Bug in 5.1 FE?

*To*: mathgroup at smc.vnet.net*Subject*: [mg52946] Bug in 5.1 FE?*From*: "Steven T. Hatton" <hattons at globalsymmetry.com>*Date*: Fri, 17 Dec 2004 05:19:44 -0500 (EST)*Sender*: owner-wri-mathgroup at wolfram.com

Could someone please verify this problem I'm experiencing on a SuSE 9.2 Intel system with Mathematica 5.1? To reproduce the problem do the following: Create a package as follows: BeginPackage["NotBroken`", "Graphics`Arrow`"] Begin["`Private`"] delta = Power[a, Rational[-1, 2]]; placeArrow[point_, d_] := Module[{}, Arrow[point, point + d, HeadScaling -> Relative] ]; End[] EndPackage[] Make the cell containing the above code an initialization cell, execute the code, and save it in package format as NotBroken.nb. In another file place: Needs["NotBroken`"] and evaluate that cell after killing the kernel. There should be no output. Now save the file as Broken.nb replace the line: delta = Power[a, Rational[-1, 2]]; with: \!\(\(delta = 1\/\@a;\)\) and change the package name to Broken. After killing the kernel try to evaluate a Needs["Broken`"]. I get This error message: Arrow::shdw : Symbol Arrow appears in multiple contexts {Graphics`Arrow`,Global`}; definitions in context Graphics`Arrow` may shadow or be shadowed by other definitions. HeadScaling::shdw: Symbol HeadScaling appears in multiple contexts {Graphics`Arrow`, Global`}; definitions in context Graphics`Arrow` may shadow or be shadowed by other definitions. Relative::shdw: Symbol Relative appears in multiple contexts {Graphics`Arrow`, Global`}; definitions in context Graphics`Arrow` may shadow or be shadowed by other definitions. Strangely, removing the HeadScaling -> Relative fromm the Arrow[] command makes the error message go away. -- "Philosophy is written in this grand book, The Universe. ... But the book cannot be understood unless one first learns to comprehend the language... in which it is written. It is written in the language of mathematics, ...; without which wanders about in a dark labyrinth." The Lion of Gaul