Theorem 2 as a word RAM program. Read the word, sort the jobs by start time, and fill the table of Section 3 over sets of thresholds written as base-$(n+1)$ numbers: the numbers are visited from the top, each is decided to be a code of a set or not by a recurrence, and the two sets $X₁$, $X₂$ of recursion (1) are found by one scan of its digits ($O(|X|)$, independent of $W$), after which its $W+1$ cells cost $O(1)$ each. The sets contribute $Σ(|X|+1) ≤ 2 (n+1)^m$ in all. The table is the monotone one ("weight at least $W'$"), so it has $W+1$ rows whatever the weights are, and the answer is one entry of the column of the first $m$ indices. The domain restricts to positive processing times, the paper's standing assumption for Lemma 1's recursion. The running time is $c (W+1) (n+1)^m + c · sortCost$: no work per machine beyond the $O(|X|)$ scan of a set.