To clarify my point, a user would use animate(duration) OR forceLayout(), but not both. In either case, you end up rendering the positions of the layers once. The default is animate(0), which animates in a finally block. forceLayout() cancels the pending animation and lays out immediately. Its useful if you need to know the final position of elements synchronously, but in general it isn't necessary.
Thanks, John LaBanca [email protected] On Fri, Feb 4, 2011 at 10:23 AM, John LaBanca <[email protected]> wrote: > I was using it in a test case. It isn't required or recommended for the > widget in general. > > Thanks, > John LaBanca > [email protected] > > > On Fri, Feb 4, 2011 at 10:18 AM, Ray Ryan <[email protected]> wrote: > >> Re: forcing layout, could that turn into a source of slowness in apps, >> where we force recalculation that turns out to be redundant? >> On Feb 4, 2011 6:41 AM, <[email protected]> wrote: >> > I updated DeckLayoutPanel to implement AcceptsOneWidget, and I moved the >> > animationDuration down from TabLayoutPanel into DeckLayoutPanel so that >> > setWidget() will automatically use the default animation time. >> > >> > I went with AcceptsOneWidget instead of HasOneWidget because its >> > compatible with ActivityManager, and because I didn't want to add >> > DeckLayoutPanel.getWidget() (getVisibleWidget is more clear). >> > >> > Also, setWidget(null) clears the current widget. That seems more >> > natural than doing a no-op. >> > >> > >> > http://gwt-code-reviews.appspot.com/1340803/diff/1/12 >> > File user/src/com/google/gwt/user/client/ui/DeckLayoutPanel.java >> > (right): >> > >> > http://gwt-code-reviews.appspot.com/1340803/diff/1/12#newcode28 >> > user/src/com/google/gwt/user/client/ui/DeckLayoutPanel.java:28: * {@link >> > com.google.gwt.user.client.ui.TabPanel}. >> > On 2011/02/02 18:41:20, sbrubaker wrote: >> >> Do you mean TabPanel or TabLayoutPanel? >> > >> > Done. >> > >> > http://gwt-code-reviews.appspot.com/1340803/diff/1/21 >> > File user/test/com/google/gwt/user/client/ui/DeckLayoutPanelTest.java >> > (right): >> > >> > http://gwt-code-reviews.appspot.com/1340803/diff/1/21#newcode32 >> > user/test/com/google/gwt/user/client/ui/DeckLayoutPanelTest.java:32: // >> > Show widget at index 1, make sure it becomes visible. >> > On 2011/02/02 18:41:20, sbrubaker wrote: >> >> You may want to note that you're testing both forms of showWidget >> > (here and >> >> below). >> > >> > Done. >> > >> > http://gwt-code-reviews.appspot.com/1340803/diff/1/22 >> > File user/test/com/google/gwt/user/client/ui/TabLayoutPanelTest.java >> > (right): >> > >> > http://gwt-code-reviews.appspot.com/1340803/diff/1/22#newcode269 >> > user/test/com/google/gwt/user/client/ui/TabLayoutPanelTest.java:269: >> > assertEquals("inserted text", >> > Rietveld has weird highlighting. There aren't actually any spaces after >> > the comma. >> > >> > http://gwt-code-reviews.appspot.com/1340803/diff/1/22#newcode271 >> > user/test/com/google/gwt/user/client/ui/TabLayoutPanelTest.java:271: >> > assertEquals("added text", >> > same >> > >> > http://gwt-code-reviews.appspot.com/1340803/diff/1/22#newcode416 >> > user/test/com/google/gwt/user/client/ui/TabLayoutPanelTest.java:416: >> > p.forceLayout(); >> > If we don't force layout (synchronously), then it happens in a finally >> > command (asynchronously) after the current event loop, which would >> > require me to turn this into an asynchronous test with a bunch of nested >> > DeferredCommands. >> > >> > http://gwt-code-reviews.appspot.com/1340803/show >> > > -- http://groups.google.com/group/Google-Web-Toolkit-Contributors
