Re: New undo element (fun . args)

[Richard Stallman]

    It turned out to be quite tricky to get the usage right as an apply
    function need to install a proper _redo_ entry which may not really do
    anything but re-install the original _undo_ entry etc.

Ifr the extensible undo entry does some other nontrivial job,
surely its redo entry should redo that nontrivial job, right?
So how can it be that the redo entry only reinstalls the undo entry?