Re: Re Jon on Error

Terry Allen wrote:
> 
> Anyone who has a single error in his document is a bozo?  Ahem.
> I don't buy any of this.  But I've said my piece, I don't think
> it matters what you specify, and as this is a vendor-sponsored
> forum, perhaps it doesn't matter what the WG says, either.

What the WG says matters only where it is acceptable to the ERB.
That is the current state of things.  It lasts until the first 
spec is solid enough for implementors to take the initiative.
Then the dominant vendor takes the field.  Capitalism at work.

What the ERB says won't matter except to those with whom 
they negotiate privately for the future of XML.  That sort 
of discussion is pretty normal in today's market and laissez-faire 
standards processes, so, no problem here.

What the specification for XML says will determine conformance 
testing.  This will come down to well-intended voluntary 
submissions of code lumps from industry because the NIST-types 
don't have the funding to build test suites that are exhaustive.
So they will default to testing by implementations.

What the programmers will do is to ignore all of the above 
and build in error recovery for the common errors their users 
are likely to make in their applications.  The Draconian 
technique will have the same effect on them as the requirements 
for link types in SGML.  IOW, useful where used.

Meanwhile, SGML will be improving and offering simplification 
options that make it possible to be deployed on the web 
without W3C/WG/ERB approval or sanction.

So, time to deflate our shoes and explore the exhibits.

len

Received on Tuesday, 6 May 1997 20:38:49 UTC