Lists Home |
Date Index |
From: "Michael Leditschke" <firstname.lastname@example.org>
> Quoting from the above link;
> "Well-formed should be a category of minority interest to
> editor-application developers, not something for public usage."
> On the face of it, this appears to be a very sweeping statement.
> Granted, well-formedness covers a range of things from the basic
> to the (for some) esoteric, but I would have thought things like
> the redundancy inherent in XML's tag structure is of use to more
> applications that simply editors.
> What is the argument behind this statement?
For editing, working with half-arsed documents of all kinds is an advantage,
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. E.g. if I provide an xmlns declaration in an
external parameter entity, I cannot guarantee that my document
has been labelled correctly.
If XML formalized headlessness, and also formalized unvalidated-but-handling-
entities-ids-and-defaults, then XML would become more robust: you don't
need to be concerned about what the parser at the other end can handle.
A receiving application should be free to pick whichever information
items it requires: however, this should be on the basis of the kind or
name or type of information item, not how the sender happened to
compose the message. WF gives a capability that is not important,
and we would be better served with a lower and a higher bar both
of which can guarantee infoset transfer.
We are a week away from the 5th anniversary of XML's publication.
Rather than armchair improvements based on aesthetics or annecdotes,
I think there are two main groups with outstanding needs (SOAP and
XHTML) and an unfortunate hole (WF as currently defined) in the reliability
department, which can all be fixed at once. These should be fixed as a first
priority, rather than farting around with individual pet improvements (such
as providing yet another way to declare something is an id.)
I tend to believe the best way to do get this started would be for TAG to add to
the goals for XML a requirement for infoset integrity, reliability and robustness,
and ask the XML WG and community for a review of XML on those grounds,
with the expectation that it would lead along the kind of split I mention above.