See also: IRC log
<Steven> Day 1
<Steven> Day 2
<Steven> Scribe: Steven
John: We need to make a
choice
... non-relevant nodes are allowed to be set with a setvalue'
how about readonly?
... Apparently setting a non-relevant node doesn't break the
model encapsulation
... but readonly does
... so the choice is saying the model creates a readonly
property that nothing else is allowed to violate
... or only an annotation that affects the controls
Steven: It seems like we have two
cow paths here, only one of which can be paved
... we need to choose the one that does the least damage
Charlie: Or restricts our future movements the least
John: We want readonly at the control level as well in the future
<Charlie> I'm more focused on trying to move us in the direction we intuit now will be the correct future direction, vs. just avoiding restricting our future options.
John: But if we want to encapsulate the model, we can call it something else
Charlie: There is an inconsistency between allowing calculate on readonly, but not a setvalue
Nick: It is about model encapsulation
Charlie: Sounds like splitting hairs
Erik: We want to provide the model with behaviour that extends the binds and so on
Charlie: I want to do with script what I can do with constraints
<ebruchez> scribe: ebruchez
John: Other example why MIPs are annotations consumed by other models is the definition: "An XForms-specific annotation to an instance data node"
Nick: It's an annotation for the
model.
... It's not a contradition, it depends who consumes it.
Charlie: What bugs me is the asymmetry between @calculate and setvalue.
Nick: @calculate coud call script via custom XPath functions.
John: Ouch.
Erik: Sometimes you want to use @calculate, sometimes setvalue.
Charlie: I am still bothered by the asymmetry.
Mark S., John is suggesting a straw poll.
John: So we have 5 in favor of readonly impacting actions, and 3 in favor of not.
Charlie: I don't think the straw poll is sufficient to decide.
<MarkS> that is too bad
Steven: So how do you suggest we decide?
<MarkS> it sounds like this is just too big for 1.1
Nick: Would be better just to leave things as they are for 1.1, i.e. vague.
Steven: I feel for Nick because he has forms that would break if we decide not to enforce readonly in actions.
<Steven> but I believe that the other option is the path of least damage
John: But he can use the @version attribute to decide what behavior to implement, so that's ok.
<MarkS> so you consign some people to never go to 1.1
Steven: I don't agree with the comment that @version solves the problem.
Nick: It means we can't move to 1.1.
John: If you are upgrading the form, you can make the changes to have the same functionality. It's not impossible, just "interesting".
Nick: readonly makes the whole subtree readonly, and it is hard to rewrite certain XPath expressions.
John: I think there exists a way, for every form, to make it work.
<MarkS> but if you have a schema with a fixed value, do you expect to be able to override that?
<MarkS> what about the model author who is not a form author?
Nick: I am not sure we can generate the XPath.
Erik: I find it funny that you can write setvalue or insert and sometimes they do something, sometimes they NOP.
John: I agree w/ Erik, [...].
Nick: Separation of model vs. UI authoring is nice.
<MarkS> Origo creates models for others to consume
<MarkS> it does not create forms
<MarkS> others do that
<MarkS> he expects redonly to be honoured
John: You can still have a model author, it's just what you expect for readonly.
<MarkS> no, he means that a node is readonly
<MarkS> no, actions are just actions, surely?
John: What you could say is that actions within the model are part of the model.
<MarkS> that is just the syntax, surely
<MarkS> if you can get the events outside the model, then you can put the actions outside the model
<MarkS> I would prefer that myself
<MarkS> I think we have good use cases for both kinds of readonly
<MarkS> and we can't really do either
Erik: If the behavior is in the model, then I don't see why the model author can't use setvalue.
<MarkS> well I do :-)
Nick: We have lots of NOPs.
<Charlie> I've changed my straw-poll vote to the hint-only option.
<MarkS> then for consistancy things like type should jut be a hint too
<MarkS> just/jut
<MarkS> If we don't know how the model works, how can anyone else be expected to understand?
Erik: I don't think we do, here setvalue can work, but it will NOPs. It's very hard to debug. It should maybe stop XForms processing then.
John: New straw poll shows 5 to 4 in favor of enforcing readonly.
Steven: We need to see who can live with either.
<MarkS> sorry
<MarkS> do what?
repeat your opinion do the IRC
<MarkS> I can't live with 2
<MarkS> I will die
<MarkS> ;-)
<Steven> Is MarkB here?
