Re: Possible to abort close stack?

emlowe@fsa.ca
Wed, 19 Jul 1995 09:50:49 -0700 (MDT)

>
> >
> > Hi, Is it possible to abort the close stack process? I want to be able
> > to allow users to close the window from the window manager, using
> > whatever method the window manager uses, but I also want to be able to
> > abort the close in certain situations.
> >
>
> You can't override the close message, since this message is only a
> notification because it isn't initiated directly by user action. You
> can remove the close box entirely by setting the "decorations"
> property, however. Other than that, all you can do is reopen the
> stack after it gets closed.
>

Well, maybe I can remove the decorations, but I can't change the
window's menu, and there are other ways to get at the menu than through
the decorations. For example, on my system, I have 4Dwm set up so that
if I click with the third mouse button on the title bar, I get the
menu, or if I Alt-click on the window, I get the menu. So the user can
always close the window via the window manager, and there isn't
anything I can do about it. This is a real big lose in my opinion,
especially for a commercial app. It's important to take all steps
possible to ensure that a user can't lose something that they have
worked on for hours, and this would seem to be a pretty big hole.

> > It's trivially easy in Tcl/Tk (nudge, nudge, Scott)
>
> Yeah, but that's because you get to (have to) deal with raw X events
> in Tk, even to the point of using the X definitions for the message
> names. This goes into the "more than enough rope" category IMHO.
> It's also going to look mighty strange to people who don't know the
> ICCCM documentation (or who are developing stacks for the Mac, OS/2 or
> Windows, nudge, nudge ;-)

And the Mac, OS/2, and Windows versions of MetaCard are available when? ;-)

I agree about the strange names, but this is important functionality
that should be available. If you don't need it, then don't worry about
it. If however, you need it, it should be there.

>
> I'll a message for this on the feature request list. It would work
> like the keyDown message: if you have a handler for it that doesn't
> pass it, the default action (inserting a character or closing the
> window) isn't taken.
>

I think that this is very important, can it be expediated?

-Earle

-- 
Earle Lowe
emlowe@fsa.ca
FSA Corporation
Performance and system software for UNIX networks.