<pfennell> I seem to have audio issues.
Steven: Good response to the
XForms lecture there.
... slides here: https://homepages.cwi.nl/~steven/Talks/2018/09-12-xml/
https://markupdeclaration.org/
<ebruchez> https://github.com/markupdeclaration
Steven: Please get involved so that we start getting momentum.
https://homepages.cwi.nl/~steven/ixml/ixml-specification.html
Steven: Please give feedback.
https://lists.w3.org/Archives/Public/public-xformsusers/2018Sep/0000
Steven: If we accept the idea,
I'm willing to spec it up.
... OK?
Erik: For group, adding an appearance?
Steven: Yes
Erik: OK
<scribe> ACTION: Steven to spec up collapsing sections
<trackbot> Created ACTION-2196 - Spec up collapsing sections [on Steven Pemberton - due 2018-09-26].
identification
https://lists.w3.org/Archives/Public/public-xformsusers/2018Aug/0029
Steven: I rather like this way of
doing it (even if I say so myself). Much simpler to understand
the current method.
... Also divides the cases of IDs that don't exist, and those
that are just currently inaccessible.
Erik: Let me type an
example
... this is how to resolve IDs.
Steven: It also deals with the difference between IDs that don't exist, and IDs that exist but, for instance, are in a repeat with zero elements.
<ebruchez> <repeat>
<ebruchez> <input id="my-input"/>
<ebruchez> <output id="my-output">
<ebruchez> <setfocus
<ebruchez> event="custom"
<ebruchez> control="my-input"/>
<ebruchez> </output>
<ebruchez> </repeat>
Steven: Why not send an
email.
... In your case, the control in the current index would get
the focus.
Erik: My reading is that it goes to the 'closest' ID, and in this case it would go to the next indexed version.
Steven: That surprises me. I don't see that in the spec.
Erik: I need to look again, but reaching the 'closest' one, regardless of the repeat index
Steven: Please send a real-life example, we need to resolve what we expect from the text.
<scribe> ACTION: Erik to send example of IDREF identification
<trackbot> Created ACTION-2197 - Send example of idref identification [on Erik Bruchez - due 2018-09-26].
https://lists.w3.org/Archives/Public/public-xformsusers/2018Aug/0016
https://lists.w3.org/Archives/Public/public-xformsusers/2018Aug/0023
Steven: This replaces a dozen functions with a single one.
Erik: Is the proposal to return
an XML document or a map?
... you need to access the parameters
Steven: That's in my proposal.
Erik: I like the functions, as easier to use.
Steven: I'm not sure if I'm convinced.
Erik: The function names are at
least checked statically, so you get some feedback. Paths don't
give you that
... so there are some benefits to the functions.
... It's an API philosophy.
... Something I don't like in XPath3 is the parse function that
gives a blob of XML.
... in the end it's balancing tradeoffs and preferences.
Alain: I do like maps.
Steven: I like them too, but we don't have them in XForms 2, and I don't think we should add them just for one function.
Philip: Agreed
Steven: If we don't agree on replacing the 12 functions with one, then we leave as is.
https://lists.w3.org/Archives/Public/public-xformsusers/2018Jun/0033
https://lists.w3.org/Archives/Public/public-xformsusers/2018Sep/0004
Steven: So what happens when you
bind to a nested bind element.
... what do we think my example should do?
Erik: I believe that we implement
what the spec says, but I'm not sure that that is what should
be done.
... if there's no result, you get a null result; if you get one
result, that is the result; if you get two results, you're back
to a null result. This is not reasonable.
... counterintuitive
<bind id="a" ref="foo/bar[expr]"/>
<bind ref="foo"><bind id="a" ref="bar[expr]"/></bind>
Steven: I would expect those to work the same
Erik: I think that's the case.
Steven: For a <repeat bind="a">...
Erik: I think you had that in the
email, and it's not the same as it now is.
... It depends on the number of foo elements, and that's
weird.
Steven: So what to do we do about it.
<scribe> ACTION: Erik to propose text for selecting items from a nested bind
<trackbot> Created ACTION-2198 - Propose text for selecting items from a nested bind [on Erik Bruchez - due 2018-09-26].
https://lists.w3.org/Archives/Public/public-xformsusers/2018Sep/0002
Steven: This is a use-case I have a lot
Alain: OK with me.
Philip: Agreed.
Erik: The only thing is what events would be dispatched in case of error.
Steven: instance-fallback
Erik: A shame we can't map that straight to submission
Steven: What event do we get then?
Erik: submit-error
Steven: Oh that's a notification
too
... I can live with that
Erik: Good for consistency.
... a shame it's got submit in the name, but conceptually
that's what it is.
Steven: Any objection to using
"submit-error"?
... Hearing none...
<scribe> ACTION: Steven to add instance fallback to spec
<trackbot> Created ACTION-2199 - Add instance fallback to spec [on Steven Pemberton - due 2018-09-26].
https://lists.w3.org/Archives/Public/public-xformsusers/2018Sep/0007
Steven: Done.
... Let me know if there's any problem.
https://lists.w3.org/Archives/Public/public-xformsusers/2018Jun/0020
Steven: This is suggested text
for catching value-changed in the model.
... Other issues: why are notification events sent to different
controls
... Do we need to catch the other notification events in the
model?
Erik: For the UI part, we have a
tree of control, and some can hold values and some not.
... but I agree it is confusing.
... if a node changes value, we set a flag (in previous
versions of the spec), and then on refresh send the event based
on the flag
... and also from the point of view that it changes value and
back before refresh. Is that a value-change?
Steven: XSLTforms does not send the event if a value is set to the same value.
Erik: Binds don't record anything
at refresh time.
... there might be a little more to it than saying "it also
happens to the binds".
Steven: How do we move forward?
Erik: I do have the same desire
to have value-changed in the model.
... can we graft the cotrol code to the model. I'm not
sure.
Philip: Is there any overlap with
mutation events?
... a value changed event in the model, is like a mutation in
the tree.
Erik: Let's think more.
Steven: Talk more next week on this.
[ADJOURN]
This is scribe.perl Revision: 1.152 of Date: 2017/02/06 11:04:15 Check for newer version at http://dev.w3.org/cvsweb/~checkout~/2002/scribe/ Guessing input format: Irssi_ISO8601_Log_Text_Format (score 1.00) Succeeded: s/SPI/API/ Succeeded: s/AG/Ag/ Succeeded: s/FO/Fo/ Succeeded: s/hte/the/ Succeeded: s/model./model?/ Succeeded: s/calue/value/ Succeeded: s/Talks/Talk/ Succeeded: s/thos to/those to/ Succeeded: s/DO we/Do we/ Present: Alain Jorge Steven Erik Owen Philip No ScribeNick specified. Guessing ScribeNick: Steven Inferring Scribes: Steven Agenda: https://lists.w3.org/Archives/Public/public-xformsusers/2018Sep/0011.html Found Date: 19 Sep 2018 People with action items: erik steven WARNING: IRC log location not specified! (You can ignore this warning if you do not want the generated minutes to contain a link to the original IRC log.)[End of scribe.perl diagnostic output]