The existing algorithm is well-behaved, but extensions to it, from
which it can greatly benefit, are already foreseen. The problem is
not thoroughly understood and some of the directions it may take depend
on discussions and slow exploration of the effects that slight, or
at times rather radical, changes make. Section
was aimed to show some of the more erratic ideas which either bear
potential or possibly be disastrous. Nonetheless, these should motivate
discussions and incompletely depict possible ways of going ahead.