W3C home > Mailing lists > Public > www-forms@w3.org > May 2006

Deferred action behaviour

From: Allan Beaufour <beaufour@gmail.com>
Date: Thu, 18 May 2006 10:38:13 +0200
Message-ID: <90d6cb0e0605180138x6eafa111tcc1be272e618141c@mail.gmail.com>
To: www-forms <www-forms@w3.org>

At the teleconf. "deferred action behaviour" was mentioned while
talking about events.

I've always thought that <setvalue/> follows the "Sequence: Value
Change", which sets the instance data, and then dispatches
xforms-recalculate, xforms-revalidate, xforms-refresh.

So I then thought that inside an action block, the dispatching of
these events would be delayed until after the last action. So these
<setvalue .../>
  <setvalue .../>
would be equivalent.

But, re-reading about actions, I suddenly see this:
"Each outermost action handler can be thought of as having a set of
Boolean flags, initially false, to indicate whether each of the
actions rebuild, recalculate, revalidate, and refresh are required
upon termination of the outer action handler."

These actions "bypass" the normal event system (which I think makes
sense). But shouldn't the <action/> dispatch events, and not use the
actions? With the current wording (1) would result in events, and (2)
would not.

It sounds wrong to me. I might have missed something though?

... Allan
Received on Thursday, 18 May 2006 08:38:19 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 21:36:17 UTC