Flux Keyboard
fluxkeyboard.com
fluxkeyboard.com
Honestly, I don't think I need an LCD in my keyboard but I'd quite like to have a go on some mAgLeV hall effect switches! On that note, if you like magnets and/or keebs you might be interested in Void switches[0] -- 3D printable hall effect switches
Ed: Untested Void Switch Reference PCB Implementation [2]
[0] https://github.com/riskable/void_switch_kicad
My stationary computer has Slovenian layout (which is where I'm from), Chromebook has German (where I also bought the laptop) layout because I couldn't find the same model for anywhere close to the same price in Slovenia (which is very common, most electronics I've bought have been from Germany or elsewhere just not Slovenia, because even including the shipping it's still cheaper.
But I still set the software keyboard layout to Slovenian, which means that the stuff that's printed on the keyboard is useless to me...
If anyone is looking for an excellent website to compare prizes and find products in Europe, I highly recommend geizhals.eu, the filter options make it so that you can basically build your own laptop, they are really amazing!
Of course, this does nothing to help if your position relative to the keyboard changes more dynamically throughout your usage.
(Combined it with a BMW-designed Level 10 case; quite the looker but monumental and monumentally oversized.)
The Stream Deck isn't doing anything particularly exotic with its key mechanism, though (from what I can tell, there's a frame between the LCD and buttons with the actual contacts on it; no Hall-effect sensors or anything like that).
Definitely my first thought when I saw this too. That brings me back to the good ol' days.
I never liked the Apple TouchBar or anything that required "active" attention away from the main screen.
There isn't anything stopping you from making the entire screen black except for the legends, probably
$400+shipping for the display panel (a competitive price from the shop). Even at volume (which an enthusiast keyboard won't be) it's still very expensive.
The ED133UT2 is also available from E Ink directly and from other shops, but I haven't seen a significantly better price. There are also some other boards with the same controller around. An alternative to using the IT8951 is to hook the display up to an MCU with an adapter board for the flex cable directly and then drive the waveforms from the MCU, this is more complicated however and a little controller talking SPI is quite nice to have.
There's also a 10.3" with an even higher resolution that is very nice for various applications.
The cost of the panel makes me cry but there's nothing like e-ink :)
I have indoor and outdoor air quality sensors in my apartment; the outdoor air quality monitor is an LCD and the indoor one is e-ink. I kind of like the LCD better, but have to use a feature to turn off the screen at night so it's not illuminating the entire room with its backlight. The e-ink doesn't emit light, but it also doesn't update as frequently, so it's often displaying information that's out of date. Because of the various pros and cons, neither technology seems like a "win" over the other; the product designer has to pick one and hope the market agrees. LCDs get the nudge because of cost, though.
Except perhaps product designers wanting to make a flashy product that draws attention all the time, 24/7, pushing through '''features''' that leverage (force) full screen animations in the driver software.
I am bitter, yes, I met too many excellent hardware ruined by stupid, oversized, look-before-function attitude software. Making it grand and flashy (eventually overcomplicated and ugly) not because it is useful but because they can.
And yeah, I totally get what you mean about the Apple TouchBar and stuff like that. It might look cool, but it's just another thing that takes your attention away from the main screen. We need interfaces that are designed to help us be more productive, not ones that are just flashy and distracting.
but, them not treating the screen as a display I think significantly limits the utility of this. You have to go through their software to customize the screen, layout, and customizations. Many of us know how hit-and-miss esoteric keyboard software tends to be.
If they stuck to making it an external display managed by the OS, polling the keyboard layout from the OS, and maybe just have their app apply custom functions like emojis and tool graphics I would have some confidence.
They are trying to tie up too much to their software which they will eventually stop supporting, like my DX1 paperweight.
I'd actually rather not have to worry about how my OS treats the extra "display" - plus I don't want running my keyboard to eat into my computer's resources in any significant way.
For me, this approach is better. Although I do have some concerns about the software eventually going EOL unless they open source it.
This is precisely why I probably won't buy one despite it looking honestly very cool.
