See also: IRC log
https://lists.w3.org/Archives/Public/public-xformsusers/2016Jan/0019.html
Steven: What was your opinion of their differences?
Erik: I didn't have time to fully
analyse it, but we can quickly run through the list
... 1. Repeat - the intent makes sense, adding and remove item
controls by default, number of items, template.
... the way that they do it is a different question
... not sure of the UI
... you can turn off the add/remove possibilities
... The other approaches to doing that, for instance ours, is
to *add* the new parts
... and there are other issues.
Steven: I like the approach of having these things supplied for you: they are difficult to do right.
Erik: True.
... Orbeon has the 'grid' component, that does this and
more
... such as reordering.
... min number of items, maximum number, initial number.
... Do you want to 'freeze' the initial values?
... quite a number of things you might want.
... I like a layered approach, such as using components.
... I prefer this to be done as a component rather than at the
level of repeat.
... 2 is the handling of templates.
... in the old days we did it differently, no source attribute,
using one of the elements from the data
... which meant jumping through hooops when deleting the last
item.
... and made submission tricky
... so we updated the insert action to make this easier.
... they have done something more like the first option, but
with a magic attribute
... I'm less charmed by that
s/^.. /... /
<ebruchez> For reference, our grid component: http://doc.orbeon.com/form-runner/component/grid.html
Steven: I do like the idea of
associating messages to binds
... because ti keeps the messages close to the tests.
Erik: There are many extensions
that people have with binds Betterforms have written about
it.
... they support something similar.
<ebruchez> https://betterform.wordpress.com/2012/12/19/taking-mips-to-the-next-level/
Erik: We can link from bind to
message, but the message is still in the UI.
... (I'll send a link later)
... there are issues with I18N
Steven: Yes, but no different from int and alert etc in the UI
Erik: True
Steven: And then 3. Extra XPath functions
Erik: Some inconsistencies
here.
... they have some XPath 2 and even one from XPath 3, plus
their own functions. Didn;t have time to analyse what they were
doing.
<ebruchez> tokenize(value, '\s+') = 'my-value'
Erik: The selected function is
less needed with XPath 2.
... but I can see the point.
Steven: Thanks for doing that, some interesting things to think about for the future.
https://lists.w3.org/Archives/Public/public-xformsusers/2016Jan/0018.html
Steven: Thanks for the comments
on 1-3.
... I have made the corrections, and I see a couple of
typos
Alain: Section 1.1, shouldn't say "the successor to HTML Forms"
Steven: Right, I'll fix
that
... One thing that struck me today, and I'm not sure if we
discussed before, but in the CSV examples,
... if the name is a non-XML name, we should record the real
name (with name="") to be able to round-trip them.
Alain: I agree.
Steven: next week, sections 1-3 if you haven't already read it, and section 4-6. The week after, section 7.
[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/;/'/ Succeeded: s/componenet/component/ Succeeded: s/ump/jump/ FAILED: s/^.. /... / Succeeded: s/"/:/ Succeeded: s/FO/Fo/ No ScribeNick specified. Guessing ScribeNick: Steven Inferring Scribes: Steven Present: Erik Alain Philip Nick(remote) Steven Regrets: None Agenda: https://lists.w3.org/Archives/Public/public-xformsusers/2016Jan/0020 Found Date: 20 Jan 2016 Guessing minutes URL: http://www.w3.org/2016/01/20-forms-minutes.html People with action items:[End of scribe.perl diagnostic output]