Erik: If we do enforce, it's going to be a lot of work for 1.1.
<MarkS> I still think we need mechanisms to give readonly in the sense Nick and I want, and in the sense that Erik wants, (but not contradicting model readonly)
<MarkS> there are good reasons for both
Steven: (...)
<MarkS> I can't hear John
<MarkS> and I miss his soothing voice
<Steven> I said: If we don't decide now, then we have to advise authors to assume case 1)
<Steven> since that includes 2) (and not vice versa)
John: Option 1 is heavier, but we may have to go with it.
<Steven> If we adopt 1) then to do setvalue on a readonly you have to have a calculate that says it is the same as another (non-readonly) value, and do your setvalues on that one
Erik: If we do 1), then we have the issue of MIP staleness.
Nick: You could disable deferred updates.
Uli: You can leave things out of action blocks.
John: It isn't a big problem to me to make the notes everywhere to explain how this works. If we choose 1), I would like the latitude to write these notes.
<MarkS> MarkB, you are making funny noises
Erik: Sure but it's a lot of complexity.
<MarkS> what is a lot of complexity?
<MarkS> what are you talking about now?
Proposed resolution: Readonly is an inviolate property of the model that other modules (including actions in the model) cannot ignore. Various spec edits will make that clear.
<MarkS> thanks
Uli: What about instance replacement?
John: Submission is IMO separate so will enforce readonly as well.
RESOLUTION: Readonly is an inviolate property of the model that other modules (including actions in the model) cannot ignore. Various spec edits will make that clear.
<scribe> ACTION: jboyer to make spec changes saying that readonly is an inviolate property of the model that other modules (including actions in the model) cannot ignore. Various spec edits will make that clear. [recorded in http://www.w3.org/2007/09/14-forms-minutes.html#action01]
<Steven> <br/>
<scribe> ACTION: nick to create samples illustrating how readonly affects insert and delete. [recorded in http://www.w3.org/2007/09/14-forms-minutes.html#action02]
<scribe> ACTION: nick to create samples illustrating how readonly affects instance replacement. [recorded in http://www.w3.org/2007/09/14-forms-minutes.html#action03]
<trackbot-ng> Created ACTION-409 - Make spec changes saying that readonly is an inviolate property of the model that other modules (including actions in the model) cannot ignore. Various spec edits will make that clear. [on John Boyer - due 2007-09-21].
Nick: We will need UI readonly in 1.2 or 2.0.
<MarkS> agreed
<trackbot-ng> Tracking ISSUEs and ACTIONs from http://www.w3.org/2005/06/tracker/xforms/
<inserted> Scribe: Nick
<trackbot-ng> Tracking ISSUEs and ACTIONs from http://www.w3.org/2005/06/tracker/xforms/
John: When you are in the model
you need to go up one level at least
... We solved the problem for setvalue with the context()
function
... We can defer this to a future version
... We could use the context attribute
Nick: It is prob. in the scope of XForms 1.2
John: Agrees that we need a simpler version, but that it is in scope of XForms 1.2
RESOLUTION: Defer issue 28 to XForms 1.2 http://htmlwg.mn.aptest.com/cgi-bin/xforms-issues/Model?id=28;user=guest;statetype=1;upostype=-1;changetype=-1;restype=-1
John: Write erata change revalidate only validates simple datatypes, and do structaral validation on submit
Erik: I don't know how we can implement this
John:
http://www.w3.org/MarkUp/Forms/specs/XForms1.1/index-diff.html#evt-revalidate
... You can do only datatype revalidation by doing full
revalidation and only take the datatype errors into
account
... or only do it node by node
Uli: If you do structaral validation how do you report errors where no controls are bound to
Erik: In schema you have the
notion of all= strict,skip, lax in xslt2 they go furture they
added other modes
... It is hard to only validate dataypes
<trackbot-ng> Tracking ISSUEs and ACTIONs from http://www.w3.org/2005/06/tracker/xforms/
John: You can look to the schema as type mips
<trackbot-ng> Tracking ISSUEs and ACTIONs from http://www.w3.org/2005/06/tracker/xforms/
Erik: talks about schema processors and xslt 2.0
John: We can deal with performance, because you know the change, so you could skip instances based on this
<ebruchez2> I was saying that XML schema defines xsd:any with lax/strict/any processing, and XSLT 2.0 takes this one step further by providing access to these to the styleshet author.
