Please look at the picture (0.87 stable):

I think there is something wrong here. The results (represented in red) should not be such great numbers. Due to large arguments, I think an error issued might be more appropriate.

Regards,
Radovan