Re: Remarks on SWI 7

On Thu, Nov 28, 2013 at 8:38 AM, Richard A. O'Keefe <[email protected]>wrote:

> the empty list (ALONE amongst the vast range of atoms used to
> terminate recursive structures)
>

I foresee a future in which young SWI-Prolog programmers use [] everywhere
as *the terminator*, or *the empty thing*, or *not a value*. Like,

    bin_tree(a, bin_tree(b, [], []), bin_tree(c, [], [])).

or maybe:

    % Add all numbers, ignoring missing values
    sum([], 0).
    sum([X|Xs], Sum) :-
        sum(Xs, Sum_rest),
        (   X == []
        ->  Sum = Sum_rest
        ;   Sum is Sum_rest + X
    ).

People are shaking their heads, "O brave new world, that has such code in
it."

I find it difficult to explain why this is not a valid approach.
-------------- next part --------------
HTML attachment scrubbed and removed