Erik: If you have REST you have a lot of instances for those requests
<trackbot-ng> Tracking ISSUEs and ACTIONs from http://www.w3.org/2005/06/tracker/xforms/
John: Maybe you don't attach schema to the 'important' instances
S/Maybe you don't attach schema to the 'important' instances/ Maybe you only attach schema to the 'important' instances
Proposed resoution: Shoot action 2006-05-31.5
<trackbot-ng> Tracking ISSUEs and ACTIONs from http://www.w3.org/2005/06/tracker/xforms/
Uli: schema can modify the
document while validating, e.g.: add default values
... so you maybe need recalculate and rebuild after a
revalidate
Erik: We don't enforce the validity
John: The order of validating MIPS and schema validation is important
Erik: schema validation should be ran before MIPS
John: I think schema validation
should not modify the dom
... can we shoot the action item
Erik: We don't specify clearly how validation should be performed, taking into account all the corner cases, but this should be done in the future
RESOLUTION: We
shoot action 2006-05-31.5
... Group and switch are valid targets for MIP
events
<scribe> ACTION: Johnb Update spec to reflect that group and switch are valid targets for MIP events [recorded in http://www.w3.org/2007/09/14-forms-minutes.html#action04]
<trackbot-ng> Sorry, couldn't find user - Johnb
John: Controls do nothing with
read-only MIP on group
... Controls do something with relevant MIP on group
Rafael: Maybe we need two things for 'UI level MIPS' and MIPs
John: Still agree on prev. resolution
<John_Boyer> 28<http://htmlwg.mn.aptest.com/cgi-bin/xforms-issues/Model?id=28;user=guest;statetype=1;upostype=-1;changetype=-1;restype=-1> rnhttp://www.w3.org/MarkUp/Forms/specs/XForms1.1/index-diff-20070912.html#expr-binding-expression
<John_Boyer> http://www.w3.org/MarkUp/Forms/specs/XForms1.1/index-diff-20070912.html#expr-binding-expression
Erik: We have two issues first is
of Michael, and the 5 other points
... we start with issue 142
... We need a clarification, and I should propose some text for
the spec
John: The restriction is that the form will not work well if you use dynamic dependencies
Erik: More advanced implementation can process can do beter and proces some/all dynamic dependencies
John: The dependency list is changing by value changes if the parameters of id are not fixed
erik: gives simple example that
is not allowed, predicate contains location-path
... He can process this easily
John: gives counter examples
Erik: Want a definition of dynamic dependency
John: It is in the spec -> A
dynamic dependency exists on any predicate unless all terms in
the test are "fixed", where fixed means either a constant, or a
value that will not change between operations explicitly
defined as rebuilding computational dependencies.
... We can say more
... proposes new definition : An expression is dynamically
dependent on a node or function return result if a
change to the node or function return result causes a change to the dependency
list of the expression.
Nick: How can we translate the definition to a form auther readable definition
Erik: Don't think we can
... Why do we need rebuild?
... The processor should be smart enough
... We need to say what will happen if you violate this
John: The form might not work as
expected
... We figured out that delete and insert should do a
rebuild
Erik: I think that the dependency algorithm is over specified
John: The master dependency ensures that calculation is done in a linear time.
<MarkS> I have to go for a while
<MarkS> be back later
<wellsk> morning all
<Steven> hi
Nick: You can skip parts in the rebuild if you are smart enough, as long as the result is the same as specified in the spec
John: There is simple difference
between visiting and referencin a node
...
<loan-record><principal></principal><duration></duration><interest></interest><mont-pay></mont-pay><total></total><interest-paid></interest-paid></loan-record>
... if you have calculates on the elements, you vistit the
multiple child nodes for name element tests, but only
references one element
... if you don't make a difference between visiting and
referencing node you get false circular dependencies
... We execute all the xpath expressions to know what nodes are
referenced and are in the dependency graph
Erik: We were going to do static analysis, and that will not work correctly for John's example
Nick: Explains an algrorithm that solves the referenced node problem using a strandard xpath engine
Erik: John's definition is more clear, and the section should not start with a negative sentence
RESOLUTION: We accept that a better definition is needed and definition for what happens when you use a dynamic dep. in the model
<scribe> ACTION: jboyer Correct section 7.3.1 Dynamic Dependencies with spec text based on the notes of issue 142 [recorded in http://www.w3.org/2007/09/14-forms-minutes.html#action05]
<trackbot-ng> Created ACTION-410 - Correct section 7.3.1 Dynamic Dependencies with spec text based on the notes of issue 142 [on John Boyer - due 2007-09-21].
