See also: IRC log
<ebruchez> Scribe: ebruchez
John: Next one is in october. Thursday-Friday October 16-17 will be virtual days.
Steven: Monday-Tuesday October 20-21 will be
the physical days.
... Tech plenary is on the 22. Physical location is the South of France.
<wellsk> 9-3 EST is Great for me! ANd good idea to chan ge to Wed-Thurs
John: Virtual days would be 9-3 EST on October 15-16.
<nick> http://www.w3.org/2002/09/TPOverview.html
<nick> 2-6 November 2009 or 16-20 November 2009, San Francisco Bay Area, California, USA (in conjunction with AC 2009) [under consideration - dates subject to change based on space availability*]
John: Then f2f for 2009: February 9-12 in California. June 8-11, London.
<John_Boyer> ACTION: MarkB and Charlie to arrange June 8-11 Face to face in London [recorded in http://www.w3.org/2008/06/12-forms-minutes.html#action01]
<trackbot> Sorry, couldn't find user - MarkB
<John_Boyer> ACTION: Steven to arrange Feb 2009 Face to Face with Raman [recorded in http://www.w3.org/2008/06/12-forms-minutes.html#action02]
<trackbot> Created ACTION-466 - Arrange Feb 2009 Face to Face with Raman [on Steven Pemberton - due 2008-06-19].
John: Schema question: is it possible to create an attribute group whose attributes come solely from another attribute group. E.g., UI attributes, I would like to distinguish between UI bindings and other types of bindings, e.g. @ref on actions. Also, we may add more attributes to UI bindings in the future.
Leigh: We would just do this in modularization.
John: Yes, just trying to figure out how to express it.
Paul: ...
Mark: ...
<prb> I believe that attributegroup can be a child of attributegroup, in order to make up larger groups from multiple attributegroups etc
Mark: What I was thinking was using redefine. So that was a little different, and I am not sure for certain that you can do it.
<prb> so attributegroup[@name="uibindings"]/attributegroup[@name="singlenodebinding"]/@ref
Leigh: Looking at schema for schema, attribute
groups can def and ref.
... Easy to do with Relax NG, especially with the compact notation.
<klotz> <xs:group name="attrDecls">
<klotz> <xs:sequence>
<klotz> <xs:choice minOccurs="0" maxOccurs="unbounded">
<klotz> <xs:element name="attribute" type="xs:attribute"/>
<klotz> <xs:element name="attributeGroup" type="xs:attributeGroupRef"/>
<klotz> </xs:choice>
<klotz> <xs:element ref="xs:anyAttribute" minOccurs="0"/>
<klotz> </xs:sequence>
<klotz> </xs:group>
John: About work tasks: those related to 1.1 are almost entirely done. So we have 1.2 and 2.0 issues to look at.
<klotz> <xs:complexType name="attributeGroup" abstract="true">
<klotz> <xs:complexContent>
<klotz> <xs:extension base="xs:annotated">
<klotz> <xs:group ref="xs:attrDecls"/>
<klotz> <xs:attributeGroup ref="xs:defRef"/>
<klotz> </xs:extension>
<klotz> </xs:complexContent>
<klotz> </xs:complexType>
<markbirbeck_> I've updated the Google calendar for the WG. It was very out of date (still had calls at 3pm, etc.). You can find it in Google as the first result when searching for XForms-related calendars.
John: Do we need separate schemas for modularization? Do we only do it at the spec level for now?
Steven: Yes, schema is an implementation of that.
<Steven> Charlie - here is the IBM Location in London http://www-05.ibm.com/uk/locations/sbnk.html
John: We need a new "chapter" in the spec that imports stuff from different locations.
Nick: I can certainly work on modularization. Maybe the model module?
Leigh: Start with p3ptype. It has the least relations with the rest.
John: Wondering about tables in the spec. Do we need more? We have columns for: element, attributes, and content. Would be nice to add events and behaviors. Behaviors would have link to description outside of the table. For elements, we use the "&" notation to mark extension. But could we use the word "extends"? Or a legend to make this notation clearer.
Leigh: The Atom IETF RFC uses Relax NG compact
in the spec. If you extract that you get a real schema.
... We are trying to figure out the notation to use.
