Dear Vladimir, Thanks for spotting this bug and for fixing it! I'm applying your patch. Also applying the second one about relational::canonical() function. (Sorry for the delay.) -richy. -- Richard B. Kreckel <https://in.terlu.de/~kreckel/> On 02.08.21 09:47, Vladimir V. Kisil wrote:
Dear All,
Currently the following code in GiNaC
realsymbol x("x"); cout << (cos(x)/exp(x)).normal() << endl;
prints value 1 with probability 50% (and the unmodified expression otherwise). I am attaching a one-line patch to fix this silly bug.
Also I am attaching a previous patch which was not commented so far. The new method may be helpful from time to time but its equivalent in a user code would be much more involved due to the private nature of the relation class members.
Best wishes, Vladimir
_______________________________________________ GiNaC-devel mailing list GiNaC-devel@ginac.de https://www.ginac.de/mailman/listinfo/ginac-devel