See also: IRC log
<trackbot> Date: 19 September 2012
<scribe> Scribe: Steven
Steven: This is just a heads up.
Steven: WHo could chair, now Leigh is gone?
John: Michael SperbergMcQueen?
Steven: Hmmm. I'll ask him
Steven: More of a heads up, for people who may be interested. Definitely of interest for XFOrms
Steven: We have 2 implementations, but we didn't actually resolve to adopt
John: [discusses use case]
Nick: It may not be enough to detect changes in the UI
John: True, because the data may change under the index, without changing the index
Steven: But in that case, you can still catch that
<nvdbleek> This is also a really handy event http://wiki.orbeon.com/forms/doc/developer-guide/xforms-events#TOC-xxforms-value-changed
Steven: Blue-skying this, ideally you would have a special instance that holds the indexes, and then use a constraint.
John: True... we even say something along those lines in the discussion of the index() function
Steven: Is there a general design that we could try for?
John: THe spec text is already there.
Steven: Are you suggesting I can do a bind already?
<bind nodeset="index('myrepeat')" ...
Steven: like that?
<nvdbleek> <xf:output value="row[index('my-repeat')]/help"/>
<John_Boyer> output or bind ref=some/node[index('X')]
John: Then when the index changes, those ref and values know they have to update
<John_Boyer> in bind case, it's a dynamic dep when in ref/nodeset but not when in calculate etc (MIPs)
<John_Boyer> a repeat index is backed by an implicit instance node
John: So technically we don't
need the event
... there may be a separate issue around how many nodes are in a nodeset and how many of them are relevant
... I don't see so many repeat tables where some rows are non-relevant
Nick: We have the relevant() function in XForms 2.0
John: If you add valid() then all
the other MIPS should be functions as well
... but valid() is the hardest to do
<John_Boyer> because valid is confluence of MIPs like constraint plus things that are determined after recalculate (schema validity)
Nick: They are good candidates to add.
<John_Boyer> so if we added valid(), then there'd be no reason not to add relevant(), readonly() etc
Nick: I'll send an email proposing them.
Steven: Do we still want the events?
John: I think you talked us down
... It's just the secondary issues. that was a problem with the event version anyway.
<John_Boyer> the index() function gives us declarative updates rather than event-driven, which is weaker as you say, Steven
Steven: Anything more to discuss on this?
Nick: No, I think we agreed to this
Steven: If we had only used something other than ID...
John: What happens if we get a clash on loading the doc?
Nick: Non valid
John: In this case we *want* to allow ID clashes.
Nick: WIthout encapsulation, you
can't embed it twice.
... you can't dispatch events to them
John: Otherwise we have to fix the IDs on loading
Steven: That solves the problem, but you still can't dispatch events
John: Will this be a real
problem? Or can app designers account for it, and design
... Some people have done this to solve the problem, but we really need a deeper solution for subforms
... this a bandaid, albeit a good one.
... we need to create something smarter.\
John: Regrets for next week.
This is scribe.perl Revision: 1.136 of Date: 2011/05/12 12:01:43 Check for newer version at http://dev.w3.org/cvsweb/~checkout~/2002/scribe/ Guessing input format: RRSAgent_Text_Format (score 1.00) Succeeded: s/TH/Th/ Succeeded: s/Definitly/Definitely/ Succeeded: s/idex/index/ Succeeded: s/index()/index('X')/ Succeeded: s/DO/Do/ Succeeded: s/say/say, Steven/ No ScribeNick specified. Guessing ScribeNick: Steven_ Found Scribe: Steven Default Present: nvdbleek, [IPcaller], John_Boyer, Steven_ Present: nvdbleek [IPcaller] John_Boyer Steven_ Alain Regrets: Erik Philip Kurt Agenda: http://lists.w3.org/Archives/Public/public-forms/2012Sep/0010 Found Date: 19 Sep 2012 Guessing minutes URL: http://www.w3.org/2012/09/19-forms-minutes.html People with action items:[End of scribe.perl diagnostic output]