We have an AIR application for Android written w Flash Builder 4.6 that receives network notifications for Alerts in a senior assisted living environment. This uses RTFMP to receive the alert.
This worked fine in AIR 22.214.171.1248
If the application was inactive, the app received the alerts and would play a sound and vibrate the tablet using a native extension.
Upgrading to AIR 126.96.36.199 has broken this behavior. This is a critical issue for our clients! They depend on this application to receive alerts in situations where a quick response is required.
It appears that in the upgraded version, the application receives the notifications and queues them. So the alert is not received until the user manually activates the application. This is not acceptable.
I have AIR for Android 188.8.131.528 on another device. How can I back this up from this device and install it on another?
Does Google have older versions of AIR for Android available for installation anywhere?
If neither of these is an option, then it would be a really bad day...
I'm going to forward this along to the Android team for their review. In the meantime, could you please open a new bug on this at bugbase.adobe.com? In the bug, please attach your application or sample code so we can reproduce the issue. If you'd like to keep this private, feel free to email the attachments to me directly at email@example.com.
I'd also recommend giving our AIR 3.3 beta release a try to see if it also reproduces there.
You can also find previous releases of the AIR runtime (and SDK) here:
This bug is in the bug base at https://bugbase.adobe.com/index.cfm?event=bug&id=3187377
Since the application is large and proprietary, I will create a simple, bare bones test case for you. This is very similar to the RTFMP examples posted on the web. I will be out on Wednesday, should have it for you by end of business Thursday.
We were able to roll back the AIR runtime and get our customer up and going. Our customers loves this app, the system can give medical alerts to the Android tablets in almost real time.
Thanks for your help!