- From: Norman Walsh <ndw@nwalsh.com>
- Date: Wed, 06 Jun 2007 12:32:46 -0400
- To: public-xml-processing-model-wg@w3.org
- Message-ID: <871wgpxcup.fsf@nwalsh.com>
/ Jeni Tennison <jeni@jenitennison.com> was heard to say:
| Henry S. Thompson wrote:
|> Jeni Tennison writes:
|>> I can't see anything in Norm's proposal that *exposes* the parameters
|>> passed to a pipeline as a (sequence of) documents?
Right. I made it possible to construct a parameter set that included
them, but I didn't make it possible to get at them as a document.
|> Ah, I missed 'exposes' in your (1). Not in V1 works for me, I'm
|> afraid.
| Actually all I need is a step like:
|
| <p:declare-step type="p:parameter-documents">
| <p:output port="result" sequence="yes" />
| </p:declare-step>
|
| which returns whatever parameters it's passed as (a sequence of) XML
| documents, and I'm happy.
I'd have said:
<p:declare-step type="p:parameters">
<p:output port="result" sequence="no" />
</p:declare-step>
and I'd have said the semantics of this step is that it returns, as a
c:parameters document, the set of parameters passed to it. So the
parameters passed to the pipeline are exposed with:
<p:parameters use-parameter-sets="#default"/>
In my proposal, the parameters are passed around as c:parameter
elements in a single c:parameter*s* document.
I guess I'd be happy to discard the c:parameters wrapper and just use
a sequence of c:parameter documents.
I'm going to take the fact that Henry, Jeni, and I are in apparent
agreement as a very optimistic sign and rewrite my proposal to
incorporate the changes that Henry and Jeni suggested.
Be seeing you,
norm
--
Norman Walsh <ndw@nwalsh.com> | The present moment is a powerful
http://nwalsh.com/ | goddess.-- Goethe
Received on Wednesday, 6 June 2007 16:40:03 UTC