Lists Home |
Date Index |
Seems to me this is a scale issue of exchange that
conflicts with framework design. In a tightly
controlled local process, I can usually get away
with this but that tends to be very local. The
whole notion of loose coupling is that one can't
aggregate because one doesn't know or want to
know. Now we are introducing a
or 'nyuk! nyuk!'. Could be comedic.
From: Jeff Rafter [mailto:email@example.com]
> How do you know no other process or person has modified the field in
> the mean time?
You just know. If I write all of the code for my application (or have
checked with the project manager) and know that no other code writes to the
database then the fact that the data is wellformed is known before I go into
the secondary parse.
> But well-formedness can't be assumed. That's the core of XML.
Right. I absolutely agree. Also, if a document is not wellformed you have to
stop parsing right then. All of this I agree with, and this is where I would
tend to disagree with Stephen. "Assumed" was probably a poor word choice--
"known" would have been better. In many situations it is very easy to know
that a document is wellformed before a second parse begins. This doesn't
violate anything in the XML spec, it just aggregates the conclusions made
throughout the process and applies them to subsequent parses.