RE: Comments on Editor's Draft 9 January 2008

> | What does "atomic step is directly evaluated by the processor" 
> | actually mean in this context? I am still a bit confused how to 
> | understand the two sentences. Can you give me an example?
> 
> Sure. Imagine that you have a command line processor that use 
> "-i port=doc" to identify inputs, "-o port=doc" to identify 
> outputs, "-ns prefix=uri" to identify command-line namespace 
> bindings, and "-pipeline qname" to identify the pipeline to run.
> 
> Then I expect the following to run the atomic step "p:xslt" directly:
> 
>   xproc -i stylesheet=x.xsl -i source=x.xml -o result=o.xml \
>         -ns p=http://www.w3.org/ns/xproc -pipeline p:xslt
> 
> If the declaration for the "p:xslt" atomic step had 
> serialization or log declarations, I would expect them to be 
> used in this case. If the p:xslt appeared in a pipeline, they 
> would not.
> 
> (I'm afraid we still don't have a clear description of how 
> p:pipeline, p:declare-step, and atomic steps really interact 
> since we made our syntax changes :-( )
> 

After reading your e-mail, I noticed a couple of mentions of "direct
invocation" in the specification. I just haven't seen it before. I think
it's a really nice feature, so perhaps it deserves more emphasis in the
specification?

Regards,
Vojtech

--
Vojtech Toman
Principal Software Engineer
EMC Corporation

Aert van Nesstraat 45
3012 CA Rotterdam
The Netherlands

Toman_Vojtech@emc.com 

Received on Monday, 11 February 2008 16:09:32 UTC