AnyEvent::Handle->starttls

[Robert Stone]

On Sat, May 25, 2013 at 12:30:19PM +0200, Marc Lehmann wrote:
> On Fri, May 10, 2013 at 09:34:28AM -0700, Robert Stone <talby at trap.mtview.ca.us> wrote:
> >         Should I be surprised by this?
> 
> No, TLS is very cpu-intensive, and there is nothing that AnyEvent can do
> about it.

        I've since discovered that in my test case, the first port was
negotiating an AES256-SHA cipher, the second was negotiating a
DHE-RSA-AES256-SHA cipher.  By setting the cipher_list option, I can get
the test to make a more fair comparison, and they benchmark similarly.

                                                -Robert