W3C home > Mailing lists > Public > public-xml-processing-model-wg@w3.org > March 2006

Re: Inputs and outputs

From: Jeni Tennison <jeni@jenitennison.com>
Date: Wed, 22 Mar 2006 15:12:27 +0000
Message-ID: <4421695B.9030903@jenitennison.com>
To: public-xml-processing-model-wg@w3.org

Norm Walsh wrote:
> / Alessandro Vernet <avernet@orbeon.com> was heard to say:
> | It
> | bothers me that those two cases would look exactly the same in our
> | pipeline language, and that one has to know about the interface of
> | each step to see when steps are connected, and to even know if the
> | steps in the pipeline are correctly connected with each other.
> 
> That's exactly why I'm currently not a fan of the implicit steps. It
> may be marginally more tedious to include the "extra" inputs and
> outputs, but it makes the pipeline a whole lot more readable and
> maintainable, I think.

On the other hand, only a minority of steps will have no output or no 
input. You could indicate cases with something like <p:no-input> and 
<p:no-output> elements instead. Using Alessandro's example:

     <p:step name="xinclude">
         <p:input href="somedocument.xml"/>
     </p:step>

     <p:step name="url-serializer">
         <p:param name="url">file:///somefile.xml</p:param>
         <p:no-output />
     </p:step>

     <p:step name="user-credentials">
         <p:no-input />
         <p:output name="credentials" label="..."/>
     </p:step>

I'm not sure whether I really like this design, just pointing out 
alternatives.

Cheers,

Jeni
-- 
Jeni Tennison
http://www.jenitennison.com
Received on Wednesday, 22 March 2006 15:12:28 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 8 January 2008 14:21:47 GMT