On Mar 4, 2008, at 5:31 AM, Johan Turkenburg wrote:
I would argue that consistency with other applications is less
important than the irritation of another confirm button. I'm surely
not the only one who grumbles 'yes of course I wanted to exit, why
else do you think I clicked the exit button'.... ;-)
First, the close button could map to a different call back than the
EXIT button. Intercepting and putting up a confirmation dialog is
trivial and amounts, inclusively, to (1) changing the already existing
binding
self.winfo_toplevel().protocol("WM_DELETE_WINDOW", self.wm_delete)
plus (2) three new lines of code:
def wm_delete(self):
if tkMessageBox.askyesno("Exit", "Exit ccp4i?"):
self.exit()
Second, I think the behavior most consistent with almost every
application created in the last 15 years is to ask to close if the
user has made unsaved changes. This could be as simple as adding a
callback to the input fields when they lose focus (which I assume are
already sub-classed any way).
Consider, for example, one's trying to import a CNS file, where one
spends 25 minutes getting the 6x,F10.5,6x,I10,etc. just right, and
after the 5th iteration one tries to close the task input window only
to close the main ccp4i window, which exits without so much as a core-
dump.
For my 4 function calculator, I don't want it to ask me if I want it
to close. For my word processor, I expect confirmation if I've made
changes. For a crystallography suite, I would err on the side of
caution.
James
--
James Stroud
UCLA-DOE Institute for Genomics and Proteomics
Box 951570
Los Angeles, CA 90095
http://www.jamesstroud.com
