rmfakecloud VS rmrl

You can upload/download files from the cloud with rMAPI. You can also use rmFakeCloud to host your own sync cloud, so the files just go to your own computer/server. This works with rMAPI as well.

There's also a sync server you can selfhost, with a little bit of added effort: https://github.com/ddvk/rmfakecloud

Device is a remarkable2, its synced via https://github.com/ddvk/rmfakecloud, and handwritting recognition / org-export using tesseract and a small rust wrapper

Definitely not low-tech, but I just can't go back to paper. Its nice being able to search up what were your todo's on a certain date and keep digital track of what you missed :)

I don’t use Remarkable’s cloud! I run rmfakecloud inside a Docker container.

Might be good to clarify then or provide an example with https://github.com/ddvk/rmfakecloud because with mention or RPi and Docker I assumed, and I imagined other would too, there would be no official RM cloud requirement.

You can get around this if you're willing to learn and put in the effort. Install Toltec and you can install rmfakecloud on your own server to sync with. Toltec will also allow you to install a bunch of other stuff and makes the device much more useful. I would get rid of mine if these hacks weren't available.

That doesn't sync with the main UI though, I'm not sure what options exist for that. You could run your own rmcloud perhaps: https://github.com/ddvk/rmfakecloud (I haven't tried yet)

Cool project! Note that if you want to support the Remarkable scribbles, there's a Python project that does that:

https://github.com/rschroll/rmrl

If you need a high detail at the cost of cosmetics (e.g. pencil texture), rmrl is a pretty good renderer: https://github.com/rschroll/rmrl

100% and yet it's still worth it. I write this as a programmer meaning that, ironically enough, my distraction is when I pick the reMarkable to start reading or sketching but despite the minimalist setup still get distracted. Not by social network notifications or the possibility of a web search but rather here for example to suggest you https://github.com/rschroll/rmrl for your PDF and if its too limited (it always is) to consider what prototypes I could build that challenge the very way I read and write. I believe it's worst for researchers because the process itself must be described in order to publish a result. One must step back and describe the experiment so that it can be both challenged and reproduced. Consequently there is always on the back of the mind a simple and justified question "Is this the right way to do that?". I believe it is quite taxing but it still worth because, and that's just my view, thinking itself relies extensively on tools. We like to imagine that it's a pure process of the mind but for any complex enough thought, we need tools. We run simulations, we sketch diagram, we organize a bibliography which represents the thoughts and experiments of others, etc. This is literally unmanageable without tools regardless of ones "intelligence". This in turns mean that the better the tool, at least in regard to the final goal, the further one can go.

Then use https://github.com/rschroll/rmrl to convert the data files you get from the previous point to pdf.

I love this tool: https://github.com/rschroll/rmrl

Reading/parsing .rm files has been already implemented by these 2 Python libraries: * https://github.com/bsdz/remarkable-layers (not maintained any longer) * https://github.com/rschroll/rmrl (looks up-to-date)

This would also be less of an issue if you ran your own infrastructure (see rmfakecloud, or use something like Syncthing or Nextcloud together with rmrl for file conversion).

This has been a gripe of mine for a while. The problem is that reMarkable isn't implementing the PDF standard correctly. Their highlights are just visual overlays but not Annotation objects. I have submitted a pull request to the RMRL python library to fix this, but the author has not been responsive. I also submitted a support request referencing the specific standard, but I'm not holding my breath. In the mean time you can always try using my version of RMRL, but it requires use of the command line and a working python install.

One way to manage the issue, would be to look into rmlr, install python and pip via Entware/Toltec, and try to have rmlr running on the RM if all the dependencies can be found for the ARM architecture.