Glider – open-source eInk monitor with an emphasis on low latency
github.com
github.com
I love people and projects like this.
When this kind of stuff gets in the open like it does here, I expect rapid innovation and disruption from the crowd.
Does anyone know why the Kindle is such a bad product? I use it because I like e-ink and the e-book market is comprehensive, but I don't think it's actually a good device.
Could you clarify who the "they" is? The way you write "not innovating on responsiveness" sounds very confident. Do you have a background in electrophoretic particle physics or industry expertise which is why you're saying that? Just want to make sure this isn't like some case where a commentor is saying "the only reason we haven't established a homeland on Mars is because rocket companies are not innovating".
A lot of time was spent integrating social features that no one uses. That time could have been spent on quality & latency.
I understand their business goals and objectives. It's still a low-quality product.
A profitable product can also be terrible.
To be clear, the displays are not created by Amazon / Lab126. Instead, they're a product of Eink Holdings, Inc.
From what I remember, most of the screen refresh algorithms etc are Eink IP. And by the way, the cost of the display module alone was eye-watering, especially when compared with LCD displays...
With e-ink, you can drive it faster, at the expense of massive power consumption or terrible ghosting / artifacting. You're not going to get the 6 weeks of use out of a battery doing that.
For reading a book, smudges / ghosting sucks, so they optimize for full screen refreshes just often enough to clear that up (that's when the screen goes black then white, followed by the update).
It's kind of a physics based fundamental limitation -- the display is closer to a mechanical display of old than an LCD.
The kindle is a product that does one thing well: display static text in any lighting condition with a similar quality to the printed page.
Layman here, but what you describe sounds very much like innovation held back by patents:
At the core, it’s really promising tech with actual major advantages over LCDs with applications already in many domains and possibly many more in the future; all you’d need really is incremental improvements, similar to the journey of LCD. Remember the shitty TFT(?) monitors from 20 years ago? Ghosting, low resolution, delay, low contrast, backlight bleeding, etc.
If we hypothetically had 20 companies competing the traditional way, throwing international manufacturing and material science know-how on these bad boys, I’d bet $100 that we’d see massive gains in ability at a fraction of marginal cost – from incrementalism alone – way before you reach physical limitations. And with a bit of luck, there might be a breakthrough in the core tech as well.
> It's kind of a physics based fundamental limitation -- the display is closer to a mechanical display of old than an LCD.
I hear you. But brilliant people have been wrong about these statements in all kinds of areas before. Could you share more detailed what those hard limitations might be?
It's not patents, it's economy of scale: LCDs ship billions per quarter and are used in phones/watches/laptops/PC monitors/TVs/car-dashes/coffee-machines/fridges/kiosks/etc etc etc, whereas e-ink screens are used in e-readers, supermarket tags, e-notes (stylus tablets), and basically nothing else.
When LCDs ship orders of magnitude more SKUs, they inevitably have lower costs. That's just economics.
Besides which, Amazon ships Kindles at-cost, there's no way they'd be price-gouged - if E-Ink tried to screw them then they'd buy E-Ink Corp. It wouldn't even be the first passive display company they bought. See: LiquaVista.
>Could you share more detailed what those hard limitations might be?
The ink in the e-ink needs to be shuffled up and down with each refresh, but if they're pushed too quickly then they pound the capsule they're in and damage it, or get permanently stuck. Either will break the display. And it's powder not a solid object, so the display needs to move all the ink, down, or you'll have ghosting.
Imagine you own E-Ink Corp. You know Amazon needs your screens. You can sell them the screens or you can sell them the company but you can price-gouge them either way. (Of course E-Ink Corp is a stock corporation. The ability to price-gouge amazon is priced in. There's no reason to assume Amazon would save money by buying the company.)
AFAIK manufacturing e-ink displays is still difficult.
I also have worked with e-ink displays for hobby projects and you’re flipping tiny balls of ink. Unfortunately e-ink displays are extremely slow and it only gets worse if you want colors or anything you might want in a display. E-ink displays look cool and sound cool but really suck to work with.
With the rapid rate of development recently, I would expect eInk displays to break the "magic" 24 fps barrier in 2025 and hit the mainstream in a major way. Considering that offices worldwide have been constructed to block out sunlight to accommodate display use, this tech has the potential to change everything.