Erik: Why do we say By default, all binding expressions refer to the first instance
within the context model. This behavior can be changed with the
instance() function.
John: The sentence is wrong
... It is redundant, so get rid of it
RESOLUTION: Remove By default, all binding expressions refer to the first instancewithin the context model. This behavior can be changed with the instance() function."
Erik: Section "7.5.2 Model
Binding Expressions" what is a Model Binding Expressions?
... Binding expressions do not always throw binding exception
but also compute exceptions?
<ebruchez> http://lists.w3.org/Archives/Public/www-forms-editor/2007Apr/0059.html
John: Every expression in a bind
is a Model Binding Expressions, and by consequence the dynamic
depency limition apply to them
... we have computed expression
<John_Boyer> Proposed Resolution: For issue 41 point 3, a computed expression is not a kind of model binding expression AND clarify that dynamic dep. issues affect both model binding expressions and computed expressions
RESOLUTION: For issue 41 point 3, a computed expression is not a kind of model binding expression AND clarify that dynamic dep. issues affect both model binding expressions and computed expressions
<scribe> ACTION: jboyer : implement issue 41 point 3, a computed expression is not a kind of model binding expression AND clarify that dynamic dep. issues affect both model binding expressions and computed expressions [recorded in http://www.w3.org/2007/09/14-forms-minutes.html#action06]
<trackbot-ng> Created ACTION-411 - : implement issue 41 point 3, a computed expression is not a kind of model binding expression AND clarify that dynamic dep. issues affect both model binding expressions and computed expressions [on John Boyer - due 2007-09-21].
<scribe> ACTION: jboyer :Why does Section "7.5.3 UI Binding Expressions": [recorded in http://www.w3.org/2007/09/14-forms-minutes.html#action07]
<trackbot-ng> Created ACTION-412 - :Why does Section \"7.5.3 UI Binding Expressions\": [on John Boyer - due 2007-09-21].
contain implement issue 41 point 3, a computed expression is not a kind of model binding expression AND clarify that dynamic dep. issues affect both model binding expressions and computed expressions
<Steven> ACTION 7=
<scribe> ACTION: jboyer :Why does Section "7.5.3 UI Binding Expressions": [recorded in http://www.w3.org/2007/09/14-forms-minutes.html#action08]
<trackbot-ng> Created ACTION-413 - :Why does Section \"7.5.3 UI Binding Expressions\": [on John Boyer - due 2007-09-21].
<Steven> ACTION- 7
<Steven> ACTION- 8
Section "7.5.3 UI Binding Expressions":
Erik: Why contain Section "7.5.3 UI Binding Expressions": "Dynamic dependences are allowed in UI binding expressions based on
the conformance profile."
John: I will take care of 4.b when processing 1
RESOLUTION: We allow dynamic dependencies in the UI
Erik: Section "4.3.4 The xforms-refresh Event": says that bindings should be reevaluated as necessary, what does this mean?
John: It is not enough if you use
the index() in the ui
... the index() should trigger a pseudo dependency
... The text dynamic dependencies in the UI are supported
implies that it will update the UI when the index is
changed
RESOLUTION: change "3. All UI bindings should be reevaluated as necessary." to "3. All UI bindings should be reevaluated."
S/3. All UI bindings should be reevaluated./3. All UI bindings should appear as if they were reevaluated./
Erik: In 6. Section "4.6.7 Sequence: Value Change": "xforms-refresh performs reevaluation of UI binding expressions" but maybe it is OK now, that we addressed point 5
<John_Boyer> adjourn for lunch
RESOLUTION: We will not change the text related to xforms-refresh in 4.6.7 Sequence Value Change
<John_Boyer> back from lunch
<John_Boyer> http://www.w3.org/TR/2007/WD-SMIL3-20070713/smil-state.html
http://www.w3.org/TR/2007/WD-SMIL3-20070713/smil-state.html#adef-expr
Charlie: They added function, that can be used in the expr languages
John: Can they acces instance data
Charlie: That is defined in the next section
Steven: Expression is a poorly name
Uli: It is more like relevant
John: It is more like if
Steven: They should use a better name
John: People don't like negative name
Uli: Maybe use active, show, ...
