I will let an IBMer who can do it authoritatively straighten out the history.
I do want to note that I am not plagued by this behavior. I have known about it for 47 years; it no longer surprises me; and I long ago worked out ways to obtain signed values where I need them. I am more querulous about programmer sloth than about the assembler's notional deficiencies. The users of the HLASM are, presumably, programmers; and if they are competent ones these 'anomalies' are 1) sometimes convenient and 2) sometimes pose small, easily circumvented problems. The notion that there is an ideal, trouble-free Platonic assembler laid up in in heaven somewhere, and that departures from its [conjectural] design are errors or, worse, sins is indeed drôle. To borrow from Clint Eastwood, it has made my day. Let me end it by recalling what Sir Thomas Browne said at the beginning of Pseudodoxia epidemica: Would truth dispense, we could be content, with Plato, that knowledge were but remembrance, that intellectual acquisition were but reminiscentiall evocation, and new impressions but the colourishing of old stamps that stood pale in the soul before. For, what is worse, knowledge is made by oblivion; and to purchase a clear and warrantable body of Truth, we must forget and part with much that we know. John Gilmore, Ashland, MA 01721 - USA
