Dear Feng,
I tested you code with old FeynCalc and can confirm your observations:
FeynCalc 4.2.0 pre1, Mathematica 5: 4.3 seconds
FeynCalc 9.0, Mathematica 9: 13.5 seconds
It is roughly a factor of 3 slowdown, unfortunately.
I will try to investigate this. The number of changes in the relevant
functions between Release-4_2_0_0 and Release-5_0_0_0-beta3 is not that
big so we should be able to trace the problem in a resonable amount of
tume. Git bissect is a particularly nice tool for doing this kind of
investigations.
Cheers,
Vladyslav
This archive was generated by hypermail 2b29 : 09/04/20-12:55:05 AM Z CEST