And, for completeness, I'll add that I'm using the Cisco CMETSP TAPI drivers. In case anyone was wondering why TAPI was using a socket to talk to a server.
On 4/19/07, Fargo Holiday <[EMAIL PROTECTED]> wrote: > > Ah, TAPI. Well, I'm glad you got it sorted out, and I'm always glad to > be of whatever assistance I can. > > Jim Dossey wrote: > > Thanks to Glenn and Fargo for their suggestions. In this particular > case, > > I'm not directly opening the socket to the server. I'm using the > Windows > > TAPI service. So I can't check for a socket error directly. But your > > suggestions gave me an idea, and I started looking for any error codes > from > > TAPI that might indicate a network error, and I found 2. So I'm using > them > > to reopen the connection after a failure. These error codes are > > undocumented, but they seem pretty consistent. > > > > On 4/18/07, Glenn L. Austin <[EMAIL PROTECTED]> wrote: > > > >> An even better solution, since it would work for cases of network > >> connection loss. > >> > >> Even better than check the delta time between the current time and > >> last time checked. > >> > >> On Apr 17, 2007, at 10:17 PM, Fargo Holiday wrote: > >> > >> > >>> Well, there may be some way to check the state with the Win32 API or > >>> WMI, but the timing would probably be tricky. Checking the time, as > >>> suggested by Glenn, might work, but strikes me as possibly unreliable, > >>> since hibernation can be triggered by a variable timer and user > >>> commands. Unless I totally missed his thought process there. > >>> > >>> Why not just provide an automated re-connect for the app? If the > >>> socket > >>> connection fails have it attempt to re-connect, up to x number of > >>> times, > >>> then maybe give a warning with a manual re-connect button. Or > >>> something > >>> to that effect. I don't think it's realistic to assume an application > >>> will have a perfect network experience, and a re-connect save you > >>> headaches from unforeseen outages. > >>> > >>> Good luck, > >>> Fargo > >>> > >>> Jim Dossey wrote: > >>> > >>>> I have an app for Windows that opens a socket to a server and > >>>> receives > >>>> various updates from the server. To run it, I simply put a > >>>> shortcut to the > >>>> app in the Startup folder in the Start menu. But I found out that > >>>> some > >>>> users don't shutdown their PC's. Instead they use Hibernate. The > >>>> problem > >>>> is that the socket gets closed on the server end, so that when the > >>>> machine > >>>> is started up again, the app is still running, but the socket is > >>>> no longer > >>>> working. Is there some way to tell either when a machine is going > >>>> into or > >>>> out of hibernate mode? > >>>> > >>>> Thanks, > >>>> Jim > >>>> Win XP, RB 2007R2 > >>>> _______________________________________________ > _______________________________________________ Unsubscribe or switch delivery mode: <http://www.realsoftware.com/support/listmanager/> Search the archives: <http://support.realsoftware.com/listarchives/lists.html>