Erik: So you are suggesting we use Relax compact in the spec?
<nick> http://relaxng.org/compact-tutorial-20030326.html
Leigh: Saying that instead of inventing new stuff, we should use an existing notation.
Mark: Good question. ...
<nick> http://www.ietf.org/rfc/rfc4287
Leigh: Events an behaviors don't have a formalization. The Atom RFC way is a solution for that.
<nick> example:
<klotz> http://atompub.org/rfc4287.html#atom.documents
Leigh: (showing an example on the
whiteboard)
... This way is more readable and at the same time it is formal.
John: What about elements?
Leigh: RNG doesn't discriminate.
<Steven> bind.attrs&=attribute p3ptype {string}
<nick> http://www.oasis-open.org/committees/relax-ng/compact-20021121.html
John: How to I give a name to an attribute, say
to indicate it's an XPath expression?
... So do we use this?
<nick> examples of types :
<nick> # Whatever a media type is, it contains at least one slash
<nick> atomMediaType = xsd:string { pattern = ".+/.+" }
Leigh: I suggest we do so.
<nick> atomNCName = xsd:string { minLength = "1" pattern = "[^:]*" }
Steven: I like the overview table. Nice to have processable schemas, but we can do that in appendices.
Leigh: See Atom 4.1.1 atom:feed. Good example.
Nick: Not sure if specxml is enough.
John: Should be fine.
Leigh: My point is that it is equivalent to the table representation. But unlike the table, you don't have the description right there.
John: Table format looks incomplete without
events/behavior. If we use a table, then I thought we should add more.
... What about repeat: it adds index().
Mark: index() is added because there is a
repeat element. So it's directly related.
... RNG compact gives you the markup, but that's easy anyway.
... (talking about modules, events, etc.) So xforms:repeat gets you
xforms:setindex an index() too.
Leigh: He's saying we are describing an
architecture.
... ...
Erik: This would gain us a normative schema.
John: As normative as it can be, yes.
Leigh: Schema used to be in the spec, and was taken out.
Erik: Also helps clarifying discrepancies between schema and rest of spec.
Nick: First, we need to figure out what are our goals.
John: What kind of authors are we trying to reach?
Leigh: Maybe JS authors trying to use the message module.
Mark: Next step is more abstract, we identify concepts, independently from the syntax.
John: But we need some kind of language bindings to bind those concepts to markup.
Leigh: The driver would do that.
Mark: We wouldn't go too abstract, sill keep events, etc.
Erik: At the moment I think we should keep elements and attributes.
Leigh: Some people may want to do this with attributes, e.g. role="...".
Erik: They may, but I am not sure we need to go there now.
John: We would be more defining Forms 1.2 instead of XForms 1.2.
Mark: ...
Leigh: We have two mechanisms proposed now: tables, or RNG.
Mark: We could try what we talked about yesterday. Then you say BTW, there is how you can do that with an XML attribute.
Leigh: We still need to describe an XML app.
Mark: Each module would have an XML app,
schema, etc.
... But explanation would be markup-independent.
John: This is more of an 2.0 thing.
Mark: I would try to do this abstraction anyway: first concepts, then markup.
Erik: I don't dislike it, but it's more work.
Mark: I am trying to push the syntax down.
... For modules we are releasing, I think it wouldn't be that much work to
take this more abstract approach.
... DOM 2 is a good example, not sure how intentional.
Erik: But we have lots of modules, we need to cover the whole of XForms.
John: It feels achievable. If we have all the same plan.
Erik: I think in the end it is still going to be even more work.
Nick: First divisions, then second step.
Erik: I agree with that.
Leigh: (showing the RNG schema for XForms)
Erik: I am pushing for the incremental approach.
John: Agreed, first divisions, then conversions to RNG or more abstract if we feel like it.
<klotz> xformstest.org/xforms.rnc.pdf
<klotz> http://xformstest.org/xforms.rnc.pdf
John: ... I don't understand why actions have to be defined in XML Events 2
Mark: E.g. conditional attributes in XML Events 2.
Erik: Actions are used as event handlers, which is why they are in XML Events
John: This is artificial.
Leigh: Mark is saying we define the concept of submission, and then xf:submission implements it.
