[
Lists Home |
Date Index |
Thread Index
]
Rick Jelliffe wrote:
> For data/document transfer, there should be no variability of the infoset possible:
> what you send is what they get. I think it is a basic matter of data integrity.
<HOWL theme="I saw the best minds of my generation . . .">
We cannot go there. The definitive, canonical, stable, predictable form or version of
an XML instance is . . . that XML instance. Yes, it must be parsed to be used. Yes, it
will likely be parsed under different circumstances and with different expectations
for each different use: that is the necessary corollary of those uses differing. Yes,
some information set will likely be built on the output of each of those parses and,
yes, those information sets will vary, as they must given the varying circumstances of
their creation--which is at the time and place of parsing, not at the creation of the
original XML instance. This one fact alone is the basis for the undeniable distinction
between the instance and any of the information sets which might be elaborated from
it.
Syntax is concrete, persistent and transmissible. Semantics may be elaborated from a
concrete instance by the operation of particular process on a particular occasion
under particular circumstances. Such semantics are specific to those circumstances and
ephemeral with them. You cannot transfer them, enforce them, suppress their
variability, or through force of will intend them. Philology and mature exegetical
disciplines have finally, painfully learned that. Us?
</HOWL>
|
