[Twisted-Python] Twisted 2.0 prerelease (close!!)

Andrea Arcangeli andrea at cpushare.com
Fri Mar 11 01:59:53 MST 2005


On Thu, Mar 10, 2005 at 04:14:52PM -0500, James Y Knight wrote:
> No, you've got the situation backwards. The situation doesn't involve 
> any hanging, it involves possible data loss of the last packet. We will 
> make a single attempt to do the final write (via calling SSL_shutdown), 
> and if it fails to write because of some transient error, we have no 
> way to know, so we close the connection anyways, even though we 
> should've retried.

Thanks for the explanation. I've also got those messages, could you post
the patch so I can apply it to pyOpenSSL until it's merged? Thanks.




More information about the Twisted-Python mailing list