Erik: Dont' see the need for it. And you can still do it, there is no standards police, you can take RNG and map that to your XML-free implementation.
Mark: I do see the need for it.
... Would it hurt to just get started and see if a common pattern emerges.
... I think our audience is wider than you think.
Erik: But I remain very unconvinced that we need to do that more work just to please a very small audience.
Mark: We've been doing this approach forever, and the result is that people don't want to implement XForms because they take it as a single monolithic thing.
<markbirbeck_> I'm not suggesting an 'extra level of indirection' in the way that Erik says. I'm suggesting something more like this:
<markbirbeck_> http://www.w3.org/TR/DOM-Level-2-Events/
Erik: True. And modularization will help. But I am unconvinced about adding this whole level of indirection.
Mark: It's a single package.
Leigh: (showing RNG schema for XForms)
<markbirbeck_> DOM 2 Events begins with a discussion of the general idea, and then much of it is described in fairly general terms.
<markbirbeck_> This kind of way of describing things is more what we want: " The event is dispatched to its target EventTarget and any event listeners found there are triggered."
<markbirbeck_> It's language-independent.
Leigh: (showing benefits of RNG compact)
John: What about events?
Leigh: I have a suggestion for events.
John: Tables are a pain in the neck. I like the RNG and what it allows to do.
Steven: Wouldn't that make it harder to include in XHTML 2.0?
John: XML Schema is just an implementation, right?
Mark: We still need an XML Schema version for XHTML 1.1 modularization.
Leigh: We can convert. Do we need all the modules separately, or just XForms as a single module to include?
John: In order to move forward: there is a desire to move towards RNC (RNG compact); we will need an RNC for XForms 1.1; in the meanwhile we need to reorg the spec; then decide how we formalize.
<scribe> ACTION: Leigh to produce and XForms 1.1 version of the Relax NG schema for XForms and ask Micah about licensing options (currently under Apache). [recorded in http://www.w3.org/2008/06/12-forms-minutes.html#action03]
<trackbot> Created ACTION-467 - Produce and XForms 1.1 version of the Relax NG schema for XForms and ask Micah about licensing options (currently under Apache). [on Leigh Klotz, Jr. - due 2008-06-19].
John: So Charlie does the data module.
Mark: I will do the message module.
... I am for releasing smaller things faster, e.g. one module out, even as a
draft.
John: Those are not rec track documents.
... It's just a question of putting a WD wrapper, whether just around the
message module or around the whole 1.2 draft.
Mark: There is a difference, the first one will
be done faster than the second one.
... Then we get a deliverable sooner.
John: Still don't see a difference in the work to do.
Mark: In modularization, you need to address things in a slightly different way.
John: But how?
Mark: I would write it first with an audience that would expect to see an independent module.
John: Is it a problem that we are chartered to produce XForms 1.2, and not 12 rec track documents?
Steven: Only from the point of view of administration.
Mark: We shouldn't be that afraid of doing what we think is best.
John: Then fine, we can produce separate documents.
Mark: Would like to get message module done, so we can have a spec out quickly and also implement it.
John: We need people to work on the modules that interest them most.
<ebruchez_> ACTION: Mark and Paul to work on message module. [recorded in http://www.w3.org/2008/06/12-forms-minutes.html#action04]
<trackbot> Sorry, amibiguous username (more than one match) - Mark
<trackbot> Try using a different identifier, such as family name or username (eg. mseaborn, mbirbeck)
<ebruchez_> ACTION: Mark and Paul to work on p3ptype module. [recorded in http://www.w3.org/2008/06/12-forms-minutes.html#action05]
<trackbot> Sorry, amibiguous username (more than one match) - Mark
<trackbot> Try using a different identifier, such as family name or username (eg. mseaborn, mbirbeck)
<wellsk> erik -- capitalize action
<ebruchez_> Keith, it seems to be working without the capitalization.
<ebruchez_> John: (looking at our timeline for deliverables)
<ebruchez_> Scribe: ebruchez_
John: (looking at our timeline for
deliverables)
... For XForms 1.1, we need implementation reports.
... There is Ubiquity, Chiba, Orbeon, Firefox, formPlayer, ...