John: They don't say anything about evaluation context
Charlie: They say something about
it som sections before
... there is informative text 'The context in which the
expressions are evaluated is as follows:'
Uli: Note that "out of scope" here refers only to the scope of the State Test module, which only involves scalar expressions, the expression context is more fully defined in the Language Profile because it includes the State Test module as well.
Steven: They say not which language to use, so the language needs to define the context
Nick: They have a language attribute
Charlie: It is on state
... The SMIL 3.0 Language Profile specifies that XPath 1.0 is
used as __the__ expression language.
John: They need links to supported Language Profiles
Charlie: No distinction between model and instance
Nick: can in xforms setvalue set an attribute that doesn't exists?
Charlie: Do they support attributes?
Nick: It maybe dependant on the language you use
John: newvalue is our insert?
Charlie: Yes
John: where is the delete action?They don't have it
Charlie: The name state is confusing if used in combination with State Chart XML
Steven: It is name spaced
John: New value is like insert
<klotz> like about 10% of insert
Nick: Maybe they have languages that only support flat space (just like variables)
John: So position should be
defined in language profile
... if in xpath points to no element how do you know where to
create the element
Charlie: The section is not well structured
John: where are they going to
insert the name element?First child?
... Is name an XPath
Charlie: Do they only support elements?
15.6.5 Data Model
Charlie: Why does it appear here
(15.6.5 Data Model)
... They could use a lot of more examples
Nick: There is a lot informative
John: They state only contains instance
Charlie: They have no seperate model an instance
Nick: Why is everything in 15.6.6 Data Model Events informative
Charlie: Let's asume that it is
normative
... they only have two events
... Do you get a statechange on a new value
John: If so is ref the ref+name of newvalue
Charlie: Are it DOM events?
<John_Boyer> and third, if they buy our point about needing a delete, then stateChange will probably have to tell what node lost a child but not which child (because it has been deleted)
Charlie: When do you get a
contentControlChange event, do you always get both
events?
... Should SMIL be included in a host language?
Steven: It has a root element
John: distinct xmlid
__namespace__ should be distinct xmlid __space__
... I want setvalue, insert, delete and a submission module on
top
Charlie: The SMIL StateSubmission
Module only says that they use XForms submissions
... What did they do with our events?
John: Could you capture an xforms-submit-done event and do a setvalue
Steven: Do they allow more then one submission element?
Charlie: How do fire a submission
John: We are in front of them,
couldn't they use XForms 1.1 submission
... submisison is outside state (read instance) like our
submission is out of our instance but inside the model
Charlie: Is this the way to go, just copy past stuf you want out of our spec?
John: They only support one
state, so submission will always replace the state
... The classic search terms example is not supported
Charlie: They are doing a profile, but maybe they are 'borrowing' not enough
John: Solution they could use only one state, if they add the target attribute of XForms 1.1
Charlie: We have been talking of AVT's for years, and they just add it without saying much about it
John: How do they activate the send
Nick: send is activated by they timing construct
John: Is submit synchronous
Charlie: You need async in a
multimedia env.
... submission is new in SMIL 3.0
<John_Boyer> in xforms 1.1, submission is asynch by default
Charlie: maybe they think it is a zero duration action, like writing to a file
<John_Boyer> which means that the result won't be available when the send finishes.
<John_Boyer> If they say synchronous, then it blocks, which may not be wha they want for multimedia presentation
<John_Boyer> if asynch, then they will need to be able to process xforms-submit-done and xforms-submit-error events
Charlie: They point to xslt for
AVT
... if you have async events what is the relation between
events and the normal timed execution, of actions
John: the events are queued in
XForms
... The stepper does it execute the events?
... Do we want to meet them om the TP?
<Steven> http://www.w3.org/2002/09/wbs/34786/MGM07/
http://www.w3.org/2002/09/wbs/35125/TPAC2007/
<scribe> ACTION: jboyer to ask Voice Browser Working Group to meet us at the TPA [recorded in http://www.w3.org/2007/09/14-forms-minutes.html#action09]
<trackbot-ng> Created ACTION-414 - Ask Voice Browser Working Group to meet us at the TPA [on John Boyer - due 2007-09-21].
