I'm in the same boat with you. We get about two hours of use before the low memory kicks in. Our application will store data when there is no network connectivity, during this time of no network we continuously look for a network connection. The application will run fine during network connectivity, it just when there is none when the issues starts to show.
Any Motorola engineers watch these groups?
Can you share the code that you are using to continuously look for a connection? If we can recreate the problem, maybe we can figure out a way around it.
If you connect to a web service every 5 seconds, I'm surprised the battery doesn't give out before memory!
If you have a need for near-instantaneous updates to the device, there are much more appropriate ways of doing that. First of, it's what notifications are for (if supported by device platform). The notifications tell the app "something is new", then you go get it from the server.
If updates might be too frequent such that notifications are inappropriate, then using WebSockets or something similar would be a better choice.
As far as your memory problem, you haven't given us much to work with other than a vague complaint. It's like going to the doctor and saying "I hurt".
It sound like you have a memory leak. You have some variables that keep growing, etc. I'd strongly suspect that once-every-5-seconds task...
Im not the dev so Im not sure of the exact calling code but he said he was using java to check for network connection. It is every 30 seconds he tries to send (via Java) i assume the stored payload.
If it fails he is still offline.
Dan Silva should be working with my dev to find a solution. I post what we do. it may be nothing.
I think the memory issue is additive to a max time of about 120 minutes. So as long as the device is not offline between charges for this time frame we may not see the low memory.