RE: Precise Definition for Interoperability Needed (Was RE: [Minu tes] 6-7 Feb 2003 TAG ftf meeting (why XML))

I think you are spot on. That would be the life cycle argument.  
When it came time to move the SGML ATOS data from its original 
form and format into the hypertext systems of IADS 
in which we used both 38784 and HTML styles of markup,
we were able to do that cheaply and losslessly about 
ten years after the original markup was created.  
The lifecycle of the information was longer than 
the lifecycle of the system on which it was created 
and kept pace with the lifecycle of the system 
being documented, in this case, a C-130 aircraft. 

The same could not be said for the graphics.  We 
had to rescan those and rekey all of the illustrated 
parts breakdown relationships.  A vote for SVG and X3D.

len


From: Tim Berners-Lee [mailto:timbl@w3.org]

My guess is that what Dan meant was that really long term persistence
is the a bility of a librarian or computer scientist to figure out what
a document meant long after the standards are gone.

XML, with its recurrent use of <> and its matching redundant
use of a tag name at beginning and end of an element, makes this
sort of rosetta stone task easier than for, say, a binary document.

Similarly, if an XML document is corrupted in the middle you can
make use of some of the rest.
There is often redundancy in the indentation vs the tag closing too.

But that's just my guess could be miles off :-)

Received on Friday, 28 February 2003 14:06:22 UTC