Hey Capn,
Is there any way that the 6.0 installer could default the []Close Window On Disconnect to ON?

For a customer this would almost always be the case. Just learned from a customer that all the "connection reset" complaints were when they were logging off the system... :rolleyes:

I'm not sure that adjusting configuration options in the installer is a good idea, because the most common installer-related request/complaint is to make it simpler/too many prompts. And any option we hard code one way or another will inevitably make some people happy and some unhappy.

I guess the reasoning behind the message was so that when the users went to lunch and then came back and their session had timed out or otherwise been disconnected, it would be nice to see what the last screen before the disconnect was.

Unfortunately, we never came up with a standard protocol for how an application should distinguish an orderly/normal shutdown from a error-induced one. One idea would be to add an optional parameter to the AG_ATEDISCONNECT function to allow you (the app) to force ATE to close quietly.

I'd appreciate that parameter as this dialog displaying is causing my ATE update routine to not work. I'm closing ATE with the AG_ATEDISCONNECT when an update is needed, and the user doesn't either pay attention enough to see the "connection reset" dialog on close it, or they don't know they need to close it.

I asked Jack about this a couple of weeks ago, and he told me there was a Registry setting that controlled this. I've not modified my code yet to default the setting to just close, but I've noticed that I sometimes see the window on my machine; eventhough, I have the just close option checked.

This is a new prompt from 5.X to 6.X i believe. With that being said, shouldn't the backward compatibility model apply and have the installer check it?

If not, and/or its not feasible, perhaps in our ZOOM flavor you could please toggle this to ON?

For what it's worth, the feature was added in 1213 (April 2011) (see 6.0 Dev Notes ), and as stated in the notes, it was supposed to default to checked (i.e. compatibility mode). And it certainly is the case that if you manually remove the setting from your registry configuration (to simulate an older configuration from before that option was added), and then go into the ATE configuration editor, you'll see that it got checked automatically.

However, in testing this now, I see there was a loophole in ATE itself was interpreting the absence of the setting as being equivalent to 0 (not checked), and if you then brought up the configuration dialog from within a connection and saved it, it would then get permanently written to the registry. So I'm guessing that is the sequence of events leading to it being unchecked (or acting like it is).

That will be fixed in 1258.0, which will also include an option in the AG_ATEDISCONNECT command to override it.

Does that cover all the bases?

All but needing the patch in 6.1

Ok, how's this:

6.0.1258.0 ashw32 (zip)
6.0 Notes

6.1.1316.5 ashw32 (zip)
6.1 Notes

Thank you sir.