- From: Richard Tobin <richard@inf.ed.ac.uk>
- Date: Thu, 25 Oct 2007 15:45:16 +0100 (BST)
- To: ht@inf.ed.ac.uk (Henry S. Thompson), Richard Tobin <richard@inf.ed.ac.uk>
- Cc: public-xml-processing-model-wg@w3.org
> For the purposes of output defaulting "Output defaulting" is not defined and is not entirely clear. Say something more explicit such as "determining whether a subpipeline with no declared outputs has an implicit primary output". > 1) It is an atomic step whose step type definition specifies a > primary output port; > 2) It is a named pipeline whose definition includes an explicit > primary output port; > 3) It is a p:choose one of whose branches has a *last step* with a > primary output port; > 4) It is a p:try whose p:group has a *last step* with a primary > output port; > 5) It is some other compound step whose *last step* has a primary > output port; Perhaps it would be better to define "last atomic step" and then say "a last atomic step that is a named pipeline is considered to have a primary output only if its definition explicitly includes one". We need a similar rule for p:pipeline inputs. Is this definition also used when checking compatibility of choose and try branches? -- Richard
Received on Thursday, 25 October 2007 14:45:40 UTC