A count of how many groups a waiting list must be split into says nothing, by itself, about when the last group is done. This paper asks what one set of decision records takes an executed calculation to have settled when it turns a number of people waiting for retraining into training cohorts, and what the records leave to time and to the order in which people go. It reads eleven authored cases from AEGIS, a studio of the RISCON family: every case in AEGIS's calculation programme whose run divides the people due for retraining by the seats of one cohort. All eleven are synthetic settings written in Korean, not real events; every post, figure and schedule in them is the author's assumption, and the English quoted is this study's own translation.
Four findings follow. The run settles one thing, the number of cohorts, and every recommendation states that number. The run's only clock stops at the end of the first cohort, so that in nine records nothing gives the time at which the last cohort ends, and only the two whose figures fit a deadline speak of the whole schedule. Who goes first is named as unknown in seven records and is never recommended. And in a few places the record's own words part from its own figures. Read beside the debates on carrying out policy from the top and from the bottom, on rationing at the front line, on queues and on manpower planning, the set counts its seats exactly and leaves its clock and its queue open.