W3C

- DRAFT -

Weekly Forms WG Teleconference

09 Apr 2008

Agenda

See also: IRC log

Attendees

Present
unl, wellsk, George_Cowe, ebruchez, Susan_Borgrink, John_Boyer, Roger, Nick_van_den_Bleeken, Leigh_Klotz, markbirbeck, Charlie
Regrets
Kenneth, Joern, Steven_(partial, on_IRC)
Chair
John
Scribe
Leigh, Uli, John_Boyer, unl

Contents


 

 

<Steven-eee> I can only participate by IRC

<Steven-eee> (at a conference, in a session)

<Steven-eee> RTP?

<Steven-eee> I wish W3C wouldn't use date space for URI like that...

<John_Boyer> Scribe: Leigh

<John_Boyer> scribenick: klotz

<Steven-eee> s/sca/space/

<Steven-eee> Yes, Shepazu, that was the backplane we were talking about

<Steven-eee> shoulld be announced any day now

<shepazu> sweet, thanks

<John_Boyer> be right there

<Steven-eee> Hi George, which company are you?

<gcowe> Hi - I'm with Origo Services

<John_Boyer> note agenda link: http://lists.w3.org/Archives/Public/public-forms/2008Apr/0055.html

John did you get my note that I'd prefer not to minute today?

<unl> i can try, but i fear it will be hard

thank you uli

<John_Boyer> Scribe: Uli

<John_Boyer> scribenick: unl

ok

Action Item List

John: any reports on action items?

<John_Boyer> Steven, were you able to do the triage of your actions?

John: lets try to make pogress

<nick> already done it

John: all, try to find a couple of hours till friday

<John_Boyer> We really need to make progress on the action items we committed to doing on the call

Bug Fix Issues

Thread for "Unfortunate name choice for target attribute"

http://lists.w3.org/Archives/Public/public-forms/2008Apr/0052.html

John: we have @target for instance replacement
... maybe we can reuse it for @replace="all"

<nick> I've hit the send button, forgot to press it. So the updated action item list is on his way

John: there is history in using @target for frame targets in HTML

Leigh: ...

<John_Boyer> 28<http://lists.w3.org/Archives/Public/public-forms/2008Apr/0052.html> rnLeigh: CSS asked us to change repeat "cursor" to "index"

Erik: most xforms implemtations run in browsers today

<John_Boyer> Leigh: to avoid confusion

Erik: so form authors are used to HTML
... we have apps like Google Docs where clicking on a link opens in a new window
... we can do this with xf:load but not xf:submission

<John_Boyer> John: (earlier) @target when used for replace="instance" submission seems to do the analogous operation to targetting a replace="all" submission to a new window or to a frame

<John_Boyer> John: (earlier) so we could even look at using target with replace="all" later

Erik: would be natural to have sth. like target on xf:submission

<klotz> +q

Erik: xf:submission replace="all" target="_blank" would be nice
... or something similar like in HTML

John: confusion is coming from having @target but behaving differently than in HTML

Erik: reusing @target for frame names is confusing
... because it would have a different content model for @replace="all"

<John_Boyer> markbirbeck: connected to modularisation

Mark: no reason why we should not have @target on xf:load and xf:submission in exactly the same way like in XHTML
... modularization brought back @target, but it is only a hint

<John_Boyer> markbirbeck: @target is regarded as a hint for replace="all" case

<John_Boyer> markbirbeck: @target is a little confusing because of the different uses between replace instance and replace all

Mark: Erik is right, that reusing/ovrloading @target is confusing

Leigh: two points: agree with Erik regarding content models for attributes
... we should free ourselves from XML Schema

<nick> it could always be an xpath expression, and then use target="'_blank'"

<nick> hmm maybe it isn't that clear I put single quotes around the _blank

Leigh: a child element for handling @replace="all"

sorry?

<John_Boyer> Ah, I see that Nick. It's hard for users to do that. They'll forget them and then not know why the target isn't working.

<klotz> it's as Raman would say bizzare for us to think that we know everything a host language needs for replace='all' so we should consider a child element instead an allow the host language to specify such attributes as target for frame.

<klotz> also i agree with erik on staying away from covariant attribute constraints until we free the specification from xml schema, which cannot express them

<nick> if _blank doesn't exists the result is the blank string ;)

John: it would be a significant change, when we change the schema
... can we just change the name?

Erik: what does it take to make such a change?

