Joint Forms/HTML WG Forms Task Force Meeting

31 Oct 2007

See also: IRC log


Nick_van_den_Bleeken, markbirbeck, Gregory_Rosmaita, anne, +1.408.996.aaaa, mjs




<scribe> Meeting: Joint Forms Task Force

<scribe> scribenick: oedipus

i do, anne

<scribe> scribe: Gregory_Rosmaita

<scribe> ScribeNick: oedipus

2 minutes -- introductions and interests (hot button issues, too)

discuss scoping (as prelude to next item)

Mark: what do people want out of task force? what would like acheived
... personally, think neeed to find bridge between solutions (XForms and Web Forms 2.0) -- discussing in Forms WG ways of making transition easiest possible; keen to see if can meet in the middle -- some of the things Web Forms 2.0 extensions could they be part of what XForms is working on? if developer moving from HTML forms to HTML5 forms to XForms 2, would like to be a smoothe process

AVK: interested in hearing what kind of input XForms WG has on WF2 (Web Forms 2.0)
... expressed outcomes in charter proposal; transition path between the two -- diffferences, mostly scripting; don't make too complex

GJR: i would like the task force to consider dave raggett's XForms Transitional draft

AVK: out of charter scope -- discussing technical spec; we are defining an architecture not a technical spec

Mark: not start with XForms Transitional -- conceptual ideas in conflict; dave's not around to support it
... think would be interesting to have proper discussion on XForms 2 extension; assess how this bridging might be achieved from WF2.0 to XForms

AVK: not sure XForms transitional compatible

Mark: background: number of things in XForms that are syntaxically unecessary; if don't express instance of XML data, get data created for you; could not require expression of model -- not a new language, but how do we tune XForms so retains its conceptual model, but easier to use syntaxically and could be used by relative novice
... need to work out: if have input control bound to firstname and another to secondname -- want to concatonate in XForms XPath concatonation -- how to bridge the 2? not by making them look the same -- different languages in model -- javascript in HTML and XPath in XForms; 2 ways of achieving -- how to get from one to the other
... meta-information will help guide us to ascertain where we are coming from; not religious about this -- javascript so commoonly used to make widgets, don't always need to create an XPath -- 2 ways of doing things

Nick: nothing against javascript or XPath -- if it is the best solution, should be able to use it

AVK: script mark HTTP-Header and meta elements -- don't know if supported in gecko

Mark: don't know how widely supported; would require us to say XPath is a scripting language
... type of issue we need to step through to achieve goal
... many good candidates for attributes; support for multiple languages in browser

AVK: single platform easy to add a lot of features -- if 4 platforms need to support different methods, it will be a waste

Mark: certain attributes contain script can be used as way to take javascript from WF2 and XPath from XForms

discuss AVK's charter proposal (latest draft): http://www.w3.org/2007/10/forms-tf/charter-proposal

AVK: should be concentrating on guidelines

Mark: trying to ensure that no particular architecture is ruled out by what we do -- if just have input control or 2 inputs and a concatonated output, nothing requires MVC architecture; don't want to rule one way or another -- implement as implementor chooses -- if someone introduces attribute, will not break

AVK: mapping simple HTML forms DOM structure to XForms

Mark: no need for conversion --

AVK: define mapping between WF2 and XForms -- WF2 makes additional parts of forms easier to use -- doesn't add much as far as a model or output --

Mark: in XForms model, if say XML node called date, anywhere put input control associated with date, can obtain it; some argue shouldn't put data on input control; don't think breaks MVC model; XForms needs ability to set view level properties -- one control read-only one not read-only -- if set forms control, all will be read-only; need way to distinguish which are which -- build on HTML4 and WF2 rather than be in conflict with them
... doesn't matter that don't shave MVC model -- can conceptually act as if there is one -- there is a model, can't see -- WF2 applies to "view level" XForms attributes apply to model

AVK: seems like a model constraint; datatypes allow you to do that

Mark: constraints are connected to model; WF2 if have 2 controls connected to same data can have separate types, right?

AVK: not sure

Mark: is it like HTML control array?

