Re: Inputs/outputs and auxiliary documents

Norman Walsh wrote:
> Alex proposes, if I understood correctly, that we can solve both of
> these problems if we make the auxiliary document relationship
> explicit:

I like this approach. However, I thought on two issues that may 
difficult it:

1) Imagine the following pipeline:

<p:pipeline>
   <p:input name="doc" />
   <p:output name="fulldoc" />

   <p:step name="xinclude">
     <p:input name="document" label="$doc" />
     <p:output name="output" label="$fulldoc" />
   </p:step>
</p:pipeline>

If I know a priori which resources are referenced inside $doc, I may add 
p:input (or p:aux-input) elements to the pipeline step. What if I don't 
know these when defining the pipeline? A typical example relates to 
gluing operations (as opposed to chunking operations), either 
XInclude-based or doc() based (think feed aggregation).


2) Chunking: how can we identify multiple outputs from chunking 
operations? Maybe referencing them on a regex-based label?


> P.S. I actually think we should just use p:input/p:output for this
> purpose. An p:input or p:output element with no name and an href
> attribute would serve the purpose and wouldn't require a new element
> name. They are, after all, inputs and outputs.

I agree. Going a bit on detail, maybe marking the main input/output is 
preferred, as opposed to marking *each* auxiliar input/output.


Cheers,
Rui