W3C

- DRAFT -

Weekly Forms WG Teleconference

02 Jul 2008

Agenda

See also: IRC log

Attendees

Present
Charlie, John_Boyer, Nick_van_den_Bleeken, prb, ebruchez, Steven, Rafael, +0208761aaaa, markbirbeck
Regrets
Leigh, Keith, Uli
Chair
John
Scribe
Steven

Contents


 

Is Forms WG going to do an official review of the XHTML Access module?

Does someone have the action item?

<nick> no action item yet

Perhaps we can talk about it today

<Charlie> while we're getting setup, connect to www.webdialogs.com/join

<Charlie> meeting code 801862

<Roger> sure

<Roger> professiona league is the key!!

<Charlie> www.webdialogs.com/join

<Charlie> meeting code 801862

<scribe> Scribe: Steven

Demo

John: We have a Flickr demo using Ubiquity
... more a demo of the usefulness of modularization

Charlie: This is thanks to Mark and Paul

[demo of Flickr thumbnail picker]

[Uses XForms model store]

Charlie: Whether we use Flickr or any other service is generalized out in the model
... This example only uses the model, but there is no data binding
... So this is all declarative, instead of writing a new script object
... it is done using Dojo, but you can put it on top of Scriptaculous or any other framework
... like YUI

John: This is one of the advantages

Charlie: It changes the world view for the script author
... Next we want to use model-view binding to control the interface

John: Also cool is that the markup can switch to a different service, like Picasa

Charlie: Same code, just different arguments to the API in the model
... Can we check these in in the Ubiquity space?

<markbirbeck> definitely...

<markbirbeck> Probably in '_samples'.

<Charlie> ok, great

John: Shows the advantages of the module approach - you can still do something useful

Mark: There's another demo we could do
... where submission is scripted
... then the next level is to use the elements to get the same effect
... to demonstrate the layering

<Charlie> That's a super great idea...we'll do that

Mark: Great demo

Reviews for the XHTML2 WG

Steven: I see that Nick has action item for the Events review

Nick: And sent it

Steven: And John for Events

John: Yes, in progress

inputmode

Steven: Finished by next week before I go on holiday

John: Great. I'll wire it in after I get back from vacation

New Modules (binding attributes)

John: We'll talk about data modules this week if we can

Charlie: Please, I'm out next week
... and it is something we need to get our heads round

<nick> http://www.w3.org/MarkUp/Forms/specs/XForms1.2/modules/instance/bindingAttributes/index-all.html

<John_Boyer> http://www.w3.org/MarkUp/Forms/specs/XForms1.2/modules/instance/bindingAttributes/index-all.html

John: This is about 29 pages of our spec; mostly function library

Erik: I complained about the decision to include the XPath finctions
... it seems orthogonal

John: Then you get yet another module, just for the functions

<nick> I find it a bit awkward too to put them in binding module, they are just added to the XPath language

Mark: I agree with Erik

<nick> And some are incompatible with XPath 2.0

Mark: you could imagine the function spec moving at a different rate (eg new functions being added without having t update the binding module)
... And we can then give our functions to other people
... creates a library of functions
... and then good to have these functions as an API call

John: OK, easy to break up

<John_Boyer> http://www.w3.org/MarkUp/Forms/wiki/XForms_1.2_Modularization

[discussion of hierarchy of modules]

John: Breaking out another module is fine, since it fits in the hierarchy

<nick> in our RNG it is

<nick> Single.Node.Binding.Attributes = attribute ref { XPathExpression }, attribute context { XPathExpression }?

<nick> Nodeset.Binding.Attributes = attribute nodeset { XPathExpression }, attribute context { XPathExpression }?

John: Does thios doc match your expectation of a module?

<markbirbeck> http://www.w3.org/TR/xhtml-role/#docconf

Steven: I only miss the overview at the beginning

John: How do you do that with a module that only adds attributes?

