Lists Home |
Date Index |
Yes. As Doc G. put it, "nothing happens until a bit changes state."
The question becomes when is it useful to build a requirement
based on an abstraction, and when does one have to specify
a physical if transient entity? A stab at it using a metaphor:
the syntax provides us a set of brown bags. An infoset gives
us a set of labels for these bags that tell us what is in
each labeled bag. I'm sure a better educated CS person can
do a better job with that.
Still, I think we have come to where so many of our conflicts
start: what we use the abstract specifications for and what
we use the physical if transitent entity specifications for.
From: Tim Bray [mailto:firstname.lastname@example.org]
Bullard, Claude L (Len) wrote:
[in response to John Cowan's pointing out how close the infoset is to
the underlying XML]
> That would tend to suggest that for the XML on The
> Web system, the Infoset specification is core and
> that XML 1.0 is a syntax mapping corresponding to
> a subset of SGML.
Except for, there's no de facto or de jure way to exchange an infoset as
an infoset. And what we do on the Web is exchange data. This is why
XML is normatively defined at a syntactical level. Without reliable
data interchange, you have nothing. With it, you have the potential to
build anything. -Tim