On 2007-01-27, at 11:58 EST, Jim Grandy wrote:
On Jan 27, 2007, at 5:12 AM, P T Withington wrote:
I don't know about your theory of SWF node instantiation and the
other bugs, but I do not believe it is relevant here.
So this really doesn't have to do with LZX attribute init order,
right?
Right.
This is what we get for naming LZX "attributes" when syntactically
they refer to XML "attributes" OR "elements", and can be databound
to either XML "attributes" OR "elements".
Er, an LZX <attribute> which is an xml element of a <node> does, in
fact, define and xml attribute of the enclosing tag, but we allow you
to give that attribute a value either directly (by its name) or by
using an attribute element. Well, except for the tricky attribute
named 'text', which defines xml text (or is that data?). To define
an element in LZX, you add an element to the node, so that is a
little more straight-forward. I can't think of a better way of
handling attributes.
The confusion comes from databinding, which allows you to bind data
elements or attributes to node elements or attributes, and it is
confusing when you bind (unordered) attributes to (ordered) elements,
as the example Phil was testing did.
But now Phil has given Dr. Doc a nice pair of examples to elucidate
this point.