Hardware manufacturers these days seem to be allergic to telling their customers how to actually talk to their gadgets, in favor of proprietary bundles of software that inevitably go unsupported within a few years.
I also think software companies should/should have to open-source products when they discontinue them like Dropbox did with Zulip (the only example I can think of offhand), but I understand why the profit motive discourages it. But surely for hardware even that tenuous rationalization goes out the window.
I think I love everything about that sentence.
An easy-to-clean keyboard could definitely find its uses in dirty environments. Even if the screen is glass, the keys should protect it pretty well. Depending on how scratch-resistant they are. For CNC, being able to customize the keyboard seems really useful.
It's interesting that they say you can play video on it due to the built-in 8 GB storage, but it's not available as a display to the host. Their point that the "mouse cursor could get lost on the keyboard" is a nice rationale, but it does feel like an odd limitation for such a powerful device.
I wonder how complicated it would be to allow customization based on SPAs. Their list of initially supported applications suggests they are not tackling web apps at launch. E.g. the Google Docs/mail suite would be interesting to me.
1. I don't look at a keyboard so there's not much point in making it look fancy.
2. I destroy keyboards in 12 months. Even good quality mechanical ones. I probably can't afford to own this.
3. It looks like bits are going to fall off it when I inevitably catch the edge of it moving from my mouse to my keyboard.
4. It doesn't have enough keys on it and the default positioned layout sucks.
5. It's got transparent bits. My backlit laptop keyboard already looks like a cat litter tray after a couple of weeks. This is going to require a lot of upkeep.
I think the only keyboard I actually like at the moment is the Cherry Stream SX TKL. Costs $26, feels like an old ThinkPad, has all the right keys in the right place, if you break you won't cry, no bits fall off it when you shake the crap out, it doesn't look like a Christmas tree on acid or ever like a cat litter tray. I had 11 keyboards to get to that and it was the cheapest one. Ugh.
Wait how? How do they fail? This is interesting to me.
Right now my favorite is the wired Apple Magic keyboard with numpad. I bought my current one new on ebay during covid at a big of a premium because they stopped making them years ago and are NLA. I don't use a Mac so the layout is funky, but all of the standard-layout knock-offs are bluetooth-only, have poor reviews, or are staggeringly expensive.
Mostly it's spills, but sometimes also a crumb or something stuck under a key and me breaking the switch in the process of trying to clean it.
I think I also have one where I just physically lost the wireless dongle. I have some spare ones from the _other_ ones I killed, but AFAICT they're "securely" paired at the factory and there's no way to reprogram those.
The other factor is that I'm a key masher. I just don't type lightly, particularly when I'm gaming - my brain refuses to accept that pressing W harder doesn't make me go faster, and my failure mode tends to be one of:
* switch breaks/wears out
* pcb stressed/flexed by my mashing causes traces or solder joints to loosen. Keys will only work sometimes, and sometimes end up putting out the wrong letter. Pushing hard on the misbehaving key will often make that one work but surrounding keys will start misbehaving. This has happened on a few keyboards!
It's a little wasteful, but for the cost of 1 good keyboard I've found out that I would have chosen the wrong layout and wrong switch type if I has sprung for a good one right away. I also learned I need a good key plate and maybe go with wire wrapping instead of a PCB for the switches.
Well for this keyboard since it's Hall effect sensors rather than switches I actually think it would hold up for you!
As for the keyboard proper, I think its too late - a colleague and I are in the process of speccing out something for me to assemble based on my usage data, and I've caught the bug :) (just yesterday I spent a lot of time looking at keycaps of all things, I don't think I ever envisioned myself spending time on that sort of thing, but here we are)
Ok sign me up.
My main question is: can the hall effect sensors be read as variable states instead of boolean keypresses? That could be really cool, especially with the display for real-time feedback. I can think of so many use cases: change the volume that hard. WASD like a joystick.
It would be even cooler if they were in an even orthogonal grid pattern like a planck keyboard. That would be a lot like the impact-pressure sensitive controllers that are popular in audio work.
