Re: [twsocket] TiggerSessionClosed...

2012-10-11 Thread zayin
Thanks for the reply Arno. >Otherwise make sure you close all clients before destroying the server component. I made sure that happens before posting the message and the clients are freed before the server. It is on shutdown so I am not worried. The issue is FastMM reports the leak each time I

Re: [twsocket] TiggerSessionClosed...

2012-10-11 Thread Angus Robertson - Magenta Systems Ltd
> It is on shutdown so I am not worried. The issue is FastMM reports > the leak each time I quit and that is a pain. It is probably a false report. I use the ICS web server which uses SocketServer, and there are no memory leaks after over 100,000 connections. Angus -- To unsubscribe or chang

Re: [twsocket] TiggerSessionClosed...

2012-10-11 Thread zayin
> It is probably a false report... MadExcept also reports it as a memory leak. -- To unsubscribe or change your settings for TWSocket mailing list please goto http://lists.elists.org/cgi-bin/mailman/listinfo/twsocket Visit our website at http://www.overbyte.be

Re: [twsocket] TiggerSessionClosed...

2012-10-11 Thread Angus Robertson - Magenta Systems Ltd
> > It is probably a false report... > MadExcept also reports it as a memory leak. So after say 10,000 connections, how much extra memory is your application using than when it started? The VCL often leaks small amounts of memory, but unless is cumulative it really does not matter. Use somethin

Re: [twsocket] TiggerSessionClosed...

2012-10-11 Thread zayin
> but unless is cumulative it really does not matter. I concur. It is just a pain. IMHO, there should be some step(s) to close the clients and close the server and not have a memory leak. -- To unsubscribe or change your settings for TWSocket mailing list please goto http://lists.elists.org/

Re: [twsocket] TiggerSessionClosed...

2012-10-11 Thread Arno Garrels
zayin wrote: >> but unless is cumulative it really does not matter. > > I concur. It is just a pain. > > IMHO, there should be some step(s) to close the clients and close the > server and not have a memory leak. I've never seen that leak, anyway Close server's listening socket in order to disabl

Re: [twsocket] TiggerSessionClosed...

2012-10-11 Thread zayin
> ..happens, for instance, _if the destination Window has been destroyed after the call to PostMessage Because the program is ending when the leak occurs that is exactly what I think is occurring. So, after I close all the clients, I call now server.ProcessMessages and that seems to handle the le

Re: [twsocket] TiggerSessionClosed...

2012-10-11 Thread Arno Garrels
zayin wrote: >> ..happens, for instance, _if the destination Window has been >> destroyed after the call to PostMessage > > Because the program is ending when the leak occurs that is exactly > what I think is occurring. > > So, after I close all the clients, I call now server.ProcessMessages > a

Re: [twsocket] TiggerSessionClosed...

2012-10-11 Thread zayin
> I try to avoid calling the message pump Ditto. > DisconnectAll That worked. So, now I close the server, call DisconnectAll and no memory leak. Thanks -- To unsubscribe or change your settings for TWSocket mailing list please goto http://lists.elists.org/cgi-bin/mailman/listinfo/twsocket V

[twsocket] Re TiggerSessionClosed...

2012-10-11 Thread zayin
> ..happens, for instance, _if the destination Window has been destroyed after the call to PostMessage Because the program is ending when the leak occurs that is exactly what I think is occurring. So, after I close all the clients, I call now server.ProcessMessages and that seems to handle the l