Jan Vroonhof writes:
Justin Vallon <vallon(a)mindspring.com> writes:
> On Sun, 18 Jul 1999, Kyle Jones wrote:
> Though barely X literately, I think I have seen this too. I would suppose
> that the "other workspaces" frames are being unmapped when the WM
switches
> workspaces. XEmacs then incorrectly believes that there would be no way
> to get back to the other frames, though the "switch to the other
> workspace" button would re-map the frames.
Yes, i would consider this a window manager bug. It should use a state
that maps more naturally on the Workspace concept. FVWM uses
Invisible+Mapped which is OK (In fact XEmacs is optimized by that
case). Telling the hidden frames they are iconified would also be
fine.
Cool. The next thing to figure out now is why Steve can duplicate the
problem even when allow-deletion-of-last-visible-frame is t. There
must be some code path where this variable is ignored.