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

RE: The triples datamodel -- was Re: Semantic We

• To: 'Elliotte Rusty Harold' <elharo@m...>, "'mayfair@t...'" <mayfair@t...>
• Subject: RE: The triples datamodel -- was Re: Semantic Web permathread, iteration n+1
• From: "Bullard, Claude L (Len)" <len.bullard@i...>
• Date: Mon, 7 Jun 2004 07:53:48 -0500
• Cc: "'xml-dev@l...'" <xml-dev@l...>

In practice with submissions to State agencies, submissions 
are quarantined until such time as they pass the standard 
for submissions in that State within an acceptable error 
rate.  Only after certification are agencies enabled to 
submit data which the State then submits on to Federal 
agencies.  The State requirements and the Federal requirements 
vary from each other and from the local agency collection 
requirements.

These do occur in practice, there are many formats (vary 
by State and by Agency), the certification period is
required and is bid in the original RFP.  Because cost 
plus contracts are the rarity in this equation, the contractor 
or system vendor works with the agency directly to ensure 
submissions comply.  If Passin's company has to keep
coming back and fixing submissions, they are going out 
of business.

XML is supposed to help do this better and cheaper.  If it 
can't, it is worthless as a format for multi-agency work. 

len


From: Elliotte Rusty Harold [mailto:elharo@m...]

At 11:16 PM +0200 6/6/04, Ari Nordstrom wrote:

>What kind of system are you talking about? And where? 'cause I haven't seen
>many of the kind you're talking about either. In most systems I've seen in
>production, if you ignore the schema whenever you feel like it (provided
that
>you can do it in the authoring environment, that is), you're going to mess
up
>something, and a manual fix will be required.

You're worrying about issues that are rare in 
practice, which was precisely my point. New 
formats are rare enough that they can be handled 
by manual intervention. And generally a new 
format does indicate something important that 
should be looked at by a person.

However, the vast majority of messages are not 
new formats. They're the same old formats you've 
seen before, and they can be recognized and 
processed automatically.

If manual intervention were required for every 
message, this approach wouldn't scale or work. In 
practice, manual intervention isn't required all 
that often after the system is initially 
developed and deployed.

• Prev by Date: What's wrong w/ nested elements?
• Next by Date: Re: What's wrong w/ nested elements?
• Previous by thread: What's wrong w/ nested elements?
• Next by thread: RE: What's wrong w/ nested elements?
• Index(es):
  • Date
  • Thread