AVK: yes, built on top of HTML

Mark: read-only -- to get in XForms have to say some data is read-only, input control bound to that becomes read-only; programmers start with UI, add more controls and data, shouldn't rule out use of MVC

AVK: guidelines - conversion from HTML to XForms model -- not about WF2 (supplemental to HTML forms) -- how do HTML forms used everywhere map to XForms

Mark: believe you can map them

AVK: not saying can't but that is the challange

Mark: HTML forms pretty easy; people get worked up over WF2 and XForms are similarities

AVK: read-only and disabled are also part of HTML forms -- not new; WF2 allows read-only on slightly more controls and disable on select -- not new feature

Mark: want to begin with new features even if only half-a-feature; harmonize work being done; if simply harmonize with HTML miss out on what is being done in other areas

GJR: from SS post to public-forms-tf -- agree on architectural consistency but must define

AVK: for some may be clear mapping between XForms and WF2 or single interface -- not sure

Mark: conceptually, see them based on same model
... if is model underneath HTML form, could still be described by MVC -- data entered into form, processed, serialized and delivered; HTML onSubmit queries all controls and submits that; could collect in a model, even if not explicit; allows further extension of forms

<mjs> my apologies gentlemen, had a traffic disaster

<mjs> can someone remind me of the call-in number?

hope you are ok

Zakim Bridge +1.617.761.6200, passcode 36767# ("FORMS#")

<mjs> not an accident, just bad traffic on the way to work

charter review period

MJS: personal view: as stands, architectural consistency is vague, but is work of TF in executing charter to find what it means; different possible interpretations need to be investigated and which make sense for what

identify regular meeting time (every other week?)

AVK: can do by email: charter review period 2 weeks from now -- 3 weeks publish as official after taking to groups

RESOLUTION: internal charter review period will end 3 weeks from today, after which will be submitted to respective WGs

AVK: produce some guidelines -- 23 november 2007 deadline -- ok?

MJS: occasional meetings good -- only as needed

GJR: when meet impasse or can't achieve consensus on list then can call teleconference to discuss

scribe's note: general agreement

chair+ none

attendees+ Nick_van_den_Bleeken, Mark_Birbeck, Anne_Van_Kestren, Gregory_J_Rosmaita, Maciej_Stachowiak

attendees: Nick_van_den_Bleeken, Mark_Birbeck, Anne_Van_Kestren, Gregory_J_Rosmaita, Maciej_Stachowiak

on the call i heard you, me, nick, mark, and maciej -- sebastian wasn't there

<mjs> what was mjs?

<mjs> oh, the .aaaa number?

<mjs> I think that was me

<anne> yeah

mjs: did you call from 408?

<mjs> I don't know why it listed me twice

<mjs> Apple is in the US 408 area code

perhaps that means that you're excused from our next telecon?

attendees- +1.408.996.aaa

chair+ none

Summary of Action Items

[End of minutes]

Minutes formatted by David Booth's scribe.perl version 1.128 (CVS log)
$Date: 2007/10/31 18:06:36 $

Scribe.perl diagnostic output

[Delete this section before finalizing the minutes.]
This is scribe.perl Revision: 1.128  of Date: 2007/02/23 21:38:13  
Check for newer version at http://dev.w3.org/cvsweb/~checkout~/2002/scribe/

Guessing input format: RRSAgent_Text_Format (score 1.00)

Succeeded: s/don't need to create/don't always need to create/
Found ScribeNick: oedipus
Found Scribe: Gregory_Rosmaita
Found ScribeNick: oedipus
Default Present: Nick_van_den_Bleeken, markbirbeck, Gregory_Rosmaita, anne, +1.408.996.aaaa, mjs
Present: Nick_van_den_Bleeken markbirbeck Gregory_Rosmaita anne +1.408.996.aaaa mjs
Got date from IRC log name: 31 Oct 2007
Guessing minutes URL: http://www.w3.org/2007/10/31-forms-tf-minutes.html
People with action items: 

WARNING: Input appears to use implicit continuation lines.
You may need the "-implicitContinuations" option.

[End of scribe.perl diagnostic output]