update-manager service should try to run the pip command after ubo-app loads completely

[sassanh]

It rarely happens that when the pod reboots to complete an update, the update-service removes the old ubo app just when ubo-app is loading, it makes the ubo-app fail to load. So while the update is in progress in the background, the user sees a blank screen and can't see the progress.

To fix this we need to make sure update-manager runs the pip install command after ubo-app has loaded completely.

One way to do so is using systemd api, but I'm not sure if it supports making a system-level service need a user-land service, and since update-manager is a system level and ubo-app is a user-land service, it may not work.

An alternative would be to have a signal mechanism (probably just creating a file in file-system or things like that) so that ubo-app can communicate it has been loaded successfully to the update manager.

Either way, we need to make sure update-manager doesn't stay waiting for ubo-app in case ubo-app has been corrupted and can't load, so we need a time out for this waiting, after which we continue the update process.

We can consider showing a message via a still image on the LCD, in case ubo-app doesn't load. Something like "the update process is in progress in the background, please be patient." This is to let the user know things are not broken.