Re: Typesetting bug

*To*: mathgroup at smc.vnet.net*Subject*: [mg96477] Re: Typesetting bug*From*: dh <dh at metrohm.com>*Date*: Sat, 14 Feb 2009 03:11:05 -0500 (EST)*References*: <gn3blp$prk$1@smc.vnet.net>

Hi Istvan, I can not reproduce your result. I guess it depends on the version. I have version7 / Windows and there it is o.k. regards Daniel replicatorzed at gmail.com wrote: > Dear Group, > > I don't know whether this "feature" is widely known or not, but it > surely gave me an hour of scratching my head, buried deep inside a > calculation: > > In[7]:= n = .5; > x = 1; > > ax = 1; > {(ax^n), x^n, ax^n - x^n} > > Out[10]= {1, 1, 0} > > In[11]:= max = 1; > {(max^n), x^n, max^n-x^n} > > Out[12]= {1, 1, -1} > > > This inconsistency happens only if exponents are typed in as real > superscripts with Ctrl+^. A FullForm reveals an underlying Times in > the last case, which ultimately is responsible for the wrong answer. > Furthermore, max^n does not work as intended if typed and evaluated in > one call, as follows: > max <Ctrl+^> n <enter> x <Shift+Enter> > because it is interpreted as a product of max^n and x. > > Istvan Zachar > >