I would like to try out /e/OS on a FP5 while waiting for the next official update to 3.1. Currently the security patch on this device is newer than the one on /e/ 3.0.4, so I would not be able to re-lock the phone yet.
If I do install that last version, will I then get an option to do a security update OTA that will change the OS’s security patch version in a way that will make it safe to lock the phone again?
Or will I need to entirely download and install the new version over adb again when it is released?
The FP5 installation guide is unclear on this course of action, as it mentions the “rollback index” only being updated when the device is locked, and rollback errors being “ignored” when unlocked, but does not clarify whether an OTA update while the phone is unlocked will get rid of that ignored error entirely upon re-locking, or whether the ignored error will first come into effect upon re-locking and brick the phone because the rollback index could not be updated while the device was unlocked.
inb4
If it is safe to install-old-then-update-ota-then-lock, yes I am aware that re-locking will still wipe the phone again and I will need to back up anything on there.
A week or two of waiting is not insignificant to me for this, as I cannot currently make or receive calls on my FP2, nor receive needed OTP texts to make payments, because my network no longer supports it.
That bit’s not true though, is it? I wouldn’t be having this conversation if there wasn’t a risk of bricking the phone right now.
Can anyone please confirm that they have specifically tested doing this after running an OTA update for the security patch, to an otherwise unsafe to lock bootloader, or do I need to be the first?
So from those posts it looks like there are a couple of people in similar situations on different devices.
If this is an anomaly, has nobody here done this before on previous releases because you haven’t typically gone for a month where the latest /e/ security release was marked as a few days behind the latest Android one?
I found it hard to believe that nobody would have tried it by now to confirm it, but that could explain it.
The business of “two dates” seems not to have been a problem for custom ROM builders and users till fairly recently. Loosely speaking all contributors of patches are aware of a “due date”.
The effect for general users was that the bulletin was usually published on 5th of the month and all patches appeared as if passed to AOSP all on that one day. One might previously regard that date as fixed without reading the small print.
Careful full reading of the bulletin makes it more clear than my very loose interpretation.