John: in PR we would have to provide a rationale for this
... otherwise going back to LC

<John_Boyer> Steven, is it possible to change target attr to some other name without going back to LC?

Erik: why not harmonizing with xf:insert which has @origin

<Steven-eee> I doubt it

<Steven-eee> that would be a pretty big change

<Steven-eee> we could restrict the LC to that feature only

John: wouldn't work since on xf:insert @origin determines the template, binding attributes determine the target

<klotz> +q

<markbirbeck> Also: http://internet-apps.blogspot.com/2006/08/using-subforms-in-xforms.html

John: it would make schema validation even harder but we cannot validate xpath expressions anyway

Erik: but it would be the first time we introduce an attribute being capable of having and id or an xpath expression
... schema support for this is not important to me
... but we would allow two different types for one attribute

John: we could use a literal xpath, as nick typed in the IRC

Erik: i understand, but you would have to write @target="'_blank'"
... everything could be an xpath expression, but my preference would be AVTs to make things dynamic

Leigh: don't want to overload an attribute with two different types
... a child element would be better

<klotz> I have no objections to leaving it as it is but. If we want to start adding stuff I would prefer the child element because it can more easily accommodate other host languages.

sorry, Mark i didn't get it

Mark: three use cases: html frames, xforms instances, and parts of document (like subforms)

Erik: we should have clear attributes for sources and targets
... @target is very general, could be used on actions too

Mark: this is a philosophical discussion

John: take back that discussion to the mailing lists

Basic XML Schema Patterns for Databinding

John: George will tell us about

George: i work for origo services in UK
... work started 3 years ago
... XML Schema 1.0 caused trouble for users

<John_Boyer> Big variation in how much XML Schema is implemented by different tools

George: especially in Web service tools because of very different coverage of Schema

please STOP! i need a break!

<John_Boyer> scribe: John_Boyer

George: schema patterns detection service. Feed schema or wsdl in, and it reports whether it conforms to pattern spec
... basic and advanced pattern specs
... advanced are recognized patterns, but you may not get a good experience from all toolkits
... last call period extends to end of month
... No impact on XForms.

John: Is there a connection between these schema patterns and *generating* XForms?

George: Not really, Mostly used in construction of good schema and wsdl

Erik: Are you saying something specifically about SOAP?

George: Yes, we're addressing SOAP-based web services because that is the arena for the toolkits

Erik: There are not many tools that take schema and produces XForms, but that is the interest from this group
... Is it possible that if someone was following these patterns that it would facilitate forms generation?

<Steven-eee> Cordys has a schema to xforms tool

George: That would probably be a dangerous thing. We're working on limitations of existing toolkits for web services that don't support schema very well.

<Steven-eee> I can imagine that other companies do too

George: If you want to generate XForms, you would want to have better schema support than the limited toolkits we are addressing.
... XForms generation should not be limited to the basic patterns

Erik: On the one hand, the basic patterns could provide a lowest common denominator, but are you saying the bar is too low?

George: Yes. Pretty much all of the Origo schemas use advanced patterns
... Most of the schema produced by BT will conform however

John: Could you explain the term 'data binding' in the title?

George: Binding the data in an XML message with host language programming construct.
... You're binding data in XML to java or C#

<unl> Axis

Leigh: Can you give some examples of the tools that are doing this?

George: Apache Axis 1 and 2
... .Net, WCF

<unl> JAXB, JIBX, etc

George: A data binding is created by a wsdl, and we want to translate that into code. But the schema patterns showing up in the wsdl docs do not always result in an ability to compile into code

<ebruchez> Thanks George

<unl> ok

John: We got the idea that maybe these patterns could be used to generate XForms "code"

<scribe> scribe: unl

<John_Boyer> markbirbeck: Origo schemas are enormous and it is hard to generate XForms from them

<John_Boyer> markbirbeck: Best practices for designing schemas with form use in mind would make a good NOTE

Mark: schemas are created in a schema-driven way

<John_Boyer> markbirbeck: yes. people get driven by schema practices, but those practices are not informed by notions of interactivity common in forms

Mark: ot allow all kinds of extension points
... but its hard to build forms on top of that
... we rolled out recently

<ebruchez> sorry guys have to step out for a momnet

Mark: i have the permission to write a white paper, especially the xforms aspects of the project

<John_Boyer> ACTION: MarkB to produce some public information about his observations on best practices for schema design with form use in mind and give link to group [recorded in http://www.w3.org/2008/04/09-forms-minutes.html#action01]

<trackbot-ng> Sorry, couldn't find user - MarkB

Forms TF

<John_Boyer> http://lists.w3.org/Archives/Public/public-forms-tf/2008Apr/0025.html

<Steven-eee> I shall try again to get Cordys to come to our meeting in Amsterdam

John: there were three emails which we should respond to
... it asks for archfirstecture documents for XForms being available
... anyone feeling guilty?