Unless you mean stylus-drawing is higher refresh, which is completely different tech as it's driven by the stylus and can't refresh that fast without the stylus at that specific point.
Anyone seriously claiming e-ink screens can hit 24FPS (the whole screen, not refreshing individual pixels separately for an interlaced illusion of higher framerate) is simply ignorant. You're talking about a whole order of magnitude difference, when the core problem is a straightforward fight against physics.
The special chips enable the eink display to have a better refresh rate, but it is still the eink displaying with the faster refresh rate.
Current cutting edge is at 14fps meaning 24fps is not impossible.
People almost never use the whole screen at once. Not when typing, not when moving a cursor. And not even when watching video, because modern compression formats update the parts that are moving.
And the laws of physics do not stop anything from moving more than 14 times per second, as you might know from your car engine or smoothie blender.
Sorry, but this is like saying the laws of physics do not stop you from moving from Mars to Earth 14 times per second. In fact, they do. The same is true of moving ink particles within a high density high viscosity physical medium and getting them to stay at a specific location once you're done moving them.
Also "almost never use the whole screen at once" makes no difference since the rate of movement of ink is the same even if you were only trying to change 1 pixel.
First of all, you can't tell me what I can and cannot do.
Secondly, if they can do 14RPM with current eink displays, then 24RPM is not very far fetched.
They can't. Anyone claiming otherwise is misrepresenting tricks as actual refresh rate. You can convincingly fake higher refreshes by e.g. refreshing pixels independently, but any given pixel can't be refreshed at 14RPM, not even close.
Please help the rest of us understand what you mean. What specific algorithms are you referring to?
They have made significant improvements in responsiveness over the years. Do you remember how slow screen refreshes were on the original Kindle? It's just that that is not a high priority for their main use case of linear reading.
If you want to use it for reading PDF reference books, you probably should look to one of the eink Android tablets that are more general purpose devices and may have a faster refresh rate.
Goodreads integration is the best! I track books I want to read with it and then when I’m ready for a new book use the integration to download a sample on the spot
I've heard similar pitfalls about Kindle Scribe, the write-able Kindle.
When I am trying to focus on reading a book, I appreciate that the Kindle doesn't have too many bells and whistles. I don't want notifications popping off and the distraction of fast Internet access.
Not saying you're wrong, just that your anecdote doesn't mean much about $100/$50 Kindles.
I'm talking about aesthetics. In this case, elegance, utility, responsiveness, durability, efficiency .
Unless of course you read books that have graphics or come in pdf form.
I consider myself a heavy user of ereaders, probably averaging over an hour a day, and have had a few in my time (Kindle paperwhite, Kindle Voyage, Kobo Aura H2O, latest the Kindle Oasis - which I got in Jan 2019). The only real time I felt like I got an "upgrade" was the slightly denser screen on the voyage, then the slightly larger screen on the Aura. The Oasis feels functionally the same for my use case (again, 99.9% text, no images or diagrams or animations, so can't compare them). But really the only reason I "upgraded" after the voyage is leaving it on a train or got damaged in my luggage one time.
And what can they really improve on it? There are denser screens - but they don't make the font size I use actually look any better. There are larger screens, but it's already about the limit of what I want for portability, and much larger would be harder to read not easier. There are faster refreshing screens, but I've never even noticed the page refresh speed, as it's lost in the time it takes to realign my eyes to the top of the page anyway.
So what's left? Looks? "Premium" materials? Thinness? The oasis is already at the point where any thinner or bezels any smaller and you couldn't hold it. I don't read comics on it, so don't really need a higher resolution, or colour panel (though can see the advantage if you did). Same with flash size - every book I've ever read on it (nearly 3000) is still on my 8gb model, and it's not even at 50% capacity.
And now my Oasis battery is starting to fade - it doesn't more than a few days anymore - and somehow it feels weird considering buying the same thing again - or pry open the back off and try to replace it - though who knows what that would do to it's water resistance. I guess the newer models have an amber backlight option, which would be nice, but still doesn't feel like an upgrade to something I purchased over 5 years ago now? Or even more as I only got that because the (functionally equivalent) predecessor got crushed on a plane.
