nuitka is slower than standard python2.7

i tested the following demo code and found nuitka is slower than python2.7

import time

t = [ i for i in range(1,10000000)]
s = [ i for i in range(1,50000)]

t1 = time.time()

def test002():
    print(set(s).issubset(t))


def test001():
    sd = True
    for i in s:
        if i not in t:
            sd = False
    print(sd)

test001()
test002()

t2 = time.time() off = t2-t1 print(“used time = %f seconds.” % off)

result of python: build_python python hello.py True True used time = 12.913735 seconds.

result of nuitka: build_python ./hello.bin True True used time = 18.586609 seconds.

i saw the offical website said “nuitka is about 3 times than standard python,” but why is that testing result?

[kayhayen]

Unfortunately, for the benefits of this to kick in all cases, some stars will have to align for me to finish this, without hurting scalability.

[sunbit]

Hello

I think this is happening in one of our legacy code that we have in 2.7 right now. Unfortunately i cannot provide a detailed info about the code right now, what i can say is that is a Flask application running on uwsgi, compiled with 0.6.7, and our tests shows that the nuitka compiled version doubles the execution time on average compared with running Cpython.

Is there any progress on this?

Thank you for your work!