This is my hope with the future PineNote running a Linux distro.

What I finally did, this June, was: I coded my own word processor for Android.

A dual boot system (Android vs Linux Desktop) would be the best solution.

There's another called Debian noroot (https://play.google.com/store/apps/details?id=com.cuntubuntu...) that appears fairly up to date but I haven't tried it. It's well rated but said to be slow. It includes XFCE and apparently starts it by default. It uses PRoot.

https://play.google.com/store/apps/details?id=studio.com.tec... is yet another, and also uses PRoot. I have concerns about their page saying they have closed-source modded versions of F/OSS operating system distros though, and that they'll only provide source to developers who register with them. That's kind of unneighborly if they're customizing BSD or Apache licensed portions of a distro and illegal if they're doing it with GPL or LGPL portions.

...But that seems to be a fork, last updated in 2013 of

https://github.com/pelya/debian-noroot

which was updated in 2020.

I'd love to test that on the Onyx Boox Max2 plus BT keyboard... I will try it and see what can be done with it.

It was started this June (2021), and I have used it for many hundred of hours with good productivity. I have not published it (yet) as I would want to polish the code first. It was quite rewarding to get it done (the best tailored product is the one you develop for yourself) - but still with a sensation of quite absurd oddity, since I had been using full-fledged word processors in the early nineties, and "there I was", almost thirty years later, with tablets much more powerful than the desktops of yore, coding a word processor because all was available for Android was unfit for use if compared to desktop tools, and I was just losing time waiting for "the product"... It's not "normal".

Wow, this sounds awesome.

I hadn't thought of the concept of writing into a multiple-column layout before. That's really really neat. It keeps the individual runs of text short (faster for proofreading, I've heard), scrolls less, and lets you keep more context on the screen.

I'm definitely very interested in this when it comes out, I don't currently have a tablet-size device but that probably won't be the case for too long and I can see myself giving this a go. A single-column option for phones (in landscape mode) might be practical.

I do completely agree that Android feels structurally broken (for want of a better way to put it) in terms of application support - it honestly feels like someone figured out how to bolt WebKit onto a feature phone and add capacitive touch and eye candy, while (somehow???) maintaining the same "smol device" fundamentals of "toy" and "not full PC". IMHO the PC experience isn't really one single cohesive vision, but rather the emergent result of years of many thousands of little independently-evolved self-sustaining threads and ideas (particularly in UX design) intersecting and bouncing off each other in a sort of cohesive balance that resulted in (comparative) miracles like Windows XP. Microsoft et al didn't invent the fundamentals of the ideology and language that emerged; rather, they just figured out how to capitalize on the net result and realized that if they knee-capped it too heavily they'd kill the thing they were trying to benefit off of.

From a distance it kinda looks like Apple/Google are trying to (re)invent everything - conventions, expectations, design language, (tens of?) thousands of tiny details, etc - from first principles, without giving sufficient consideration to whether their fussing about accidentally edits and rewrites the technological aspects of the goal narrative. I think advertising and the exponential over-valuation of user data are to blame for the current insanity - and sadly so, since it's obvious that won't change soon. I think the fundamental incentive to refine and optimize for technical experience/competence has become diluted by All The Ad Revenue™ from the current status quo, producing the current ecosystem, app store offerings, app quality level, fragmentation, etc. Basically everything seems to be working as intended, or in other words I don't mean that "current status quo is fundamentally bad and evil" (for want of a better way to put it), I rather mean that between ad revenue, internal politicking, Google/Apple/Microsoft competition, antitrust regulations, overall public perception of how invasive ads/tracking is, etc etc, the system as a whole has basically reached a steady state that simply doesn't have room for pure technical excellence, which I don't think will change until the ad bubble collapses. I'm tentatively hopeful that might actually happen in my lifetime, although as a counterpoint, hopefully it doesn't take so long the world ends up in a WALL•E-style dystopia with everything paperclip-maximized around ads instead of garbage. Haha.

However, this all provides a unique opportunity for anyone prepared to make the effort to make apps that are actually interesting: can I interest you in a "Buy" button? :P

You mentioned cleaning up the source code - and while I certainly won't dissuade you from open-sourcing something like this :D (it would be a genuine net improvement, and you could list on F-Droid as well), I'm reminded of the "source code is <license> on GitHub, but ready-to-go APK in Play Store is $.$$" model, which I think could work quite well for this sort of thing. There are a lot of small, focused apps out there that have loyal followings.

Maybe you could do an "early beta" sort of thing to cue users to understand the first versions are free but you plan to change that at some point. (The Play Store management dashboards probably make it straightforward to grandfather everyone who started with the beta into free updates for life or something like that.)

> NPU (Neural Processing Unit) Capabilities: * Neural network acceleration engine with processing performance of up to 0.8 TOPS; * Supports integer 8 and integer 16 convolution operations; * Supports the following deep learning frameworks: TensorFlow, TF-lite, Pytorch, Caffe, ONNX, MXNet, Keras, Darknet

Very odd for an EPD tablet!