On Wed, 2 Mar 2011, Andrej Vodopivec wrote:
> This is more likely a regression in maxima. I get very different
> results with maxima 5.20 (ccl 1.4) and 5.23 (ccl 1.7-svn) on the same
> computer.
I can confirm that.
maxima 5.20.1 on ECL 10.2.1: flat
maxima 5.23.2 on ECL 11.1.1: increasing
The behaviour of integrate(log(x^2+y^2),x,0,1) did change between these
two maxima versions (5.20 did not need assume(y>1) ).