From cbmvax!uunet!CUVMA.BITNET!LOJBAN Fri Jan 17 17:11:58 1992 Return-Path: Message-Id: <9201171904.AA20365@cunixf.cc.columbia.edu> Date: Fri, 17 Jan 1992 14:01:33 -0000 Reply-To: David Cortesi Sender: Lojban list Comments: Warning -- original Sender: tag was cortesi@CRICKHOLLOW.INFORMIX.COM From: David Cortesi Subject: [Major: The dreaded word "only"] X-To: lojban mailing list To: John Cowan , Eric Raymond , Eric Tiedemann Status: RO [Shoulson...] > "lat." is not a valid cmene for the same reason as above [Major...] > "T" is the name of the truth constant in most implementations of lisp. > A discussion of a lisp program might well want to refer to "the thing > named 'T'". Wouldn't that be: la ty? (or la ga'e ty?) As in xu la fy cu fancu cupra la ga'e ty (yes or no: the f function-produces the T) Discovery: the above tanru is the only way I can find to refer to "the value returned by function f()." Certainly for programmers, >the< key element of a function is what it returns, its value. I think that's true for purer-math people also; one normally speaks of "f(x)" meaning, "the value that results from applying f() to x". Unfortunately "fancu" lacks a place for "value" entirely! While "cupra" lacks a place for "from input," leaving no way that I can find to say, "the value of f() applied to x." This is definitely something to consider in those last-minute place-structure revisions... cupra pra produce x1 produces x2 by process x3 fancu function x1 is a function of variable(s) x2 (plural/set) over domain x3 defined by expression x4 Following would not be appropriate IMO, it's the wrong kind of "return." xruti xut xru return x1 returns x2 to x3 from x4; x1 moves x2 back to x3 from x4 Following is the wrong kind of "value": vamji vam va'i value x1 is a value of x2 to x3 for use x4 Following would require an abstractor for "execution of function f" as its x2, which would be tedious. jalge jag ja'e result x1 is a result of action/event/state x2 Dave cortesi@informix.com