Would someone be willing to update the disable sickel timeout patch
first introduced in 4.10 to work with current firmwares? And/Or find a way to make the timeout value user-configurable rather than hard-coded at 20 or 30 seconds (or even better, be able to disable sickel from kicking in entirely; it's really easy to force a hard power-off with the power button if need be if the device really has locked up after all)?
Spoiler:
Code:
sickel reboot timeout (20 sec):
- Enabled: no
- Description: increases sickel reboot time if no ping from 10s to 20s
- PatchGroup: sickel reboot timeout
- ReplaceBytes: {Offset: 0x003498, FindH: 42 F2 10 71, ReplaceH: 44 F6 20 61}
- ReplaceBytes: {Offset: 0x0034AA, FindH: 42 F2 10 71, ReplaceH: 44 F6 20 61}
- ReplaceBytes: {Offset: 0x0036F6, FindH: 44 F6 20 61, ReplaceH: 49 F6 40 41}
sickel reboot timeout (30 sec):
- Enabled: no
- Description: increases sickel reboot time if no ping from 10s to 30s
- PatchGroup: sickel reboot timeout
- ReplaceBytes: {Offset: 0x003498, FindH: 42 F2 10 71, ReplaceH: 47 F2 30 51}
- ReplaceBytes: {Offset: 0x0034AA, FindH: 42 F2 10 71, ReplaceH: 47 F2 30 51}
- ReplaceBytes: {Offset: 0x0036F6, FindH: 44 F6 20 61, ReplaceH: 4E F6 60 21}
I'm asking because
I saw in the 4.18 firmware thread some users complaining that some books now make their devices inconsistently reboot when open, and I suspect a larger timeout would help work around it. On my device with a ridiculously large collection/database, everything is slower so anything I open that needs the Adobe renderer (regular epubs, pdfs, cbz, etc.) takes longer and will usually trigger sickel consistently (I personally needed at least 120 seconds myself to be safe in a few cases; maybe longer now since my collection has only grown since 4.10 but 30 seconds definitely isn't enough for me anymore), but I now use koreader for those files. An updated sickel patch (with a larger default timeout or even better, sickel disabled entirely) would allow me to start using nickel for those files again, though (which would be nice since I find koreader eats up more battery in the same amount of reading time compared to nickel which I doubt much can be done about without entirely recoding koreader).