Please provide source code for kernel and bootloader


#14

We surely appreciate open sourcing the android parts of your codebase - but most interest in this thread is for the kind request to allow us flash custom OS to our boox and opensourcing kernel + bootloader.

This least gives us alternative in the future for more privacy minded users - as it is not possible to really use android without a google account, while not really burdening Onyx with anything. In addition, Android seems to have very large latency on eink screens. The solution to that is allowing community to design their own apps on top of generic linux - as remarkable is doing with great success.

I believe Onyx hardware has much more potential, which is now hindered by the fact that we can’t really modify it. Perhaps someone will eventually reverse engineer their way into booxes, but it will be a wasted effort. For some reason Onyx was providing kernel and bootloader as well instructions how to access the device for your old series - hope you can foster this kind of openness into the future.


#15

I have a M92 since 2012. On MobileRead, many discussions are present about the software. The M92 is a pure Linux device and the improvements were very important compared to the first version of the firmware - thanks to Booxtor/ereader.store (Mobilearead) and Onyx developers. This device is done to ‘read’, as attended for an ‘e-reader’. It is missing few days of work on the software to make the M92 a perfect reader. I would like to have the full code for the M92.

But many users complain about the lack of applications available on Android and push Onyx to use Android.
An e-ink device could not be a tablet, a small computer to type a thesis in a WYSIWYG mode, a display for all type of videos, etc.

The HDMI port is a good thing, even if I prefer a SD-Card port to an HDMI port.

I’m waiting to received my ordered Max2.

A better access to the code could give at the community the possibility to work and improve some part of the software. The modifications could be integrated by Onyx’s developers.

For me, it’s important for Onyx to keep concentrate on the main functions of an e-reader :

  • read : read all formats, crop functions, improve to the maximum the speed when the page are turned, the changing of page, have many opened files open simultaneously even if they haven’t the same format, etc.
  • annotate, export correctly the annotations.

#16

On a side note, I’m an hobbyst android developer myself, and tried developing a small app yestrday to test the response of the Max2 device, noticing the fact most apps I installed from the play store were effectively unusable. Well, the app (a simple “draw stroke” test) runs surpringly well, beside screen refresh latency of course, but no abysmal loading times (it’s a very simple app, but loads instantly), no crashes, solid 50fps, pretty useable overall.
Interestingly, it starts losing on framerate when too much “black” to “refresh” is on screen (which makes me think it’s something related to e-ink? white has no “refresh cost”, black has? just wondering) and probably just refreshing smaller portions of screen -say, only where you are drawing- would limit the issue (and I’ll probably test in the future if I have some more time),
All this just to say most apps available around are heavily burdening the system unnecessarly, and of course don’t think at e-ink screens nuances, if one keeps the app sleek and crafts with care, even complex devices with large screens, and e-ink screens, can be tackled a bit more successfully through Android. Of course, this is something most devs don’t really take into account when making their apps…
hope a more widespread use of these awesome devices will bring better optimization practices, softer apps… and more powerful, less “picky” e-ink hardwares :slight_smile:


#17

Sorry, but a 1000$ device with GPU, 2GB RAM and 4 core processor is not “just an e-reader”.

It is fine if minimalism works for you, but most of us would like to take the full advantage of this hardware, not tip-toe because Onyx graced us with buggy and unfinished apps (which hopefully will soon change) :slight_smile:


#18

It is good to keep in mind that android software stack has almost decade of optimization for LED screens. It was never conceived someone would run it on a eink-screen.

Hence the poor performance, bugs, horrible battery life and so on.

At this point I am not even sure if to blame onyx, or blame android itself. Not using correct tool for the job backfires horribly.


Can't you make onyx's software to be stabilized?
#19

Is the source code for the PDF & EPUB readers and Notes app for the Max2 available? If not, would you consider release the source code for these? That would be very helpful for the community.


#20

Yes please make it open source. It will be to your competitive advantage!


#21

I disagree with the OP though: Building on top of android is the way to go. The UI is good enough and having full android power is just fantastic. Alternative: build on top of Ubuntu, but they have not so many good apps and are not optimized for tablets … at all!


#22

+1 to open the bootloader and +1 to open source the kernel.


#23

I agree, the kernel needs to be opened (perhaps also to comply with the GPLv2 license), but more important because that would open for the community to make the software better. The latter also applies to the firmware. But perhaps most importantly is the source for the user applications provided in the firmware distribution. By opening these developers would have access to them as examples of well designed e-ink optimized apps, but also the possibility to enhance the apps themselves.


#24

As the user applications are missing a lot of features, having them open source greatly increases the devices value.
E.g. I want to automatically export all my fingerings of my ~20 pdfs, so I can move and rename them safely on my desktop pc.
Losing these fingerings would be very very unfortunate. Having access to the source of the neo2reader app, I could implement this backup feature by myself.


#25

Seeing that this thread has been going on for half a year without any response from Onyx, I think we should abandon any hope of peaceful cooperation.

What needs to happen is someone reverse engineering the reader, or perhaps wait until ReMarkable hopefully release larger version of their open source pad.

My Max 2 has been relegated to simple PDF reading, almost every other feature has bugs, including notes and so on.


#26

I’ve requested this as well at Wishlist for Firmware 2.x , and there’s another request open at Developer support?, waiting for a response.

I’m happy to see that the repo referenced at Please provide source code for kernel and bootloader seems to be active, with about 30 commits this year.


#27

+1 to open the bootloader and +1 to open source the kernel.


#28

Yes, please.

I’m fine with developing on Android, though.


#29

+1 to open the bootloader and +1 to open source the kernel.


#30

+1 to open the bootloader and +1 to open source the kernel.


#31

+1 on both Kernel and Bootloader


#32

Created an account just to +1 this.

Thanks!


#33

It seems unlikely that this will ever occur. If you poke around and dump the system files, you’ll find evidence of IP theft from other apps, dictionaries, and open source projects. It’s possible to get root on this device, but without a bootloader unlock I don’t know what other options there are. For people who are interested, there is some discussion going on at the mobileread forums.