From: SourceForge.net <no...@so...> - 2007-01-23 19:45:29
|
Bugs item #1642851, was opened at 2007-01-23 13:43 Message generated for change (Comment added) made by willisbl You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=104933&aid=1642851&group_id=4933 Please note that this message will contain a full copy of the comment thread, including the initial issue submission, for this request, not just the latest update. Category: Lisp Core Group: Includes proposed fix Status: Open Resolution: None Priority: 5 Private: No Submitted By: Nobody/Anonymous (nobody) Assigned to: Nobody/Anonymous (nobody) Summary: notequal isn't negation of equal Initial Comment: The user documentation for 'notequal' says that notequal(a,b) is the negation of equal(a,b). It's not: (%i1) is(equal(x,%i)); (%o1) false (%i2) is(notequal(x,%i)); A putative fix is (defun mnqp (x y) (let ((b (meqp x y))) (cond ((eq b '$unknown) b) ((or (eq b t) (eq b nil)) (not b)) (t `(($notequal) ,x ,y))))) With the exception of test in rtest_taylor 117, this change is OK with the testsuite taylor(atan(1/sqrt(1-b))/sqrt(1-b),b,1,2) With the proposed mneq code, test 117 asks for the sign of b. If you answer pos (or neg), you get a sign called on %i error. And it seems that the asksign keeps some data about b (it shouldn't). With the old mnqp code, test 117 fails with the sign called on %i error, but it doesn't go thru asksign. Advice? ---------------------------------------------------------------------- >Comment By: Barton Willis (willisbl) Date: 2007-01-23 13:45 Message: Logged In: YES user_id=895922 Originator: NO I forgot to log in... ---------------------------------------------------------------------- You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=104933&aid=1642851&group_id=4933 |