So, I wrote a simple script to test my branch by looping through rebooting the phone and asking indicator-datetime what time it thought it was.
* Wednesday's runs: 'automatic' time setting with London specified as the tzid
* Thursday's runs: 'manual' time setting (as inspired by Ren's comment #8) with America/Los_Angeles specified as the tzid
1000+ reboots later, I checked the outputs to see if they always increased over time or if there were outliers that jumped to the wrong timezone. The output looked like this:
So, I wrote a simple script to test my branch by looping through rebooting the phone and asking indicator-datetime what time it thought it was.
* Wednesday's runs: 'automatic' time setting with London specified as the tzid
* Thursday's runs: 'manual' time setting (as inspired by Ren's comment #8) with America/Los_Angeles specified as the tzid
1000+ reboots later, I checked the outputs to see if they always increased over time or if there were outliers that jumped to the wrong timezone. The output looked like this:
> variant string " 7:40 AM"
> variant string " 7:42 AM"
> variant string " 7:43 AM"
> variant string " 7:44 AM"
> variant string " 7:45 AM"
> variant string " 7:47 AM"
> variant string " 7:48 AM"
For these two cases, at least, It looks like this branch is working.