Revision: 2341 http://vexi.svn.sourceforge.net/vexi/?rev=2341&view=rev Author: clrg Date: 2007-09-27 08:25:20 -0700 (Thu, 27 Sep 2007)
Log Message: ----------- Add a comment explaining just why renderAll is necessary Modified Paths: -------------- trunk/core/org.vexi.core/src/org/vexi/plat/Platform.java Modified: trunk/core/org.vexi.core/src/org/vexi/plat/Platform.java =================================================================== --- trunk/core/org.vexi.core/src/org/vexi/plat/Platform.java 2007-09-27 15:22:38 UTC (rev 2340) +++ trunk/core/org.vexi.core/src/org/vexi/plat/Platform.java 2007-09-27 15:25:20 UTC (rev 2341) @@ -305,6 +305,13 @@ private static volatile boolean rendering = false; private static volatile boolean again = false; + /** + * This indirectly invokes reflow, which always preempts any render() + * call. Without it, any boxes [in a surface tree] that are marked + * for reflow but not yet dirtied [because any visual adjustments are + * dependent on the mechanisms of reflow] will not get updated until + * another event causes the box tree surface to queue render() again. + */ public void renderAll() { if (rendering) { again = true; return; } synchronized(Scheduler.class) { This was sent by the SourceForge.net collaborative development platform, the world's largest Open Source development site. ------------------------------------------------------------------------- This SF.net email is sponsored by: Microsoft Defy all challenges. Microsoft(R) Visual Studio 2005. http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/ _______________________________________________ Vexi-svn mailing list Vexi-svn@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/vexi-svn