<scribe> ACTION: Charlie to respond to http://lists.w3.org/Archives/Public/public-forms-tf/2008Apr/0025.html [recorded in http://www.w3.org/2008/04/09-forms-minutes.html#action02]

<trackbot-ng> Sorry, couldn't find user - Charlie

<nick> http://lists.w3.org/Archives/Public/public-forms-tf/2008Apr/0024.html

John: next is about what Architectural Consistency actually means

<nick> http://lists.w3.org/Archives/Public/public-forms-tf/2008Apr/0017.html

John: Anne's response is much more concrete than Maciej's
... we need the high level principles as well as the specifics

<nick> http://lists.w3.org/Archives/Public/public-forms/2008Apr/0056.html

John: we need to respond to Anne's and Maciej's emails
... Sebastian indicated that he can't continue to work for TF
... anyone in the group who wants to join the TF?
... Keith, would you like to join?
... Erik, what about you?

Keith: i could do

Erik: i'm little wary to get into this ...

John: thank you, keith
... you are the face of the Forms WG in the Forms TF but don't hesistate to put back work to the group
... back to the question of how to respond to Anne and Maciej
... any volunteer or should i do the first try on public-forms?

<scribe> ACTION: John to create a strawman for a reponse [recorded in http://www.w3.org/2008/04/09-forms-minutes.html#action03]

<trackbot-ng> Sorry, amibiguous username (more than one match) - John

<trackbot-ng> Try using a different identifier, such as family name or username (eg. jkugelma, jboyer)

<nick> If I know the groups standpoint, and I can agree with it, I can do that

Mark: i can do this

John: we got to move forward with the task force, so please respond until friday

Charlie: the Backplane XG officially kicked off.
... see homepage for signing up

John: Mark, Steven, whate about the message module?

Mark: i already started with it but have to talk back to Steven

<Steven-eee> I'll be home monday

<wellsk> /leave

<markbirbeck> The point I'm making is that the presentation is all. :)

<markbirbeck> The actual mechanics and architecture of the message module is not that difficult.

Summary of Action Items

[NEW] ACTION: Charlie to respond to http://lists.w3.org/Archives/Public/public-forms-tf/2008Apr/0025.html [recorded in http://www.w3.org/2008/04/09-forms-minutes.html#action02]
[NEW] ACTION: John to create a strawman for a reponse [recorded in http://www.w3.org/2008/04/09-forms-minutes.html#action03]
[NEW] ACTION: MarkB to produce some public information about his observations on best practices for schema design with form use in mind and give link to group [recorded in http://www.w3.org/2008/04/09-forms-minutes.html#action01]
 
[End of minutes]

Minutes formatted by David Booth's scribe.perl version 1.133 (CVS log)
$Date: 2008/04/09 16:18:27 $

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)

FAILED: s/sca/space/
Succeeded: s/sace/space/
Succeeded: s/overlaod/overload/
Succeeded: s/JAX-RPC/JAXB/
Succeeded: s/it/first/
Found Scribe: Leigh
Found ScribeNick: klotz
Found Scribe: Uli
Found ScribeNick: unl
Found Scribe: John_Boyer
Inferring ScribeNick: John_Boyer
Found Scribe: unl
Inferring ScribeNick: unl
Scribes: Leigh, Uli, John_Boyer, unl
ScribeNicks: klotz, unl, John_Boyer
Default Present: unl, wellsk, George_Cowe, ebruchez, Susan_Borgrink, John_Boyer, Roger, Nick_van_den_Bleeken, Leigh_Klotz, markbirbeck, Charlie
Present: unl wellsk George_Cowe ebruchez Susan_Borgrink John_Boyer Roger Nick_van_den_Bleeken Leigh_Klotz markbirbeck Charlie
Regrets: Kenneth Joern Steven_(partial on_IRC)
Agenda: http://lists.w3.org/Archives/Public/public-forms/2008Apr/0055.html
Got date from IRC log name: 09 Apr 2008
Guessing minutes URL: http://www.w3.org/2008/04/09-forms-minutes.html
People with action items: charlie john markb

[End of scribe.perl diagnostic output]