Creating copies of pycurl handles for threaded use

Question

I\'m trying to reduce the amount of time that requests I make using pycurl library uses. I\'ve noticed that reusing handles instead of creating them during every connection resu