XProc Editors Draft 2007-07-19: Section 2.1 Comments

Another few editorial comments.

It's not clear to me whether a pipeline is an atomic step or a compound 
step. From the point of view of the pipeline invoking it, it's an atomic 
step, but from inside the pipeline it's a compound step, right?

In the description of "ports" (third from last paragraph), it's not 
clear whether all port (inputs and outputs) must be uniquely named, or 
whether inputs must be uniquely named and outputs must be uniquely 
named. A statement like "no port can have the same name as another port 
on the same step" would perhaps clarify things.

Final paragraph: "Steps may have access to any number of parameters, all 
with unique names." I think it's possible to have two parameter ports, 
each of which is passed a parameter called 'foo'. So parameters don't 
have to have unique names, although the parameters on a particular 
parameter port do have to have unique names. I think this paragraph 
needs rewording to something like:

   "Steps have parameter ports, on which parameters can be passed. The
    parameters passed on a particular parameter port must be uniquely
    named. A step can have zero parameter ports, and each parameter port
    can have zero parameters passed on it."

I have a feeling that this section is mixing the concepts of "step" and 
"step type" fairly freely, but I can't put my finger on any particular 
instance where this causes real confusion.

Jeni
-- 
Jeni Tennison
http://www.jenitennison.com

Received on Sunday, 22 July 2007 20:44:45 UTC