John: I will make the agenda and Erik will chair the next meeting
Steven: Some people said it would
be a light meeting at the TPA
... We need more comitment from people, to come the f2f
meetings
John: We pre-alocated trips
Steven: remote confernce turnout was low
John: w3c says that had to go to 3 out of 4 meetings
<Steven> 2 out of 3
John: We can do 3 meetings a
year
... If we go down to 3, we will slow down
<Steven> Proposed 31 Jan/1 Feb (XHTML), 4-6 Feb Forms, North America (Possibly North Carolina)
<Steven> Proposed 1-2 May 1/3 (XHTML), 5-7 May (Forms), Amsterdam
<Steven> Proposed 10-11 July (XHTML), 14-16 July (Forms) Victoria, Canada
<Steven> Technical Plenary 20-25 October, Paris, France
<Steven> http://www.w3.org/MarkUp/Forms/wiki/FaceToFace
Steven: Do we want to continue to offer remote
John: It is hard to go out to f2f meetings
Charlie: You have the time zone differnce
<wellsk> Reservations are confirmed for RTP, NC have been made for the Jan 3-Feb 6
Steven: You can do the time difference when you stay local
<wellsk> My two sense, attending remote is difficult -- I'd rather be f2f
<Steven> Jan 31 I hope wellsk
<wellsk> yep
<wellsk> Jan 31-Feb 6
<wellsk> thurs/Frid/M/T/W
<Steven> ah! GOod!
<Steven> I'd forgotten that
John: De we provide call in service in future meetings?
Steven: What do we want?
Charlie: Maybe we should try a f2f with everybody on the phone?
<klotz> as long as steven brings a mike
Erik: It is hard
<Steven> :-S
<MarkS> what about telepathy?
<MarkS> that would be nice and quiet
John: We should stay with 4
meetings a year, should allow dail in for invited
experts,
... We had a good f2f
John: We going to have future
versions
... An editor has to go to ALL f2f meetings
... You have to spend more time, and prob. in your spare
time
... Erik is a possible one
Erik: The commitment is great
John: our charter only says
XForms 2.0
... We will need to update the charter schedule on the homepage
at some time
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/witha/with a/ Succeeded: s/aout/about/ Succeeded: s/CH/Ch/ Succeeded: s/STeven/Steven/ Succeeded: s/encp/encap/ Succeeded: s/CH/Ch/ Succeeded: s/liue/lie/ Succeeded: s/Somtimes/Sometimes/ Succeeded: s/what do you/how do you/ Succeeded: s/if/of/ Succeeded: s/somthing/something/ Succeeded: s/feature/future/ Succeeded: s/In schema you have the notion of all= strict,skip, in xslt2 they go furture they added lax/In schema you have the notion of all= strict,skip, lax in xslt2 they go furture they added other modes/ Succeeded: s/Roger/Rafael/ Succeeded: s/sentance/sentence/ Succeeded: s/wath/what/ Succeeded: s/defintion/definition/ Succeeded: s/ we introduce the trem compute expressions/ we have coputed expression/ Succeeded: s/coputed/computed/ Succeeded: s/Erik: /john :/ Succeeded: s/as they/as if they/ Succeeded: s/expression/expr/ Succeeded: s/Charley/Charlie/G Succeeded: s/clasic/classic/ Succeeded: i/TOPIC: Issue 28/Scribe: Nick Succeeded: s/Setven/Steven/ Succeeded: s/comitment/commitment/ Found Scribe: Steven Inferring ScribeNick: Steven Found Scribe: ebruchez Inferring ScribeNick: ebruchez Found Scribe: Nick Inferring ScribeNick: Nick Scribes: Steven, ebruchez, Nick ScribeNicks: Steven, ebruchez, Nick Present: John Uli Charlie Steven Erik Nick Rafael Rogelio MarkS Leigh MarkB Agenda: http://lists.w3.org/Archives/Public/public-forms/2007Sep/0054 Got date from IRC log name: 14 Sep 2007 Guessing minutes URL: http://www.w3.org/2007/09/14-forms-minutes.html People with action items: jboyer johnb nick spec update[End of scribe.perl diagnostic output]