W3C

- DRAFT -

XML Processing Model WG

Meeting 154, 01 Oct 2009

Agenda

See also: IRC log

Attendees

Present
Norm, Paul, Henry, Mohamed
Regrets
Vojtech, Alex
Chair
Norm
Scribe
Norm

Contents


Accept this agenda?

-> http://www.w3.org/XML/XProc/2009/10/01-agenda

Accepted.

Accept minutes from the previous meeting?

-> http://www.w3.org/XML/XProc/2009/09/17-minutes

Accepted.

Next meeting: telcon 8 Oct 2009

Mohamed gives possible regrets.

Plans for interim CR draft

Norm summarizes the status; getting there but not quite ready.

Norm: I propose a publication date of 12 Oct with a commitment from the editor to get a new draft available by close-of-business on 8 Oct.

No objections heard.

Proposal: Publish the current editor's working draft, with additional changes as seen fit by the editor, as a new interim CR draft on 12 Oct.

Accepted.

Review of open CR issues

-> http://www.w3.org/XML/XProc/2008/11/cr-comments/

Norm: We resolved 163, or rather directed the editor to resolve it, last time.

Norm summarizes his change to the declared output bindings.

Proposal: The editor got it right.

Accepted.

Proposal: This part of 2.2 has been successfully clarified as well.

Accepted.

Norm: I think that closes 163.

166 Steps with no type in p:library

Norm: I think the status quo is that it isn't an error to do this.
... but it does seem a little pointless.

Mohamed: I think we should say something about this.

Norm: I think there are two choices: make it an error or just add a note about it.

Norm/Mohamed discuss the fact that an implementation-defined mechanism could run them even if they don't have a type.

Paul: I don't think we should make it an error.

Norm: It looks to me fairly harmless.

Proposal: Add a note to the spec to inform readers, but not make it an error.

Accepted.

167 Runtime semantics of p:try

Norm expresses his feeling that p:try is still in the spirit of this rule.

Henry: It is worth putting an explicit note in the spec about this. If all it said was the result of a multi container step will always be the result of exactly one of it's pipelines, that would be true. But that's not quite true for p:try.
... I think we should add "with the possible exception of side-effects in the main branch of a try/catch in which the try fails."
... Pipelines are going to have side effects.

Some discussion of whether or not the xpath expressions in p:when clauses can cause side effects.

Henry: I think it's worth recasting this.

<scribe> ACTION: Editor to attempt to clarify this prose, noting that side effects may arise in p:group w/i p:try and from XPath expressions evaluated by p:when or from the bindings within p:xpath-context. [recorded in http://www.w3.org/2009/10/01-xproc-minutes.html#action01]

Henry: I'd say "with the possible except of side effects, such as..." so we don't seem to be enumerating them all.

Progress on the default processing model?

Henry reports no progress.

Henry: I will try to get something done before the face-to-face.

Any other business

Norm expresses the goal of getting to PR this month.

Mohamed: Last time I looked at the test suite, coverage wasn't that good. Has it improved?

Norm: I didn't think it was that bad after we added serialization tests.

Adjourned.

Summary of Action Items

[NEW] ACTION: Editor to attempt to clarify this prose, noting that side effects may arise in p:group w/i p:try and from XPath expressions evaluated by p:when or from the bindings within p:xpath-context. [recorded in http://www.w3.org/2009/10/01-xproc-minutes.html#action01]
 
[End of minutes]

Minutes formatted by David Booth's scribe.perl version 1.135 (CVS log)
$Date: 2009/10/01 20:00:55 $