W3C home > Mailing lists > Public > public-xformsusers@w3.org > December 2016

Re: Core Attribute Collections

From: Erik Bruchez <ebruchez@orbeon.com>
Date: Thu, 1 Dec 2016 10:44:09 -0800
Message-ID: <CAAc0PEU=xC6DopuOoabf-gbqTZGmx4uPk3F9bg2ueMWUfG1_eQ@mail.gmail.com>
To: Steven Pemberton <steven.pemberton@cwi.nl>
Cc: XForms <public-xformsusers@w3.org>
> Well OK, but wouldn't it better to make it more explicit where context and
> model are allowed, rather than allowing them everywhere?
> At present
> <model model="m">
>    <instance model="j" context="..">
>      <data/>
>    <instance>
> </model>

They don't make sense here so I agree they shouldn't be on Common.

> and many other meaningless things are allowed syntactically, and so not
> checkable with validating parsers.
> I would prefer for @model/@context to be in Binding, because they really
> are allowed anywhere a binding is allowed, and then specifically mention
> other non-binding places where they are allowed.

But I also find they don't belong with Binding, because  @ref and @bind
really don't make sense on `xf:action` and others elements like that.

So maybe:

- we introduce a new Context Attributes including @model/@context
- Binding implies/includes Context
- we are specific on which elements Context applies when they don't support

This would include:

- var
- param
- body
- result
- item
- choices
- dialog
- case
- submission
- all action elements which don't have Binding
  - action
  - show
  - hide
  - setindex
  - toggle
  - setfocus
  - dispatch

I am not sure about:

- send
- recalculate
- revalidate
- rebuild
- refresh

The last 4 need a model, either implicitly or explicitly, but they do not
use an XPath context. In this case, @context would just be harmless and
could still be allowed.

(I went quickly through the elements and might have made mistakes.)

Received on Thursday, 1 December 2016 18:45:02 UTC

This archive was generated by hypermail 2.4.0 : Friday, 17 January 2020 19:37:47 UTC