Lists Home |
Date Index |
That is one of those interesting conundrums of
the current XML zeitgeist. Given the Infoset
definitions and subsequent APIs, one may be
programming to an XML data model but never
actually passing a single pointy bracketed
bit of data. Contrast this to the current
zeitgeist where some, quick to hit the coding
keys but tardy to RTFM, use XML as a super
ASCII delimited file that is dumped to a
known directory location and sucked into the
next processor. These are distinctly different
integration models, IMHO.
Is there a confusion between loose coupling
as an API/data model and loose coupling as file
types on the wire? While both are viable, I
wonder if we would have the same confusion if
the InfoSet were taught before the syntax.
On the other hand, how many systems work as
described, that is, by focusing up front on
the InfoSet and leaving the syntax for those
cases where one is actually exchanging a file.
From: Paul Brown [mailto:email@example.com]
I am of the opinion that SAX, ((J?D)|X)OM, etc. are all aspirants for a "service-oriented" architecture for document processing. In my world, XML is one of the least-likely inputs -- well-formed or not -- and outputs, but the generic functionality provided by XML tools -- data type/structural expression/validation, structural rearrangement (e.g., XSLT), etc. -- is of significant value.