So I think my point is that for the "Reading Books" use case, a kindle has already reached the maxima. Any "improvement" would almost be a waste - why improve refresh times when it won't actually affect the user experience one bit?
Is there something I'm missing? Is this just a local maxima where you'd look back on and feel stupid for not seeing the "obvious" improvement path? Or is ereader "development" just a waste now as it's now just a commodity?
A handle. My hand is not L-shaped. Make it hand-mirror shaped and shift as much weight as possible into the handle. Preferably a fold-down handle.
A folding screen. Or rather, two screens with a hinge, no need for fancy flexible screen tech. It doesn't matter if there's a visible bezel between the screens, you're only displaying text. Ideally, you'd cram it just small enough that it can fit in your pocket when folded.
Also, once you have a handle you don't need bezels.
I can't imagine holding a hand mirror for any length of time comfortably - especially if there's only a single orientation you can grip it with the screen upright - but that may be the weight distribution as you mentioned.
Folding would be useful for putting it in a bag or fitting it somewhere, which hasn't really been a big issue for me. A flat profile is already fine when putting it in a laptop bag - it's thin enough it just slides down the side of the pocket. I wouldn't be surprised if decreasing the total volume is pretty much at it's limit already with being able to be held - but maybe putting that is a more square package when stored would be easier in some cases?
So maybe zines and comics is where ebook readers are heading.
Its UI seems oriented to promote Amazon Store and I never used it, sending books over e-mail and deleting after read, that's OK with me. I'd prefer for its library to have folders and I'd prefer for it to work as USB stick like other ebooks do, so I can connect it to PC and organize things inside as I want, but those are not necessary.
So may be Kindle is bad, but rest are worse, I don't know a single ebook brand of Amazon scale. They all seem to be Chinese no-names which come and go without investments to quality and reputation.
https://us.amazon.com/gp/help/customer/display.html?nodeId=T...
Have you tried Calibre? https://calibre-ebook.com/
Works perfectly.
I like tools that do one thing well. The Kindle has hit that spot for a long time. There were incremental improvements (faster processor, 3G/4G, front light, higher DPI / contrast, etc), but it's surprising how similar a 2010 kindle is to a 2024 one.
It's the same complaint people make about iOS devices degrading to force upgrades.
I don't think it's deliberate but I do think it's deliberately neglected.
Main purpose is reading a lot of sheet music documents, making annotations and share with choir/orchestra. I discovered that Calibre is excellent tool for organisation of large database of sheet music documents.
What surprised me most is how responsive drawing/writing by stylus is. Weird part is that apps need optimalisaton for fast responsivness which some apps have and some don't. Which mean eink technology capabilities are limited by SW quite significantly.
Because it’s made by Amazon
The Kindle excels at being a low-cost device for reading novels, or linear non-fiction with no graphics etc. For anything fancier, it's simply not built for that.
Pocketbook particularly takes ages to reflow text if you rotate it. I think it's reflowing the entire ebook to get page numbers + chapter positions? Very annoying if you forget to turn off the accelerometer.
Oh, and speaking of responsiveness, I found that it depends on temperature a lot. E-ink apparently has a sweet spot at room temperature, but when I use the device outdoors, in either sub 10 ˚C temperatures or over 30 ˚C the screen changes noticeably slower.
it's neither, it's a physics issue. Moving electrically charged particles isn't going as fast as making current flow flow through semi conductors
But the software is awful and the application of the eink hardware is terrible too.
Taking as many books as I want onto an airplane for a business trip is great. My kids and wife read theirs literally daily. I’d have to look up how old they are, but the newest one is 2 years old and I think mine is around 10 (it’s a first gen paper white).
The e-ink display updates fast enough to not distract me from reading. The battery lasts multiple business trips, even on my very old unit.
I’m surprised how negatively you feel about it, given my and my family’s very good experiences.
The complaints I have are all qualities within their control that seem to be lagging due to neglect (responsiveness, UI, stability, degrading performance over time, keyboard, search)
Now ignore the library aspect and compare reading a single book to reading a single e-book and the gaps will be more visible.
I read 100+ books a year on my Kindles.
The device itself is neglected, underpowered, unresponsive, unreliable , sluggish with a terrible UI.
Decompose the concept.
