See also: IRC log
https://lists.w3.org/Archives/Public/spec-prod/2016JanMar/0036.html
https://w3c.github.io/tr-design/src/sample.html
Steven: There is a new TR style,
required from 1 March
... so I will look into implementing it.
<scribe> ACTION: Steven to implement new TR style [recorded in http://www.w3.org/2016/02/24-forms-minutes.html#action01]
<trackbot> Created ACTION-2054 - Implement new tr style [on Steven Pemberton - due 2016-03-02].
report back
http://www.w3.org/2016/02/17-forms-minutes.html#action02
Steven: Thank you for doing that
Alain (http://www.agencexml.com/direct/trimmed/typed-hello.xml)
... What did it involve
Alain: Added a type
... input field can be edited, but the instance value is not
necessarily updated.
Steven: What did you have to change in the implementation?
Alain: I had to separate the
inpur node and the instance value
... look at the source
Steven: The implementation has to recognise the type specially?
Alain: Yes.
Erik: Who does the trimming?
Alain: It's in incremental mode
Erik: Yes, but at which point?
Alain: When refreshing.
... it compares values after removing spaces.
Steven: SO a user couldnt add a type like this?
Alain: No, the implementation recognises the type.
Steven: So if you wanted a type to remove all spaces, and another type to reduce all spaces to single ones, you would have to introduce two more types and recognise them specially in the implementation
Alain: Yes.
Erik: When I type, something gets added to the data, and then I type a space, that gets written to the data, and on a refresh it gets trimmed?
Alain: Yes
Erik: The value for the output is trimmed on refresh, and the value of the input is not trimmed if it has focus?
Steven: So you have special knowledge on the input and output?
Alain: No, not on the output
Erik: If I tab out of the field?
Steven: Then the input value gets trimmed as well?
Alain: Yes.
Erik: We do it differently.
... The value in the input gets written to the data, and gets
trimmed at the next recalculate
... On refresh, the control just gets the value.
... Something special would need to be done to deal with
incremental
Steven: Yoou said that you have a whitespace flag on the value?
<ebruchez> xxf:whitespace="preserve | trim"
Erik: It's a new MIP
Steven: Is that better than a type?
Erik: Not sure. You can add this
to any type.
... for instance, an email type can have whitespace trim.
Steven: If it were a MIP, we could add other values, I'm thinking of a credit-card value, that would remove all spaces.
Erik: Normalise spaces would be
another.
... collapse spaces.
... it is annoying that in 2016 there are still websites
telling you not to type spaces.
... there are other possible transformations, where whitespace
is a subset, but an important one.
Alain: Phone numbers with dots and dashes, same problem.
Steven: I like the MIP solution,
because you can apply it to any type.
... would it be more difficult to do it in your implementation
Alain?
Alain: No.
Erik: You can actually see it as a calculation.
Alain: It's related to
'collations'.
... for instance uppercase to lowercase.
<ebruchez> calculate="normalize-space()"
Steven: It could be done with a function that removes the spaces, and a calculate, with two values, one for input and one for output.
<ebruchez> calculate="xxf:trim()"
<data> <in/><out/></data>
<bind ref="out" calculate="trim(../in)"/>
<input ref="in"/>
<output ref="out"/>
Erik: Clearly.
... but this would be heavy for a large form.
Steven: you were suggesting a calculate that calculates itself.
Erik: Yes.
... but this has problems for incremental mode.
... that's a drawback.
Steven: We don't have anything at present in XForms that does anything like that...
Erik: I don't think so; I use it in forms, though less now that we have the whitespace MIP.
Steven: Let's take a week to let this all sink in.
Alain: Yes. Though I think it may
be a more general problem.
... It's a two-way problem.
<ebruchez> <input format="..." unformat="..." ...>
Erik: We have things like this in
our implementation for dealing with the input/output
differences
... flexible, but you need to write the expressions every
time
... better really to have it at the model level.
http://www.w3.org/2016/02/17-forms-minutes.html#action04
Steven: Erik, anything on this?
Erik: I sent an email:
https://lists.w3.org/Archives/Public/public-xformsusers/2016Feb/0040.html
... we can use the other exceptions as a model for how to solve
it. Not difficult.
... See mail.
Steven: Looks convincing, anyone got any problems with it?
Alain: No
Philip: No
<scribe> ACTION: Erik to implement replace link-exception [recorded in http://www.w3.org/2016/02/24-forms-minutes.html#action02]
<trackbot> Created ACTION-2055 - Implement replace link-exception [on Erik Bruchez - due 2016-03-02].
Steven: Organising XML Days in AMsterdam later in the year, sometime in the autumn. More details to follow.
[ADJOURN]
This is scribe.perl Revision: 1.144 of Date: 2015/11/17 08:39:34 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/TH/Th/ Succeeded: s/iit/it/ Succeeded: s/calculates/calculate/ Succeeded: s/calculates/calculate/ No ScribeNick specified. Guessing ScribeNick: Steven Inferring Scribes: Steven Present: Steven Alain Erik Philip Agenda: https://lists.w3.org/Archives/Public/public-xformsusers/2016Feb/0037 Found Date: 24 Feb 2016 Guessing minutes URL: http://www.w3.org/2016/02/24-forms-minutes.html People with action items: erik steven[End of scribe.perl diagnostic output]