On 10/29/07, Fabrizio Caruso <caruso at dm.unipi.it> wrote:
> It seems that the behaviour of batch has changed.
>
> My test do not work because things that are
> equal are considered different if they are not identical
Yes, batch/test used to compare the actual and expected output by
resimplifying using ratsimp. But that didn't detect many kinds of
bugs. For example, factor(x^2-1) would accept x^2-1 as a
factorization! Other bugs which were not detected where when Maxima
didn't perform a simplification it was supposed to. For example, if
in some context sin(-x) didn't simplify to -sin(x), test would still
think the test succeeded.
> and I haven't found a way to force them to be identical.
> Example:
> computed result is:
> ... - (2 a x^2 + 3 b x + 4 x)/4 ...
>
> I write the identical thing as expected output
> but maxima reorders the addends and shows during the test:
> ... -(4 c + 3 b x + 2 a x^2)/4 ...
You should use precisely the output of the computed result as shown by
"string". If this doesn't work, something is wrong, and you should
file a detailed bug report showing how to reproduce it.
-s