Have I bricked my Forma, perhaps by using KOReader?

[NiLuJe]

@Pajamaman: Stop. That's not a KOReader crash. You can tell, because KOReader hasn't, in fact, crashed .

That's what a (faked) KOReader crash looks like:

[notStanislaw]

Thanks again for all the help.

Past days I've been pretty busy, but also just enjoying reading.
I've done a tiny bit of stresstesting. Nightmode on, quickly flipping pages in epubs as well as pdfs. No issues there. Favorites screen hasn't changed and it opens just fine.

So far the battery holds up just fine as well.

So in the end I opted to cancel the exchange and sent the new one back, without even having opened the package. It's still under warranty for a couple of years, so if something's amiss later, it'll probably get resolved just fine.

Such a nice device, it's so much better than lugging around a thirty tonne hand-me-down android tablet with a short battery span for reading texts.

I should probably open a new thread for the following, but I had the idea to contribute to KOReader by making a dvorak keyboard layout for it. Those lua files are simple enough. At the very least I'd do it for myself, to have the same layout across devices I use, but I'd gladly share it.

[notStanislaw]

My Forma froze again last weekend, this time as I was about to load more content into it. It was on the "Starting USBMS session" screen.
Even the charging LED remained on. Again, no matter what I did with the buttons, nothing worked. I had to wait multiple days for it to be completely discharged as there was no dedicated reset button or a way to remove the battery, as far as I know.
I suppose I'm lucky it was only halfway charged and the LED kept being on to drain it faster.
Anyway, I'll just assume the hardware is to blame and I'll ask for an exchange, as this is the second time this happens. The Forma and KOReader have become essential for my thesis workflow and I can't be waiting for multiple days every time it decides to freeze.
Not sure why I share this exactly. Again, I suppose it might help someone who is in the same boat.

[PeterT]

Did you try holding the on off switch down for 20 to 30 seconds? My experience has always been that a hung Kobo will power off at that point.

[notStanislaw]

Thanks for the input, PeterT!
I did, multiple times, the only result being my fingertips hurting when I spend an hour or so trying anything from 1-40 seconds of holding that weird mushy button. I've tried and I hope to save time troubleshooting by putting it back into a factory state and exchanging it for one that hopefully has a more reliable power button and even better, doesn't have these random freeze issues.

[NiLuJe]

I may have mentioned this the last time, but the long-long-press to hard-reboot may require the device to be unplugged. (Since you've mentioned an USBMS session, it's ambiguous whether you tried that with the device sill plugged in or not ).

Also, FWIW, since you've mentioned the charging led: it's somewhat meaningless, because mostly entirely controlled by software (and that mostly form userland, in fact).

[NiMa]

Originally Posted by NiLuJe

I may have mentioned this the last time, but the long-long-press to hard-reboot may require the device to be unplugged. (Since you've mentioned an USBMS session, it's ambiguous whether you tried that with the device sill plugged in or not ).

Also, FWIW, since you've mentioned the charging led: it's somewhat meaningless, because mostly entirely controlled by software (and that mostly form userland, in fact).

Software... do you mean Nickel? Or early U-Boot and kernel?
From what I learnt hooking my Kobo to serial, the long, non-blinking light indicates that U-Boot's loading the kernel, then when it starts to blink, it means the kernel is booting up Nickel or whatever else OS is in there (like Alpine and InkBox).

But yes, I agree that the light is turned off in userland. For InkBox, I had to write in some brightness value in /sys/class/leds/some-path-that-i-dont-remember to stop it.

Also, on my Glo HD, the light turns on (but doesn't blink) when I turn it on and there's no microSD card in it. On the Mini though, it doesn't. Probably that the board is in cause there...

[NiLuJe]

All of the above ;o).

(In this case, what I was getting at is that it mostly almost never has anything to do with charging when it's handled by lower level stuff than Nickel ).

[notStanislaw]

Originally Posted by NiLuJe

I may have mentioned this the last time, but the long-long-press to hard-reboot may require the device to be unplugged. (Since you've mentioned an USBMS session, it's ambiguous whether you tried that with the device sill plugged in or not ).

Also, FWIW, since you've mentioned the charging led: it's somewhat meaningless, because mostly entirely controlled by software (and that mostly form userland, in fact).

I did unplug the device when I noticed it was frozen and didn't want to mount or be found by calibre. Nothing worked. I suppose it's cool to know the led is meaningless. So, does that mean the software decided it should be kept on?

Let's hope the new one will not have the same issues. I will save a backup of the contents of the old one. If there's interest in any logs or settings files etcetera, let me know.

[NiMa]

Originally Posted by notStanislaw

I did unplug the device when I noticed it was frozen and didn't want to mount or be found by calibre. Nothing worked. I suppose it's cool to know the led is meaningless. So, does that mean the software decided it should be kept on?

Let's hope the new one will not have the same issues. I will save a backup of the contents of the old one. If there's interest in any logs or settings files etcetera, let me know.

Strangely for no particular reason my Aura decided to corrupt its eMMC yesterday. I tried to turn it on, and sometimes the led stayed on and did not blink until I force powered off it. I don't know if the LED in the Forma is managed by the board or just U-Boot/kernel, but it could indicate some interesting things.
Fortunately, I hooked mine up to serial console and after some reboots it decided (how) to de-corrupt (how) the eMMC and boot into the recovery partition. Now it is all working fine. Maybe it was just mmcblk0p1 the problem, but still, it's strange.

If you can, maybe hooking yours to serial (a simple Raspberry Pi Zero could do, with some Dupont connectors, or you could also get an USB to serial adapter (3.3V))to see if the device (if you still have it) shows any signs of life and check the logs if they are printed out.