#3546 Incorrect Laplace transform cos(sin(t))

[John Scott]

Hello.

On Maxima 5.42.1 and 5.41.0, it incorrectly gives the Laplace transform of cos(sin(t^p)) for seemingly any p > 0.

(%i2) laplace(cos(sin(t^p)),t,s);
Is p positive, negative or zero?

positive;
Is p an integer?

no;
Is 2 p - 1 positive, negative or zero?

positive;
       1
Is p - - positive, negative or zero?
       2

positive;
                                       1
(%o2)                                  -
                                       s

In fact, it appears that the questions can be answered in any way at all and it will still give 1/s, which however is the leading term in the Laplace transform's asymptotic expansion.