Re: Steps with no input/output ports

On Thu, Oct 9, 2008 at 2:50 AM,  <Toman_Vojtech@emc.com> wrote:
>
> Hello,
>
> Just wondering: Do we need to say anything special about atomic/compound
> steps that are declared as having no input ports and/or no output ports?
> For instance, consider the following declaration of an atomic step:
>
> <p:declare-step type="foo:do-some-magic"/>
>
> The fact that the step has no inputs/outputs can have interesting (or
> surprising) consequences for the user if he uses it in a pipeline. For
> instance, the XProc processor is completely free to decide when to
> execute the step since it has no explicit connections to other steps.
>
> My feeling is that these "degenerated" cases are covered sufficiently in
> the spec, but perhaps there are cases that we didn't foresee? Has
> anybody thought about this?
>

I believe we are fine.  The step executes as some point before the end
of the pipeline execution.  There are no known consequences because
there are no outputs.  If there are side effects, we don't know about them.

While this is pretty strange and probably would only be used for some kind
of "magic step", we can't say why we'd disallow them.  That is, there isn't
a good technical reason to disallow a "NOP" step.


-- 
--Alex Milowski
"The excellence of grammar as a guide is proportional to the paucity of the
inflexions, i.e. to the degree of analysis effected by the language
considered."

Bertrand Russell in a footnote of Principles of Mathematics

Received on Wednesday, 15 October 2008 17:57:59 UTC