murdering tkinter programs

[Eric Brunel]

jeff wrote:
> At the start of a Tkinter program, you create a root window and then start
> the mainloop().
> 
> So why isn't it standard practice to do the reverse when shutting down the
> program, via an exit command or quit button ?
> - that is, call .quit() first, then .destroy()
> 
> All the tutorials and introductions only seem to call destroy() in their
> example programs.

Well, this is actually using a side-effect of the destroy method on the main 
application window, which actually quits the main loop. But you are perfectly 
right; one should do:

from Tkinter import *
root = Tk()
Button(root, text='Quit', command=root.quit).pack()
root.mainloop()
root.destroy()

So clicking on the button actually quits the main loop, but does not actually 
destroy anything; you can see that by adding a time.sleep(2) after the 
root.mainloop(). This is usually not considered as a problem, because Tkinter 
programs rarely have anything left to do after the main loop is over. So the 
next thing that happens is that the Python interpreter actually exits, and the 
window disappear.

Note also that the quit method is available on all Tkinter widgets, and always 
quits the main loop:

from Tkinter import *
root = Tk()
b = Button(root, text='Quit')
b['command'] = b.quit
b.pack()
root.mainloop()
root.destroy()

This can be quite handy if you're in a module that doesn't know anything about 
the main window, but that can make the program quit anyway.
-- 
- Eric Brunel <eric (underscore) brunel (at) despammed (dot) com> -
PragmaDev : Real Time Software Development Tools - http://www.pragmadev.com