InvokeLater() - will just once be enough?

Question

I use the recommended code to start my interactive program, which uses Swing. :

    public static void main(String[] args) {
        javax.swing.SwingUtilities.invokeLater(
        new Runnable() {
          public void run() {
            createAndShowGUI();
          } } ) };

It creates a JFrame, call it "Foo", and ends when the user closes that window (by using the OS' Close- Window X icon or another OS way to close the app)

I want to display another window, "Bar", get user input, close that window with my own Swing calls, and then display the "Foo" window.

I could reuse the same JFrame for Foo and Bar, and just clear it out in-between. But I prefer to make each of them more independent in their design. So, an ActionListener in Foo would need to close Foo's frame, and call the code that displays Bar.

Do I need to use InvokeLater() to call the code that displays Bar? To get more design independence, should my main() be starting and synchronizing two threads?

score 2 · Answer 1 · answered Aug 16 '13 at 12:36

2

User triggered action listener code is executed in the EDT, so you do not need to wrap it again. That said, it may well be that using CardLayout would be more appropriate than multiple frames.

answered Aug 16 '13 at 12:36

kiheru

6,588
25
31

TYVM. In what way is CardLayout better? (i have never used it or read about it.) – ElHombre55 Aug 16 '13 at 13:06
in general users don't want multiple windows, you use just one frame – nachokk Aug 16 '13 at 13:09
@ElHombre55 It provides an easy way to switch window (or any other component) contents. Essentially a cleaner way to reuse the same frame. The layout of the contents can still be designed independently. CardLayout just gives a nice way to select the displayed component tree. – kiheru Aug 16 '13 at 13:11
@nachokk agreed. My UI is that only the menu displays; interactions; only the subject displays, in the same place; interactions. – ElHombre55 Aug 16 '13 at 20:22

score 2 · Answer 2 · edited May 23 '17 at 11:52

I could reuse the same JFrame for Foo and Bar, and just clear it out in-between. But I prefer to make each of them more independent in their design. So, an ActionListener in Foo would need to close Foo's frame, and call the code that displays Bar.

Do I need to use InvokeLater() to call the code that displays Bar? To get more design independence, I am unclear on the operation of the EDT.

you would use pack() and setVisible(true) wrapped into invokeLater in all cases for
1. for new Top-Level Container - alyways
2. for container created but never dispayed - always
3. for container once visible, then hidden and again visible on the screen - always
4. doesn't matter if is invoked from Swing Listener (by default on EDT) or not
to my point 3rd. to avoiding any unwanted Graphics lacks in the case that you reused Top-Level Container (is specifics, very short moment, but could be visible, but not, never annoying)
- old value is visible, then immediatelly refreshed to current
- old JComponent is visible, then immediatelly refreshed with current JComponents
- relayout/ pack() , the same issue as a.m.
invokeLater to delay (in most casses with success) this event to the end of EDT
see my view translated to the code demonstration about

InvokeLater() - will just once be enough?

2 Answers2