Hi everybody, I wrote an external method that performs the authentication to an AFS server, calling a C subroutine that I wrapped with SWIG. The subroutine executes fine, and control returns to the python code of the ext method that called it , but when the etx method exits (immediately after exiting, I guess) I get an error with this stack trace: 2002-07-08T10:28:19 ERROR(200) ZServer uncaptured python exception, closing channel <ZServer.HTTPServer.zhttp_channel connected aa.bb.cc.dd:1660 at 0x8b2f12c channel#: 2 requests:> (exceptions.IOError:[Errno 107] Transport endpoint is not connected [.../carlo/zope/Zope-2.5.1-linux2-x86/ZServer/medusa/asyncore.py|poll|107] [.../carlo/zope/Zope-2.5.1-linux2-x86/ZServer/medusa/asyncore.py|handle_write_event|398] [.../carlo/zope/Zope-2.5.1-linux2-x86/ZServer/medusa/asynchat.py|handle_write|146] [.../carlo/zope/Zope-2.5.1-linux2-x86/ZServer/medusa/asynchat.py|initiate_send|208] [.../carlo/zope/Zope-2.5.1-linux2-x86/ZServer/medusa/asynchat.py|refill_buffer|195] [.../carlo/zope/Zope-2.5.1-linux2-x86/ZServer/Producers.py|more|36] [.../carlo/zope/Zope-2.5.1-linux2-x86/ZServer/medusa/http_server.py|log|289] [.../carlo/zope/Zope-2.5.1-linux2-x86/ZServer/medusa/logger.py|log|247] [.../carlo/zope/Zope-2.5.1-linux2-x86/ZServer/medusa/logger.py|log|77] [.../carlo/zope/Zope-2.5.1-linux2-x86/ZServer/medus2002-07-08T10:28:19 ERROR(200) ZServer uncaptured python exception, closing channel <ZServer.HTTPServer.Traceback (most recent call last): File ".../carlo/zope/Zope-2.5.1-linux2-x86/z2.py", line 707, in ? asyncore.loop() File ".../carlo/zope/Zope-2.5.1-linux2-x86/ZServer/medusa/asyncore.py", line 200, in loop poll_fun (timeout, map) File ".../carlo/zope/Zope-2.5.1-linux2-x86/ZServer/medusa/asyncore.py", line 111, in poll obj.handle_error() File ".../carlo/zope/Zope-2.5.1-linux2-x86/ZServer/medusa/http_server.py", line 437, in handle_error asynchat.async_chat.handle_error (self) File ".../carlo/zope/Zope-2.5.1-linux2-x86/ZServer/medusa/asyncore.py", line 421, in handle_error self.close() File ".../carlo/zope/Zope-2.5.1-linux2-x86/ZServer/HTTPServer.py", line 293, in close p.more() # free up resources held by producer File ".../carlo/zope/Zope-2.5.1-linux2-x86/ZServer/Producers.py", line 36, in more getattr(self.logger, self.method)(self.bytes) File ".../carlo/zope/Zope-2.5.1-linux2-x86/ZServer/medusa/http_server.py", line 289, in log ' - %s [%s] "%s" %d %d "%s" "%s"\n' % ( File ".../carlo/zope/Zope-2.5.1-linux2-x86/ZServer/medusa/logger.py", line 247, in log self.logger.log ('%s%s' % (ip, message)) File ".../carlo/zope/Zope-2.5.1-linux2-x86/ZServer/medusa/logger.py", line 77, in log self.write (message) File ".../carlo/zope/Zope-2.5.1-linux2-x86/ZServer/medusa/logger.py", line 53, in write self.maybe_flush() File ".../carlo/zope/Zope-2.5.1-linux2-x86/ZServer/medusa/logger.py", line 65, in maybe_flush self.file.flush() IOError: [Errno 107] Transport endpoint is not connected Not to mention, the ZServer process aborts execution. Does anybody know what kind of bug might cause such error and (possibly) how to fix it? Thanks in advance. Regards, Carlo.