See also: IRC log
<trackbot> Date: 12 November 2014
<ebruchez> can't seem to be able to join, Zakim keeps disconnecting me
http://www.w3.org/2014/10/31-forms-minutes.html
Steven: A successfull session. The issues I've split off as separate agenda items
http://lists.w3.org/Archives/Public/public-forms/2014Oct/0041.html
Steven: Opinions?
Erik: This is an attr on
submission. XForms namespace is only relevant if the instance
is inline
... the main problem is not that the namespace is kept around
but that it is in scope
... there might even be 10 around still in scope, polluting the
instance.
... XSLT has an exclude attr, which we added as an
extension.
... both make sense in general.
... You want a clean data model
Steven: What do think of Alain's suggestion of a default includenamespaceprefixes=""
Erik: I wouldn't want it.
... the real problem is the extra namespaces that you don't
want around.
Alain: Users can't understand why the extra namespaces are there. We shouldn't distinguiosh whether instances are inline or not
can someone else scribe?
My phone has died.
I'll rejoin
Steven: "If absent, all namespace nodes present in the instance data are considered for serialization"
https://www.w3.org/MarkUp/Forms/wiki/XForms_2.0#The_submission_Element_2
Erik: This is the XPath data
model.
... there are namespaces in scope
Steven: In scope yes, but are they actually in the instance?
Erik: You can tackle it in two
ways.
... filter out the non-useful ones, or exclude them
altogether.
... If I have an external instance, which specifies namespaces,
even if they are not used, when I serialise out, if we change
submission then I will lose those
<alain> Sorry, I have to leave now...
Philip: Is "present" the same as "in scope"?
Steven: That's my question
https://www.w3.org/MarkUp/Forms/wiki/XForms_2.0#The_xforms-submit_Event
Steven: Look at the last example before that link. The two namespace nodes that have been serialised are not "in the instance data" by my reading, so I disagree with the example.
<ebruchez> 3.2.2 The instance Element
<ebruchez> "If the initial instance data is given by inline content, then instance data is obtained by first creating a detached copy of the inline content (including namespaces inherited from the enveloping ancestors)"
Steven: So the important words
are "instance data" (not just "instance")
... so I agree that the namespace nodes are really there, and I
agree with Alain that this is not very user friendly.
Erik: It is cleaner to say that
the instance data is clean from the beginning, rather than
having to clean it up later.
... and we should be able to roundtrip external instances
Steven: Absolutely agree. The problem is with all the extras that we get from the host language.
Erik: It comes down to an
incompatible change somewhere.
... maybe 3.2 is the best place
... we can look at XSLT to see what
Steven: Within W3C it is understood that a number change (2.0) means expect incompatible changes
Erik: I think there are few cases
where this would really make a difference.
... I don't have a problem with making the change on the
instance element level.
... I would have a problem with changing it at the submission
level.
<scribe> ACTION: Erik to investigate dealing with excess namespace nodes in section 3.2 the instance element [recorded in http://www.w3.org/2014/11/12-forms-minutes.html#action01]
<trackbot> Created ACTION-1986 - Investigate dealing with excess namespace nodes in section 3.2 the instance element [on Erik Bruchez - due 2014-11-19].
http://lists.w3.org/Archives/Public/public-forms/2014Oct/0045.html
http://lists.w3.org/Archives/Public/public-forms/2014Nov/0000.html
Steven: I added @name and @value
to the header element
... you suggested that no header should be produced if there is
no name
... seems good
... you also suggested that we should clarify the evaluation
context.
... I agree
<scribe> ACTION: Steven to update header@name @value to include Erik's comments [recorded in http://www.w3.org/2014/11/12-forms-minutes.html#action02]
<trackbot> Created ACTION-1987 - Update header@name @value to include erik's comments [on Steven Pemberton - due 2014-11-19].
http://lists.w3.org/Archives/Public/public-forms/2014Nov/0002.html
Steven: What to do about
email.
... we don't want to exclude any formats that may still be in
use.
Erik: There seems to be a
disparity with reality
... in our implementation we stopped using regexes
... there is code out there that validates email
addresses.
... we actually use code to validate, not a datatype
Steven: Are you saying that you have to use code, and that regexps are not enough?
Erik: We found it easier to use someone else's code.
Steven: So we have to reverse engineer this?
Erik: Not sure.
<ebruchez> steven@com
Steven: XForms should not exclude correct email addresses. I don't know if steven@com is right or not; we should exclude it if it is wrong, and allow it if it is OK.
<ebruchez> the other case is john@example.org/
<ebruchez> (ending with a "/")
Steven: This needs more research. Either our regexp is wrong, and then it should be excluded.
<scribe> ACTION: Steven to contact WAI about correct international email addresses [recorded in http://www.w3.org/2014/11/12-forms-minutes.html#action03]
<trackbot> Created ACTION-1988 - Contact wai about correct international email addresses [on Steven Pemberton - due 2014-11-19].
[ADJOURN]
This is scribe.perl Revision: 1.140 of Date: 2014-11-06 18:16:30 Check for newer version at http://dev.w3.org/cvsweb/~checkout~/2002/scribe/ Guessing input format: RRSAgent_Text_Format (score 1.00) Succeeded: s/;/'/ Succeeded: s/tio/tion/ Succeeded: s/nn/n/ Succeeded: s/cc/c/ Succeeded: s/belm/blem/ No ScribeNick specified. Guessing ScribeNick: Steven Inferring Scribes: Steven Default Present: Steven, pfennell, [IPcaller], ebruchez Present: Steven pfennell [IPcaller] ebruchez Alain Agenda: http://lists.w3.org/Archives/Public/public-forms/2014Nov/0003 Found Date: 12 Nov 2014 Guessing minutes URL: http://www.w3.org/2014/11/12-forms-minutes.html People with action items: erik steven[End of scribe.perl diagnostic output]