W3C home > Mailing lists > Public > public-html-commits@w3.org > August 2008

html5/spec Overview.html,1.1211,1.1212

From: Ian Hickson via cvs-syncmail <cvsmail@w3.org>
Date: Wed, 06 Aug 2008 07:57:38 +0000
To: public-html-commits@w3.org
Message-Id: <E1KQdtq-0004Uq-Ue@lionel-hutz.w3.org>

Update of /sources/public/html5/spec
In directory hutz:/tmp/cvs-serv17257

Modified Files:
Log Message:
Simplify garbage collection for ports even further. Define dicarding of Document objects better for ports. Prevent inactive documents from receiving messages. (whatwg r2024)

Index: Overview.html
RCS file: /sources/public/html5/spec/Overview.html,v
retrieving revision 1.1211
retrieving revision 1.1212
diff -u -d -r1.1211 -r1.1212
--- Overview.html	6 Aug 2008 07:25:14 -0000	1.1211
+++ Overview.html	6 Aug 2008 07:57:36 -0000	1.1212
@@ -44024,6 +44024,17 @@
    href="#messageport0">MessagePort</a></code> object, and remove the event
    from the queue.
+  <p>If the <code><a href="#messageport0">MessagePort</a></code> is owned by
+   a <code><a href="#window">Window</a></code> object and the
+   <code>Document</code> that was the <a href="#active">active document</a>
+   in that <a href="#browsing1">browsing context</a> when the <code><a
+   href="#messageport0">MessagePort</a></code> was created is not <a
+   href="#fully">fully active</a>, then an opportunity doesn't exist &mdash;
+   events in such cases must only be dispatched once the
+   <code>Document</code> in question becomes <a href="#fully">fully
+   active</a>. If that doesn't happen before that <code>Document</code> is
+   discarded, then the events are lost.
   <p>The <dfn id=close3
@@ -44102,11 +44113,42 @@
   <h5 id=ports><span class=secno> </span>Ports and browsing contexts</h5>
-  <p>Ports are unentangled when the <code>Document</code> that was the <a
-   href="#active">active document</a> of the <a href="#browsing1">browsing
-   context</a> corresponding to the <code><a href="#window">Window</a></code>
-   object that owns them is <a href="#discard" title="discard a
-   document">discarded</a>.
+  <p>When a <code>Document</code> is <a href="#discard" title="discard a
+   document">discarded</a>, if there are any <code><a
+   href="#messageport0">MessagePort</a></code> objects that:
+  <ul class=brief>
+   <li>are entangled, and
+   <li>are owned by the <a href="#browsing1">browsing context</a> that
+    contained that <code>Document</code>, and
+   <li>were created while that <code>Document</code> was the <a
+    href="#active">active document</a> of that <a href="#browsing1">browsing
+    context</a>, and
+   <li>are entangled with a port that is either not owned by that <a
+    href="#browsing1">browsing context</a> or was not created while that
+    <code>Document</code> was the <a href="#active">active document</a> of
+    that <a href="#browsing1">browsing context</a>,
+  </ul>
+  <p>...then the user agent must run the following steps for each such port:
+  <ol>
+   <li>
+    <p>Let <var title="">surviving port</var> be the port with which the
+     <code><a href="#messageport0">MessagePort</a></code> object in question
+     is entangled.
+   <li>
+    <p>Unentangle the two ports.</p>
+   <li>At the next available opportunity, after any scripts have finished
+    executing<!-- XXX queue -->, <a href="#firing2">fire a simple event</a>
+    called <code title=event-unload>unload</code> at <var title="">surviving
+    port</var>.
+  </ol>
   <h5 id=ports0><span class=secno> </span>Ports and garbage
@@ -44126,12 +44168,6 @@
     live code, despite having a strong reference to each other.</p>
-  <p>When an entangled message port is about to be garbage collected, it must
-   be unentangled. Because of the aforementioned strong reference, this can
-   only happen if either both ports are about to be garbage collected as a
-   pair, or if the entire <a href="#script2">script execution context</a> of
-   the port is being discarded.
   <h2 id=syntax><span class=secno>8. </span>The HTML syntax</h2>
   <h3 id=writing0><span class=secno>8.1 </span>Writing HTML documents</h3>
Received on Wednesday, 6 August 2008 07:58:17 UTC

This archive was generated by hypermail 2.3.1 : Wednesday, 7 January 2015 15:09:18 UTC