Why does Mathematica 5.0 fail where Mathematica 4.1 works ?

*To*: mathgroup at smc.vnet.net*Subject*: [mg46996] Why does Mathematica 5.0 fail where Mathematica 4.1 works ?*From*: Oleksandr Pavlyk <pavlyk at phys.psu.edu>*Date*: Fri, 19 Mar 2004 01:35:49 -0500 (EST)*Organization*: Penn State University; Department of Physics*Reply-to*: pavlyk at phys.psu.edu*Sender*: owner-wri-mathgroup at wolfram.com

Hi, I was looking at excellent talk by Daniel Lichtblau "Tactics for solving equations in Mathematica" from http://library.wolfram.com/infocenter/Conferences/337/ In the "Tricky Solve problem" chapter we find the following k = E; Solve[ { y == x^y , y == k }, x ] So here are outputs in the freshly started notebooks in two versions of Mathematica Mathematica 4.1 on Solaris returns E^(1/E) and Mathematica 5.0 on Windows returns (1/y)^(-1/y) where y is an unknown variable. Both versions complain that inverse functions were used and hence some solution maybe lost. See output from 5.0 attached below. I would not be surprised of somebody tells me this bug has been fixed in 5.0.1. In fact this is what happened with all earlier bug reports posted on this mailing list. I must ask our sysadmin if our departmental subscription qualifies for a free upgrade to Mathematica 5.0.1. ;) I can't help but feel that Mathematica should provide bug (not limitation) fixing patches for Mathematica users free of charge. But that's me... Sasha In[1]:= $Version Out[1]= 5.0 for Microsoft Windows (June 11, 2003) In[2]:= sol = x /. First[Solve[{y == x^y, y == E}, x]] From In[2]:= \!\(\* RowBox[{\(Solve::"ifun"\), \(\(:\)\(\ \)\), "\<\"Inverse functions are being used by \\!\\( Solve\\), so some solutions may not be found; use Reduce for complete \ solution information. \\!\\(\\*ButtonBox[\\\"More?\\\", \ ButtonStyle->\\\"RefGuideLinkText\\\", ButtonFrame->None, \ ButtonData:>\\\"Solve::ifun\\\"]\\)\"\>"}]\) Out[2]= \!\(\((1\/y)\)\^\(\(-1\)/y\)\)