On Friday 29 September 2006 21:53, Kevin L wrote: > Package: cl-statistics > > This is strange: cl-statistics compiles fine in CMUCL and *appears* to > compile fine in SBCL, but SBCL still throws an error.
From the log, and from experience, I guess that the error is 'compile-file warned while'. This is a pretty irritating issue but the sbcl people claim it is a feature, not a bug :-(. This is because the warnings are of possibly broken code (they are not STYLE-WARNINGS): > ; compiling (DEFPACKAGE :STATISTICS ...); > ; caught WARNING: > ; STATISTICS also exports the following symbols: > ; (FISHER-Z-TRANSFORM NORMAL-MEAN-CI-ON-SEQUENCE > FALSE-DISCOVERY-CORRECTION) > ; See also: > ; The ANSI Standard, Macro DEFPACKAGE instead of : > > ; compiling (DECLAIM (OPTIMIZE # ...)) > ; compiling (IN-PACKAGE :STATISTICS) > ; compiling (DEFMACRO TEST-VARIABLES ...); in: LAMBDA NIL > ; (LET* ((STATISTICS::DUMMY (CDR #:WHOLE60))) > ; (BLOCK STATISTICS:TEST-VARIABLES)) > ; > ; caught STYLE-WARNING: > ; The variable DUMMY is defined but never used. > So actually there is a bug in cl-statistics :-S Groetjes, Peter -- signature -at- pvaneynd.mailworks.org http://www.livejournal.com/users/pvaneynd/ "God, root, what is difference?" Pitr | "God is more forgiving." Dave Aronson|