rjnoe@ihlts.UUCP (Roger Noe @ N41:48:31, W88:07:13) (02/01/84)
>I think someone should point out that the traditional syntax of Roman >numerals prohibits strings of symbols of ascending order whose length >exceeds two, for the rather obvious reason that subtraction is not >associative. > Jim Crandell, C. S. Dept., The University of Texas at Austin Someone DID point that out. I did in the very article which was cited in Jim Crandall's article (which is cited above). But this is just tradition; there is no definitive syntax of Roman numerals. All we have to go by are examples actually found in authentic Roman works. Who says a tool cannot or should not do something untraditional as long as it is advertised as doing so? All are at liberty to modify my arabic.c to recognize only Roman numerals of one type or another. Roger Noe ihnp4!ihlts!rjnoe