There is an old, entrenched mathematical tradition of assigning different meanings, often entirely unrelated ones, to the minuscule and the majuscule of some SINGLE alphabetic character. Some familiar examples are
o t and T, h and H, f and F, o π and Π, σ and Σ, δ and Δ. For obvious reason---xy is ordinarily taken to be the arithmetic product or the logical conjunction of x and y---multicharacter symbols are less used. There are, however, some few of them that are used heavily. Some examples are o the circular functions sin, cos, tan, and sec, o the hyperbolic functions sinh, cosh, tanh and sech (rare), o the bounding functions max, min, sup, inf, o absolute value, abs, o signum, with signum(x | x < 0) = -1, signum(x | x=0) = 0, signum(x | x>0) = +1, for all real x. I have used the word 'entrenched' advisedly. These symbols are, for example, employed in the Roman alphabet in published Russian or Ukrainian mathematical papers that are otherwise written in (very slightly different versions of) the Cyrillic alphabet. Traditionally, they are all written in minuscules; and it is true that SIN x or even Sin x instead of sin x looks odd, even very odd; but no shudder went through the skies when its was written as SIN(X) in FORTRAN for many years. My personal views and practice are in complete agreement with those of Tony Harminc. All-majuscule source programs reflect obsolescent circa-1956 hardware technology, and they are hard to read. I therefore write my own source programs in mixed case comprised chiefly of minuscules. Those who wish to write them only in majuscules, in conformity with a tradition that they judge holy, are and should be free to do so; but this is not a practice that I recommend to students. The separable issue of case sensitivity in responses, even in macro-instruction keyword parameters, is one I feel much more strongly about. If such a keyword, call it &keyword, may be responded to with any of yes|no|maybe, then any case or mixture of cases should be allowable and treated as equivalent; indeed any CIDT of one of them should be supported and usable too. [Check boxes or pulldown menus may of course be used instead, but there is no place for case sensitivity in data entry, with the single but important exception of what is found within quoted strings, which should be preserved. Others may have very different views; but theirs are, of course, wrongheaded. John Gilmore, Ashland, MA 01721 - USA
