On 7/15/09 5:56 PM, Jean-Claude Dufourd wrote:
Dear Robin,
Thanks for your email.
Robin Berjon a écrit :
There would then be a good case for the WG to spend some time on
devising a proper workaround.
Anyone sharing my opinion that the widget update feature will be
consistenly implemented (even if optional) ?
You seem to be mixing up Widget Updates and the update() method, is
that the case?
JCD: I understood the widget.update() method being discussed to be the
scripted version of the Widget Updates mechanism.
No. Read the spec. It says it's just a way of checking if an update is
available (by asking the UA to check) and _not_ a way for a widget to
update itself. Apples and oranges, as they say! (no pun intended)
Please tell me if I am wrong. The current draft seems to still say that.
You are wrong. The current draft does not specify anything normative
about the behavior of widget.update().
And I am really not sure that a script-triggered version of the update
There has _NEVER_ been a "script-triggered version of the update" or any
such functionality specified by the Working Group. To imply otherwise is
misleading.
mechanism can be discarded off-hand.
Yes it can: I discarded it from the spec.
The point is moot. The issue has been resolved. Widget.update() does not
exist (and has absolutely nothing to do with the PAG or whatever:
widget.update was deemed useless by the Working Group long ago. We've
just been too busy to publish a new draft without widget.update()).
So, please stop bringing up this closed issue.
As the Working Group awaits the findings of the PAG on how to proceed
with Widget Updates, I find it inappropriate that you continue to
discuss this matter here.
Once the PAG findings are made public, and the WG has a chance to
respond and act, then, if you have concerns, raise them here.
Kind regards,
Marcos