<klotz> javascript:{var%20all,fC;all=document.all;fC=document.createElement;for(var%20i=all.length;i--;){var%20o=all[i];var%20id=o.id?o.id:o.name;if(id&&o.canHaveChildren){var%20c=fC('a');c.innerHTML='#'+id;c.href='#'+id;c.style.background="yellow";o.appendChild(c);}}}void%200;
John: So we know where we are with the roadmap.
<klotz> http://www.gate2.net/TnT/Bookmarklets.html
<klotz> javascript:{%20try%20{%20var%20doc%20=%20window.document;%20var%20anchors%20=%20doc.anchors;%20for(%20var%20i=0;%20i<anchors.length;%20i++%20)%20{%20var%20a%20=%20doc.createElement(%20'a'%20);%20a.href%20=%20doc.location.href%20+%20'#'%20+%20anchors[i].name;%20a.appendChild(%20doc.createTextNode(%20'[*]'%20)%20);%20var%20pn%20=%20anchors[i].parentNode;%20try%20{%20pn.insertBefore(%20a,%20anchors[i]%20);%20}%20catch(%20ex%20)%20{%20alert(%20ex%20);%20}%20};%20al
<klotz> ert(%20doc.anchors.length%20+%20'%20anchors%20made%20visible.'%20);%20}%20catch(%20ex%20)%20{%20alert(%20ex%20);%20}%20void(close());}
John: Once we have the UI module, it should be easy to define the streamlined syntax module.
Steven: (demo of the PicoForms plugin)
<markbirbeck> free wifi at the airport Steven? how come you never told us?
<Steven> Really?
<markbirbeck> "WIFI_SG" seems to be open and free.
<Steven> What's the name of the connection
<Steven> well
<Steven> watch out that it isn't a phishing setup
<markbirbeck> Was waiting to put in CC details, and suddenly IRC sprang into life
<markbirbeck> Sure.
<John_Boyer> ACTION: Nick to write the bind module (bind element, nodeset attribute) [recorded in http://www.w3.org/2008/06/12-forms-minutes.html#action06]
<trackbot> Created ACTION-468 - Write the bind module (bind element, nodeset attribute) [on Nick Van Den Bleeken - due 2008-06-19].
"Free Public WiFi" thing, see http://tinyurl.com/y5aflc
<markbirbeck> Sure...I wouldn't connect to another person's computer. But there are only two unsecured networks here. One is called KPN and the other WIFI_SG. There are signs everywhere saying "WiFi Zone", with tables and chairs around them, but no instruction leaflets. So I took a guess that the 'zone' was called WIFI_SG. And it worked for free. Note this person was using it in April: http://is.gd/vLD
<John_Boyer> http://lists.w3.org/Archives/Public/public-forms/2008Mar/0097.html
<Steven> http://www.w3.org/TR/xhtml-modularization/
<Steven> http://lists.w3.org/Archives/Public/www-forms-editor/2008Jun/0000.html
<John_Boyer> Meeting over
This is scribe.perl Revision: 1.133 of Date: 2008/01/18 18:48:51 Check for newer version at http://dev.w3.org/cvsweb/~checkout~/2002/scribe/ Guessing input format: RRSAgent_Text_Format (score 1.00) Succeeded: s/Cahrlie/Charlie/ Succeeded: s/psp/p3p/ Succeeded: s/audiuence/audience/ Succeeded: s/WDs/documents/ Succeeded: s/eric/erik/ Found Scribe: ebruchez Inferring ScribeNick: ebruchez Found Scribe: ebruchez_ Inferring ScribeNick: ebruchez_ Scribes: ebruchez, ebruchez_ ScribeNicks: ebruchez, ebruchez_ Present: MarkB(live) Uli Charlie John Erik Leigh Nick Rafael Rogelio Steven Keith(remote) PaulButcher Regrets: MarkS Agenda: http://lists.w3.org/Archives/Public/public-forms/2008Jun/0014.html Got date from IRC log name: 12 Jun 2008 Guessing minutes URL: http://www.w3.org/2008/06/12-forms-minutes.html People with action items: charlie leigh mark markb nick paul steven[End of scribe.perl diagnostic output]