W3C home > Mailing lists > Public > public-xml-processing-model-wg@w3.org > June 2007

Re: p:input vs. p:parameter-input

From: Norman Walsh <ndw@nwalsh.com>
Date: Thu, 21 Jun 2007 08:57:26 -0400
To: public-xml-processing-model-wg@w3.org
Message-ID: <87bqf9pirt.fsf@nwalsh.com>
/ ht@inf.ed.ac.uk (Henry S. Thompson) was heard to say:
| Norman Walsh writes:
|
|> / ht@inf.ed.ac.uk (Henry S. Thompson) was heard to say:
|> | . . .
|> | I can live with all this -- I note that the combination of the last
|> | two means in practice that 'parameters' is not allowed on the "call"
|> | to a step.
|>
|> I don't follow. If the step being called has a parameter input, then
|> they can use it, they simply use p:input to do so.
|
| What I mean is that the following is not allowed:
|
|  <p:xslt>
|   <p:input port="parameters" parameters="yes">
|    . . .
|   </p:input>
|   . . .
|
| [yes, we could allow that, but not "parameters='no'", but that would
| just be silly, I think.  Both RelaxNG and XSDL can express the
| necessary context-dependent constraint.]

Yes, we could allow that, but let's not. :-)

                                        Be seeing you,
                                          norm

-- 
Norman Walsh <ndw@nwalsh.com> | The laws of conscience, which we
http://nwalsh.com/            | pretend to be derived from nature,
                              | proceed from custom.-- Montaigne

Received on Thursday, 21 June 2007 12:57:31 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 8 January 2008 14:21:53 GMT