UI is great! I can easily read on the thing. What is “terrible”? Who cares if the device is underpowered or whatever? I tap on the screen and things happen. I’ve owned nearly every kindle ever released and never had any of the complaints you do.
It’s great for reading, I’m not mining bitcoin on the thing.
It’s an excellent single-use device.
What exactly do you want the Kindle to be?
Despite being a mature device it has the design and performance of a beta product
Beginning of this year, they were apparently looking for a Windows 3.11 Administrator [0]
[0] https://www.heise.de/news/Deutsche-Bahn-sucht-Admin-fuer-Win...
It turns out I had a very minor stigmatism. My eye doctor did not recommend correcting it, but upon correction my dry eye and eye fatigue completely went away.
So, get your minor astigmatism corrected via computer glasses, regardless of the eye doctor, best practice for minimum prescription strength.
And like you say, even though it's a minor abberation, the difference it makes in fatigue is incredible, especially after a day of coding work behind a screen..!
Straight up the best 200 bucks I spent in terms of QOL improvement. Well that and a very nice thick blanket to sleep under.
What is this?
I got computer glasses from that, no special blue filtering or whatever, just set to give me the best view.
There are a lot of special lens technologies and coatings that are advertised as being able to reduce eye strain when using digital devices. Most of these have little or no evidence to support their efficacy, so be wary if they try to upsell you on expensive extras.
So much great information in your readme about Eink. I've already read a decent bit of it, and am going to reference it for the years to come!
I get the feeling that all this flashing and latency of e-ink displays is entirely due to the open-loop nature of the controller.
how far are we from magazine quality look and feel using eInk?
like there is a scene from an old 80s sci-fi movie dude whips out a game magazine and the screenshot of games are fully animated videos...
ive been waiting 30 years now for this tech
There seems to be a direct tradeoff between contrast/image-brilliance and latency/frames-per-second. Many monitors can be switched along that spectrum, but the current tech doesn't seem to be able to deliver both simultaneously at a reasonable price.
You have to fight with the settings to just get something that is vaguely usable, and set up the OS with custom colors and an accessible theme. But even then it’s so much more straining to use than a normal monitor, because of the reflections and shit contrast.
I think you could probably get something more useful by building a 28“ wall paper of game boy colors.
Worst purchase ever.
https://en.wikipedia.org/wiki/Structural_coloration#Variable...
Flexible e-ink displays
Color eink displays
High-ish refresh rate eink displays
The problem is AFAICT there's no device that covers all 3. Color displays have pretty big tradeoffs in terms of resolution/contrast/latency. However they're still way ahead of where they were 5-10 years ago.
The reason e-inks are so expensive and progressing so slowly is that it's a niche product which doesn't drive much R&D spending (compared to LCDs, which sell billions per quarter and you're reading this comment on).
Will you be working on such a product? :-D
and I'm surprised I keep reading such claims on HN and everytime anyone ask which patent you're referring to, you get no concrete answer. It is like saying "microsoft holds lots of patents on operating systems" and that's why operating systems are so terrible and everyone is still using windows.
https://github.com/Modos-Labs/Glider/blob/main/assets/glider...
It lets you auto "re-flow" diagrams using different algorithms (e.g. fewest intersections between arrows, center-out, etc).
Anyway, I appretiate the waveform format documentation and tools. Might kick me back to working on my Pocketbook display driver.
The Limitations section was interesting to read about. As I was thinking about it, I began to wonder if 1 memory cell per pixel could be used in an analog fashion (if it could mimic the panel response) with a high speed ADC. Seems more complicated and less accurate that what's described, and perhaps not cheaper if possible at all.
Yes, it has. Even purely from the consumer facing side: the latest panels are far better in grayscale depth, sharpness and refresh rate (and their controller still kind of suck). Not to mention colour e-ink panels in consumer products seeing rapid improvements gen over gen.
> and no sign it will, as it's controlled by one company.
That's less true today, and there are multiple implementations of this idea that are being commercialised.
Though, sadly, e-ink the company bought one of them...
I was hoping to get my hands on some functional DVI/HDMI controller. Well never mind, I'll wait.
based on this tweet: https://twitter.com/zephray_wenting/status/17901730074884506...
> Open-source E-ink monitor. Mirror of https://gitlab.com/zephray/glider