Steven: I'll see if I can find one

http://www.w3.org/TR/2008/PR-xhtml-modularization-20080611/abstract_modules.html#s_xhtmlmodules

http://www.w3.org/TR/2008/PR-xhtml-modularization-20080611/abstract_modules.html#s_imapmodule

<markbirbeck> http://www.w3.org/TR/xhtml-access/

<nick> http://www.w3.org/TR/2008/PR-xhtml-modularization-20080611/abstract_modules.html#s_common_collection

<nick> 'Collection Name' and 'Attributes in Collection' no element

John: I need advice on writing schemas for modules
... especially when the module only has attributes

Mark: Shane is doing a relax implementation of modularization at the moment
... and W3C is talking modularization more generally
... so there is no hurry

<nick> http://lists.w3.org/Archives/Public/public-forms/2008Jul/0004.html

John: What is this discussion on modularization in W3C?

Mark: Oh wait, I'm mixing up two things
... Oh no, wait, it is about ndvl

Steven: It is a team internal discussion, not W3C wide

<nick> http://relaxng.org/

<nick> http://relaxng.org/spec-20011203.html

Steven: maybe the validator supporting Relax

<John_Boyer> http://www.w3.org/MarkUp/Forms/specs/XForms1.2/modules/instance/bindingAttributes/index-all.html

John: The doc supports two attribute groups
... so should the table list these two?

Steven: Yes, like in http://www.w3.org/TR/2008/PR-xhtml-modularization-20080611/abstract_modules.html#s_common_collection

John: Yes, and I could make the attributes clickable

Steven: Good idea

John: I talk about binding, and I cleaned the spec up a bit
... adding terms to the glossary
... look at Evaluation Context in XForms 1.1
... I have broken those things apart
... and referred to them as objects
... to make them reusable
... Are the examples of adding attributes to a collection?

<nick> Single.Node.Binding.Attributes |= attribute bind { xsd:IDREF } | (attribute model { xsd:IDREF }?, attribute ref { XPathExpression }, attribute context { XPathExpression }?)

<nick> Nodeset.Binding.Attributes |= attribute bind { xsd:IDREF } | (attribute model { xsd:IDREF }?, attribute nodeset { XPathExpression }, attribute context { XPathExpression}?)

http://www.w3.org/TR/2008/PR-xhtml-modularization-20080611/abstract_modules.html#s_editmodule

<nick> in RNG ;)

<John_Boyer> Nick, nice :-)

http://www.w3.org/TR/2008/PR-xhtml-modularization-20080611/abstract_modules.html#s_bdomodule

Steven: Just done in prose by the overview

John: What did we decide about using RNG in the spec?
... well let's go with that
... When we add stuff to a collection, so we respecify the collection

Steven: No

Nick: Though there are some sticky cases
... @context is not allowed on bind for instance

John: When we exclude attributes (eg bind doesn't take @model)
... we will add the attributes to the group, but then exclude them when used
... Uli had the action to universally apply @action
... but that is no longer needed
... M12N is quite challenging, because there is wording we use now that we can't use any more
... I have added notes to the wiki when that sort of factorization kicks in to remind the module writer to include other things that have been moved
... we have to decide what @model on bind would mean

Steven: We could have two or more groups of attributes if we wanted

John: Yes, but we have to ensure that the changes end up to have the same effect as now, even though they are spread over different places

Nick: Yes, I noticed this; it involves lots of rewriting

<scribe> ACTION: jboyer to split off XPath functions module from binding module [recorded in http://www.w3.org/2008/07/02-forms-minutes.html#action01]

<trackbot> Created ACTION-475 - Split off XPath functions module from binding module [on John Boyer - due 2008-07-09].

<scribe> ACTION: jboyer to provide a better overview for the binding attributes module [recorded in http://www.w3.org/2008/07/02-forms-minutes.html#action02]

<trackbot> Created ACTION-476 - Provide a better overview for the binding attributes module [on John Boyer - due 2008-07-09].

<nick> yes probably a table with columns 'Collection Name' and 'Attributes in Collection'

John: Please read the new draft

<John_Boyer> http://lists.w3.org/Archives/Public/public-forms/2008Jul/0001.html

New Modules (Data Module)

Charlie: I have cloned the spec xml tree

John: Do you have CVS access yet?

