On Tue, 2006-02-14 at 17:49 -0500, Lillian Angel wrote: > On Tue, 2006-02-14 at 14:41 -0500, Lillian Angel wrote: > > When the location of a Dialog is changed before it is visible, it does > > not take any effect. The Dialog would appear at some arbitrary location. > > This is now fixed. > > > > 2006-02-14 Lillian Angel <[EMAIL PROTECTED]> > > > > * gnu/java/awt/peer/gtk/GtkDialogPeer.java > > (setVisible): New method to override super. Need to set the > > native bounds of the component, so it appears at the > > correct location. > > > > This patch caused some problems, because if the bounds is set before we > make the Dialog visible, it could be resized to 0x0. We just want to > make sure that the location is correct for all Windows. I made this > patch, but I would like approval before I commit it.
Can you run the gnu.testlet.java.awt tests from Mauve as well as the Window/Frame/Dialog related vte tests? I think we found that we needed to set the width and height, as well as the location (i.e. we needed to call setBounds) before showing the window. Tom > > Thanks > Lillian > > 2006-02-14 Lillian Angel <[EMAIL PROTECTED]> > > * gnu/java/awt/peer/gtk/GtkDialogPeer.java > (setVisible): Removed method. > * gnu/java/awt/peer/gtk/GtkWindowPeer.java > (setLocation): New method. > (setLocationUnlocked): New method. > (show): Changed to use setLocation instead of setBounds. > * java/awt/Component.java > (show): Should call peer.show(), not peer.setVisible(), so the > location of the component is correctly set. > * native/jni/gtk-peer/gnu_java_awt_peer_gtk_GtkWindowPeer.c > (Java_gnu_java_awt_peer_gtk_GtkWindowPeer_nativeSetLocation): > New function. > (Java_gnu_java_awt_peer_gtk_GtkWindowPeer_nativeSet > LocationUnlocked): New function. > * include/gnu_java_awt_peer_gtk_GtkWindowPeer.h: > Added declarations for Java_gnu_java_awt_peer_gtk_ > GtkWindowPeer_nativeSetLocation and > Java_gnu_java_awt_peer_gtk_GtkWindowPeer > _nativeSetLocationUnlocked. >
