Date
  • Thread