----- Original Message ----- From: "Arno Garrels" <[EMAIL PROTECTED]> To: "ICS support mailing" <twsocket@elists.org> Sent: Wednesday, August 22, 2007 9:46 PM Subject: Re: [twsocket] ThreadDEtach with THttpConnection problem
> Fastream Technologies wrote: >> But it always stops at that dispatchmessage! Persistent and insistent. > > Now that I have some minutes, I checked the source code again, > however I cannot find any bug, may be sorting out things helps a bit? > > When a window is being destroyed all pending messages in thread's > message queue addressed to the destroyed window are being removed > as well. I just wrote a little test project that confirms the API doku. > So it seems impossible that GetMessage/PeekMessage will ever retrieve > messages to an already destroyed window. > > Since winsock does not try to send any further notification messages > to the destroyed window (as Francois already mentioned) this can be > sorted out as well. > > Remains the case when the window is not being destroyed, but registered > message numbers have been unregistered. Means there may be still messages > pending in thread's message queue addressed to an existing window, however > they won't be handled by the component but passed to DefWindowProc which > will throw them away. > > Do you register new custom messages, thru AllocateMsgHandlers? If so > are they unregistered properly by overriding FreeMsgHandlers as well? > > Also critical: When the component has no window handle and for some > reason property Handle is called an auto-attach is performed by the > component which _must_ be avoided. > > Remains the question whether important messages may be removed upon the > call to ThreadDetach()? I don't think so, given that it's called after > SessionClosed, though the message pump may be called from InternalClose, > however that should not hurt at all. As I wrote earlier, I call ThreadDetach in a private handler that is postthreadmessage'ed from SessionClosed, not directly there. Regards, SZ > > -- > Arno Garrels [TeamICS] > http://www.overbyte.be/eng/overbyte/teamics.html > >> >> On 8/22/07, Arno Garrels <[EMAIL PROTECTED]> wrote: >>> >>> Fastream Technologies wrote: >>>> Is TranslateMessage really needed in a socket worker thread? >>> >>> I guess it's not needed, however I use it in all of my pumps. >>> >>>> All I know is I see >>>> "EAccessViolation" at DispatchMessage() of worker thread. >>> >>> The debugger isn't very reliable in multi-threaded applications >>> it happens that it stops at one line whilst the error is at a >>> different location. >>> >>> -- >>> Arno Garrels [TeamICS] >>> http://www.overbyte.be/eng/overbyte/teamics.html >>> >>> -- >>> To unsubscribe or change your settings for TWSocket mailing list >>> please goto http://www.elists.org/mailman/listinfo/twsocket >>> Visit our website at http://www.overbyte.be >>> >> >> >> >> -- >> CSA, Fastream Technologies >> Software IQ: Innovation & Quality >> www.fastream.com | Email: [EMAIL PROTECTED] | Tel: +90-312-223-2830 >> Join IQWF Server Yahoo group at >> http://groups.yahoo.com/group/IQWFServer Join IQ Reverse Proxy Yahoo >> group at http://groups.yahoo.com/group/IQReverseProxy > -- > To unsubscribe or change your settings for TWSocket mailing list > please goto http://www.elists.org/mailman/listinfo/twsocket > Visit our website at http://www.overbyte.be -- To unsubscribe or change your settings for TWSocket mailing list please goto http://www.elists.org/mailman/listinfo/twsocket Visit our website at http://www.overbyte.be
