XML Processing Model WG

Meeting 186, 06 Jan 2011


See also: IRC log


Norm, Henry, Mohamed, Paul, Vojtech, Alex


Accept this agenda?

-> http://www.w3.org/XML/XProc/2011/01/06-agenda.html


Accept minutes from the previous meeting?

-> http://www.w3.org/XML/XProc/2010/12/16-minutes.html


Next meeting: telcon, 20 Jan 2011?

Per “Any Other Business” below, the 13 Jan telcon is cancelled; next meeting is 20 January 2011.

Review of the template note

-> http://www.w3.org/XML/XProc/docs/template-note.html

Norm points to Mohamed's comments: http://lists.w3.org/Archives/Public/public-xml-processing-model-wg/2010Dec/0011.html

Norm: Anyone think I got the rules for parsing "{" and "}" wrong?

No comments heard.

Mohamed proposes renaming p:in-scope-names to p:set-in-scope-names

Norm: I'm not moved.

Vojtech: We also have p:value-available() to check if an option is set; so maybe values would be better in the name.

Norm: Any other comments?

Mohamed: I'm persuaded the the verb question isn't relevant here.

Norm: I'm not sure I like values better, but I won't lie down in the road over the name.

Vojtech: No, p:in-scope-names is ok with me.

Norm: Anyone else?

None heard.

Norm: I propose to leave the name unchanged. Any objections?


Norm: Now on to p:document-template; Mohamed proposes instead p:template-document and points out, in particular, that p:document-template would be another step starting "p:document", so makes completion harder.
... I'm sort of moved. I'm not thrilled with p:parameterize-document, but p:template-document works.

Vojtech: What about just p:template?

<MoZ> +1

Henry: I have to say I like that...

Norm: I can't think of any problem with p:template. Anyone prefer *not* to name it p:template?
... I think the proposal is to rename p:document-template to simply p:template


Norm: The rest of Mohamed's note observes that the error links are broken and we don't have any examples.

Mohamed: The declaration of the steps aren't the same as the declarations in XProc; the background color is missing.

<scribe> ACTION: Norm to produce a new draft. [recorded in http://www.w3.org/2011/01/06-xproc-minutes.html#action01]

Mohamed: what about the error namespace?

Vojtech: Yes, don't we encourage users to use our error namespace?

Norm: That was specifically for err:XD0030, I think, not the errors namespace.

Vojtech: Or maybe it was the xproc-step namespace?

Norm: Yes, that rings a bell.

Brief searching doesn't turn up the relevant prose from the spec.

Norm: So where are we?

Vojtech: Saying we don't allow the error namespace for custom errors is what I'd like, but I think that would be a breaking change.

Henry: Yes, but if users are doing that, they're already in danger of walking on each other.
... Given that we didn't publish a policy for that little symbol space, people use it at their own risk.

Norm: Yes, I'm with Henry, if you started with XC0067 for your private errors, you've made an interesting design choice, but the consequences are small.

Vojtech: Perhaps we could say that we discourage users from using the err: namespace?
... And perhaps something similar for the XProc step namespace?

Norm: I'd be ok with that.

Norm: I think the proposal is to add a note of the form "Users are discouraged from using the error namespace..."


Norm: How about we do this New Orlean's style? I'll publish a draft this week. If no one objects in email next week, I'll send it off to be published as an official WG note.

<ht> +1


Review of comments on the processor profiles document

-> file://localhost/projects/w3c/WWW/XML/XProc/2010/11/lc-comments/

-> http://www.w3.org/XML/XProc/2010/11/lc-comments/

Norm: There aren't any new comments.

Henry: I haven't looked at it.

Norm: I think all we need to do is close the loop with David Lee that we're not comfortable adding more profiles

Henry: What about Vojtech's comment?

Vojtech: I think it's obvious that we expect a namespace aware processor.

Norm: I think that is what we meant, but if it's not clear...

Vojtech: We refer to the term "namespace well-formed document", I think that naturally assumes a namespace aware processor.

<ht> Yes, that's what I was looking for

Norm: I think you're right. Namespace well-formed is absolutely definitive, I think.
... So we can close your issue without change?

Vojtech: Yes, I think so.

<scribe> ACTION: Henry to close the loop with David Lee to get his assent to not add new profiles. [recorded in http://www.w3.org/2011/01/06-xproc-minutes.html#action02]

Norm: If that works out, then I think we should begin the process of getting this published as a PR.

Definition of an XProc processor

-> http://lists.w3.org/Archives/Public/public-xml-processing-model-wg/2010Nov/0052.html

Norm: Vojtech made a proposal that I liked.
... I'll draft an erratum to add that definition to the spec.
... Any other business?

We've got stuff we can do in email, I propose that we *don't* meet next week.

Next meeting is 20 January. Any objections?

None heard.

Norm: Any regrets for 20 January?

None heard.


Summary of Action Items

[NEW] ACTION: Henry to close the loop with David Lee to get his assent to not add new profiles. [recorded in http://www.w3.org/2011/01/06-xproc-minutes.html#action02]
[NEW] ACTION: Norm to produce a new draft. [recorded in http://www.w3.org/2011/01/06-xproc-minutes.html#action01]
[End of minutes]

Minutes formatted by David Booth's scribe.perl version 1.135 (CVS log)
$Date: 2011/01/06 19:00:58 $