Charlie: No

John: Please work with Steven to get it
... I'm still working on the file hierarchy
... so perhaps you should wait before starting with CVS

<John_Boyer> http://lists.w3.org/Archives/Public/public-forms/2008Jul/att-0001/index-all.html

Charlie: I'm not sure how to refer to other modules and so on

<nick> ACTION: module to Actions.Attributes, Action.Content, ActionName (event name) [recorded in http://www.w3.org/2008/07/02-forms-minutes.html#action03]

<trackbot> Sorry, couldn't find user - module

John: There are two things going on here
... exporting, and importing
... is there an example of a module doing both?

http://www.w3.org/TR/2008/PR-xhtml-modularization-20080611/abstract_modules.html#s_bdomodule

John: How do we refer to them? Bibref?
... that will be difficult because we don't have specs yet

Steven: Relative references?

John: What should they be named?

<John_Boyer> http://www.w3.org/MarkUp/Forms/specs/XForms1.2/modules/instance/bindingAttributes/index-all.html

<nick> http://www.w3.org/TR/xhtml-role/

John: How do we group them?
... should there be XForms in the name?
... binding attributes, or XForms binding attributes?

Mark: I would prefer XForms for the branding
... but maybe better not to include to encourage adoption
... Looking at these things as a sort of API
... just becasue something fires an event, we don't need to include the events module
... the insert event should be in the insert module

John: But who causes the insert event

Mark: A sort of controller
... the data module needs to know things have been added and deleted

John: We do allow script access to the instance data
... bad idea
... and the repeat module may not work when the mutations happen by the backdoor

Mark: Charlie has shown you can do repeat-like stuff without the repeat module
... so that gives us some cluses about module boundaries

<nick> couldn't we do it in multiple iterations, first just create the xforms specific modules and in a next iteration split up thos XForms modules?

Nick: couldn't we do it in multiple iterations, first just create the xforms specific modules and in a next iteration split up those XForms modules?

Mark: Good point

John: So we'll have to make bibrefs to things that don't yet exist

Charlie: That's a problem for me

Nick: Let's first finish the bigger XForms modules, and then do the splitting out for other purposes

Mark: It might be easier to break out the low level modules first

John: If you have the instance module, you still need a way to put data in there
... if you use script for that, would the ability to use script be removed at higher levels?

Mark: You could define insert and delete in terms of DOM mutations
... so you get the same effect
... but a basic insert module would be like an xinclude, and useful in itself

John: But that makes our insert and delete events useless

<Roger> bye

[adjourn]

Summary of Action Items

[NEW] ACTION: jboyer to provide a better overview for the binding attributes module [recorded in http://www.w3.org/2008/07/02-forms-minutes.html#action02]
[NEW] ACTION: jboyer to split off XPath functions module from binding module [recorded in http://www.w3.org/2008/07/02-forms-minutes.html#action01]
[NEW] ACTION: module to Actions.Attributes, Action.Content, ActionName (event name) [recorded in http://www.w3.org/2008/07/02-forms-minutes.html#action03]
 
[End of minutes]

Minutes formatted by David Booth's scribe.perl version 1.133 (CVS log)
$Date: 2008/07/02 16:22:42 $

Scribe.perl diagnostic output

[Delete this section before finalizing the minutes.]
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/anothere/another/
Succeeded: s/bidning/binding/
Succeeded: s/mark/Mark/
Succeeded: s/ordering/hierarchy/
Succeeded: s/attriv/attrib/
Succeeded: s/and/... and/
Succeeded: s/idff/diff/
Succeeded: s/stiff/stuff/
Succeeded: s/ruies/ries/
Succeeded: s/thos/those/
Succeeded: s/aa/a/
Found Scribe: Steven
Inferring ScribeNick: Steven
Default Present: Charlie, John_Boyer, Nick_van_den_Bleeken, prb, ebruchez, Steven, Rafael, +0208761aaaa, markbirbeck
Present: Charlie John_Boyer Nick_van_den_Bleeken prb ebruchez Steven Rafael +0208761aaaa markbirbeck
Regrets: Leigh Keith Uli
Agenda: http://lists.w3.org/Archives/Public/public-forms/2008Jul/0006.html
Got date from IRC log name: 02 Jul 2008
Guessing minutes URL: http://www.w3.org/2008/07/02-forms-minutes.html
People with action items: jboyer module

[End of scribe.perl diagnostic output]