More work on TMS and TIMESIN



Am Dienstag, den 16.06.2009, 23:25 +0200 schrieb Dieter Kaiser:

I have added further code to PLUSIN to do simplifications like the
following:

(%i12) 1/sqrt(2)+1/sqrt(2);
(%o12) sqrt(2)

Therefore we get for the integral of the bug report ID 1927178
"integrate(sin(t),t,%pi/4,3*%pi/4)" the expected result:

(%i17) integrate(sin(t),t,%pi/4,3*%pi/4);
(%o17) sqrt(2)

This works for the following too:

(%i13) 3/sqrt(2)+1/sqrt(2);
(%o13) 2^(3/2)

But not for this:

(%i14) 2/sqrt(2)+1/sqrt(2);
(%o14) sqrt(2)+1/sqrt(2)

Because we simplify 2/sqrt(2) immediately, PLUSIN does no longer add the
two terms. We have to factor the result to get the old result:

(%i15) factor(%);
(%o15) 3/sqrt(2)

I have run the testsuite and the share_testsuite. The testsuite gives 11
results which have changed. In the share_testsuite I have counted 7
changes. 5 changes are due to the testfile rtest_to_poly_solver.mac. I
will have a look at these tests to see what has changed.

Dieter Kaiser