02 Jan 2021

reMarkable 2: thoughts on reinventing paper

reMarkable 2 is a very nice product but its shortcomings are a nice illustration of the hidden complexity that can occur when trying to replace even simple things like paper.

I initially thought about writing a real world review but I do not write reviews here, it is against my editorial line. I will try to lean more towards what can be learnt about product design and development.

I love paper, it is cheap and versatile. A good notebook, a few nice pens and pencils cost less than 30€. Assuming you write a lot, a year of notebooks may cost you about 100€, not much.

reMarkable 2 works, it works well. It is clearly in the first tier of its class but it is a little rough on the edges.

But let’s take some altitude, a decade ago (and even more) people were seeing e-ink as a “coming soon” replacement for paper. Let’s analyze this.


Put a pen on paper and the ink is already there, to be fair Remarkable achieved quite something here, it feels good to write here.

Once you have written, you want to read, and the experience is quite good too.

Turning a page gets more challenging. My swipes are not always detected, when they are the latency is taking between 300ms and 500ms. That is enough to swipe, wait one second to be sure that something is wrong, try again. It regularly takes me more than three seconds to turn a page.

That is where thinking about searching something gets nightmare-ish. When you could simply flip pages with paper, you have a lot of dead time suddenly happening at the same moment in your user experience. To find your pages you need to relies on splitting them between files and you need to get those filenames right!


Draw a line, easy.

Prolungate this line, hard.

Understand that where you think you were writing is not precisely where your device think you were writing.

Clearly you will not draw as good as you were doing with paper. I think that the thickness of the screen glass is the best explanation for this problem.

But if you think this is only affecting the drawing experience, guess again. This problem has side effects on edition and deletion. You never manage to write exactly on the line you want to edit, and you always erase more than the tiny stroke your were aiming.

The different pens, brushes and pencils are a nice add. Some are gadgets, but I was impressed by the fun I had with the brush sensitivity. You can maybe make a bit of calligraphy, but getting to the paper experience will require another technology jump.


With computers came the joy of copy pasting, if it is obvious that on this side of things digital alternatives should lead, a bit of discussion is relevant.

The Photoshop-like experience: layers, hide, show, copy, paste, select and move arbitrary area is pretty nice.

Yet I see some missed opportunities, I would love to see a word disappear when I strike it out, but it would likely trigger problems when I doodle.

I think the features depth and expectable behaviors would justify for different modes matching different contexts on different power user archetypes: writing, note taking, reading / editing, creative.

But regardless of the aforementioned proposal, I think it is relevant to observe that possibilities offered by a simple digital alternative to paper quickly become complex user interface problems and ask the makers to balance simplicity of the first moments with richness desired by power users. All this obviously impacting the likelyhood of bugs.


Another dimension that feels obvious with digital is connectivity. And yes handwriting recognition is not a given. Much like speech to text or natural language processing, interfacing human language with machine understandable data or structured data is an open problem.

As a product designer, I guess you thought you were making a piece of paper but suddenly you need to handle data-in, mostly PDF I guess, and data-out maybe edited PDF, written notes that the user expects as text or simply drawings.

Putting data in has been solved marvelously for Amazon’s Kindle that accepts files from incoming email or provides a system wide virtual printer. Remarkable is not as good here and a few “corrupted file” messages remind you that PDF is not that easy of a format.

Getting data out is a way bigger problem, handwriting recognition works better if you know which language that is being used. If you write in many different languages it gets bad. A lot of human scribbles are not characters but have a meaning, it is really frustrating to not be able to cross out a word or black out a character.

Minor mistakes make the experience really frustrating and often useless. That is why close to ten years after the introduction of Siri, “personal assistants” are either not used or limited to recognizing songs and setting cooking timers.

Your instinct tells you that your data is all in one place on this comfortable cloud. Which should mean that all your notes and all your PDFs are available for searching. For some reason: no. The “search feature” only searches in file names. I guess that hand writing recognition is using a third party API and systematically sending user data would generate sizeable costs.

Paper does none of these, but it does not pretend do, you have no expectations about it. It fails you less.


A library is beautiful, it reflects one’s character and taste. I put on display books I want to be judged by or that can trigger discussion or curiosity.

Compared to this, a Kindle is a dull thing.

Similarity reviewing my old notebooks is something I do quite often. It is a bit of a nostalgic activity but also a motivating one, as it enables me to realize how much was achieved. It is way less likely to occur without the physical artefact around the content.

Plus you can damage a book or a notebook, but you are probably not at risk to lose your notes because you spilled your tea over it. Those transistors are fragile beings.

EDIT: about 2 minutes after losing those articles I went to pick up my tablet to realize I had lost my marker tip …


We underestimate how much we achieve with paper and how much we expect from our digital tools. But toying with this tablet and trying to look for flaws unveiled an incredibly complex topic for something than one can perceive as a simple replacement for a simple thing. Overall reMarkable sells a fine product, yet I do not know if it will replace my dear notebooks and the handful of pens I always carried with me. I got sold on this alternative to paper, but using it made me wonder if I needed one.

I like to benchmark old school ways and more “modern” ways. In the past I have benchmarked rewriting notes by typing, or dictation followed by editing, it takes roughly the same time, but the editing is a major pain.

I have done a similar thing here, exporting my handwriting, and then editing. It is fine, clearly better than rewriting everything, but the few errors are really painful to edit. Editing a text is about making it better, not fixing OCR errors.

Fräntz Miccoli

This blog is wrapping my ideas and opinions about innovation and entrepreneurship.

Since a bit of time now, I am the happy cofounder, COO & CTO of Nexvia.

Ideas expressed are here to be challenged.

About me The Dark Side Twitter