[p:]declare-step vs [p:]declare-component

I actually think the [p:]declare-component element was a much better
choice here.  I'm not sure how I missed this.

The definition of the [p:]step element says the 'type' attribute
points to a 'signature'.  The definition of 'signature' is:

[Definition: The signature of a component is the set of inputs, outputs, 
and parameters that it is declared to accept.]

Later, is says:

"Each type of step (e.g. XSLT, XInclude) has a fixed signature, declared 
globally or built-in, which all its instances share, whereas each 
instance of a construct has its own signature declared locally."

IMHO, that seems a bit broken in that it uses 'step' where one would
expect 'component'.

When declaring a type that can be used as a value of a 'type' attribute 
on 'step', I think we're declaring a signature of some 'component' and
it should be named something like:

    * [p:]declare-component or,
    * [p:]declare-signature or,
    * [p:]declare-component-signature or,
    * [p:]declare-component-type.


--Alex Milowski

Received on Monday, 2 October 2006 21:07:17 UTC