Ken, did you change anything in cells as to how circular references
are dealt with?
Out of a sudden (???) test-gtk won't run, but complains about tons of
circular references. Pretty much every fm-other and fm^ produce a
circular reference, when they query the siblings by asking the (kids
(fm-parent self)), while, of course, self first needs to be created
for the kids slot to be accessible.
It seems like this always "just worked" against all odds, and now the
hounds come after me. First I took it as a sign to get started
converting (value (fm-other :something)) into (widget-value
:something), thus replacing tree search by hashtable lookup.
However, I came across this in buttons.lisp:
:new-tail (c? (and (upper self box)
(not (eql (first (kids (fm-parent self))) self))
'-from-widget))
i.e. this rule queries kids of parent of self -- which is as circular
as it can get.
Now why did this work in the past, and why is it broken now?
I don't understand ...
As always, help is greatly appreciated.
Peter
_______________________________________________
cells-gtk-devel site list
cells-gtk-devel@common-lisp.net
http://common-lisp.net/mailman/listinfo/cells-gtk-devel
