W3C

- DRAFT -

XML Processing Model WG

Meeting 252, 08 Oct 2014

Agenda

See also: IRC log

Attendees

Present
Norm, Jim, Loren, Alex, Vojtech
Regrets
Chair
Norm
Scribe
Norm

Contents


Accept this agenda?

-> http://www.w3.org/XML/XProc/2014/10/08-agenda

Accepted.

Accept minutes from the previous meeting?

-> http://www.w3.org/XML/XProc/2014/10/01-minutes

Accepted.

Next meeting: 15 Oct 2014

No regrets heard.

Disposition of action items

Norm runs through the resolutions.

Norm: Anyone have questions, comments, or concerns about the errata or the disposition of those actions?

None heard.

Issue #62, 2.5.1 Specify types of variables, options, and parameters. See proposal.

-> https://github.com/xproc/specification/issues/62

-> https://ndw.github.io/specification/langspec/var-types/head/

Alex: I see the 'as' attributes in the syntax.

-> https://ndw.github.io/specification/langspec/var-types/head/diff.html

Alex: We've never laid out the context for this very well.

Norm: Fair enough.
... I've created issue 80 to track this.

Jim: Are we planning to throw an error if the types don't match?

Norm: Yes.

Norm: Any further discussion?
... I propose to accept the var-types proposal. Any objections?

<jfuller> +1

None heard.

Issue #38, 2.7.5 Syntax: allow p:inline to be optional See pull request #77.

-> https://github.com/xproc/specification/issues/38

-> https://github.com/xproc/specification/pull/77

-> http://lists.w3.org/Archives/Public/public-xml-processing-model-wg/2014Oct/0000.html

Jim: Basically, things have moved on a little since that email.

Norm: How about I leave this on the agenda for a week. I'll help get the toolchain setup so that we can review the changed spec.

Issue #53, 3.9 Consider dividing the specification.

-> https://github.com/xproc/specification/issues/53

Norm: Everytime we talk about this we waffle a bit. Last time we talked about splitting, Liam was concerned. But since he never turns up for our calls...

Alex: What would we do, exactly?
... There could be a language specification and a second spec with a vocabulary of steps.
... It requires people to look at multiple specs, but that's hardly uncommon.

Norm: Yep.

Alex: You could imagine a pipeline implementation that came with no steps. Just your own custom steps.

Norm: One motivation for a separate spec for the vocabulary is so that it can be revised on a different schedule.

Alex: There's lots of stuff we could do tactically if we had a separate spec.

Norm: Is there anyone opposed to separate specs?

None heard.

Norm: Shall we split the XProc 2.0 spec into two specifications: a core language specification and a step vocabulary specification?
... Anyone in favor?

Jim: I think it's a good idea.

Norm: Any objections.

<alexmilowski> +1

None heard.

Norm: Ok, I'll take a stab at it.

Issue #37, 2.7.4 Syntax: allow <p:input port="portname"/>

-> https://github.com/xproc/specification/issues/37

<scribe> ACTION: A-252-01 Norm to see if the WG mailing list can be subscribe to the github issue tracker. [recorded in http://www.w3.org/2014/10/08-xproc-minutes.html#action01]

Some discussion of the problems associated with our new github-based tool chain.

Jim and Vojtech argue that this change is confusing wrt the default readable port.

Vojtech: I also think it conflicts with the input declaration.
... I like the idea of a pipe attribute, which is similar to the suggestion that we allow an href attribute on p:input.

Alex: There are two separate things going on, the shortcut for empty and the idea of a shorthand for referencing other things.
... On the empty side, I wonder how much this has to do with parameters.
... That's where I've used it a lot and I don't know where else I've used it.
... The shorthand to refer to other things is a different usability question.

Norm: I think I'm hearing consensus *not* to make the change proposed in issue #37.
... Anyone disagree?
... I propose that we reject this request.
... Any disagreement?

None heard.

2.1 Simplify parameters

->http://lists.w3.org/Archives/Public/public-xml-processing-model-wg/2014Oct/0016.html

<Zakim> ht, you wanted to ask about 2.7.8, and other similar bits of the spec

Henry: The new 2.7.8 is p:make-map()
... I don't think "markup errors are ignored" is acceptable.
... It occurs to me that we need to perhaps be more explicit about this. We might need to think about giving a general purpose option between strict and lax.
... Where what I have in mind is that the default behavior is we'll ignore individual parameter bindings that we can't make sense of but if we can find ones we can make sense of we'll use those.
... I think we have to be clear that it has to be a document.
... If it's not clear how to make sense of it, you have to halt and catch fire. And there should be an option to specify that behavior if the document isn't valid.
... I wonder if there are other things like this that we need to treat in a more-or-less uniform manner.

<ht> By 'like this', I mean little XML languages in the c: namespace

Any other business?

None heard.

Summary of Action Items

[NEW] ACTION: Norm to see if the WG mailing list can be subscribe to the github issue tracker. [recorded in http://www.w3.org/2014/10/08-xproc-minutes.html#action01]
 
[End of minutes]

Minutes formatted by David Booth's scribe.perl version 1.138 (CVS log)
$Date: 2014/10/08 15:43:21 $