I would love something like that too use as a macro pad: keys that have arbitrary meaning you can reference. Keys that behave differently depending on context, and show you the context in real-time. That would be incredibly useful.
But my keyboard itself? I don't want to look at that. That place is for muscle memory alone.
From the article:
> They also feature a software adjustable actuation point in 0.1mm increment
So probably yes.
Yes, there's hall-effect keyboards out now that you can configure to have certain keys provide an analog input, for use in things like racing games. Since the actuation point is entirely software, you can keep track of actuation distance.
You might want to have a look at Elgato's Stream Decks.
If I do want to look at my keyboard, why have it physically resemble a typewriter from 1800?
I'm definitely not looking to remember where a key is relative to the rest.
I really like how it's Not derivative. It doesn't look like an Apple knockoff product or some other coattail riding project. Refreshing.
Keep at it! This is the kind of innovation we need!
I have one. Somewhere. Because the keys need to be as close as possible, it's less than ideal as a keyboard, and because even though the keys are right next to each other there are still gaps between the keys, it's less than ideal as a touchpad.
That's when it worked. It wasn't reliable for me.
[0] https://www.lenovo.com/ca/en/p/accessories-and-software/keyb...
The only problem is that I use a split (ergo) keyboard :/
If I could buy a prebuilt, I think it'd be my endgame.
I don’t use it and know little about the project, I just remembered that it’s a split keyboard with an optional trackpoint attachment.
Anecdotally, around 2010 I’ve worked for a company that used Thinkpad laptops, the majority of people would always choose touchpad over trackpoint.
Unsolicited advice: you should definitely never visit /r/mechanicalkeyboards.
I did in fact find a keyboard like that online -- some noname brand, and the promise was good, but not the execution. A high quality version would be welcome.
[1]: https://github.com/qmk/qmk_firmware/blob/master/docs/feature...
What is interesting is they're using hall effect (magnetic) sensors, which have a vastly longer life than normal mechanical switches. That isn't new either, the Dreamcast had them way back in the day, and today Wooting sells keyboard with the same tech. This enables some very useful gaming features.
First, you can adjust the initial actuation point, so the keyboard registers a much lighter press, resulting in faster effective response times. Like Wooting, it goes all the way down to 0.1mm.
Second and probably more importantly, it supports rapid trigger, meaning the key instantly releases the instant you start to lift up your finger, and then it doesn't have a set actuation point so the key doesn't need to lift up most of the way before you can press it down again. This allows for much, MUCH more responsive controls which is a noticeable advantage in eSports games and some rapid-clicker games like Osu.
Thirdly and not very important at all, Wooting supports analog controls, so you can play driving games with your keyboard, or adjust your movement speed in a shooter. There are all kinds of issues with support in games, though, and it's unclear if the Flux supports it at all.
The problem with Wooting is it's expensive; the keyboard costs $175. These magnetic switches need to get down to a more reasonable pricepoint.
And the problem with the Flux keyboard is, well, it's even more expensive. Starts at $299 in the kickstarter phase, and will retail at $450.
I do find it encouraging to see more manufacturers using magnetic switches, and hope prices come down at the low end quickly. They're just plain better.
I think that's close to an inherent problem with the bill of materials. Taking a look at DigiKey, the cheapest in-stock linear hall effect sensor that I see still runs about $0.40 apiece in large quantity. Turning that into a usable keyswitch requires at least two more magnets, and I'm sure the assembly isn't particularly easy.
Without some great cost reduction in hall effect sensors, I think a magnetic-switch keyboard will always be more expensive than a mechanical-switch keyboard, which will be more expensive than a rubber dome keyboard.
No doubt they're considerably more expensive than regular mechanical switches, but it's unclear how much of that expense comes from their being pretty rare right now and not benefiting from economies of scale. That's my hope, at any rate, because again-- they're just plain better.
>The keyboard will cost between US$299 - $350 after discounts for pre-orders with a retail price of US$450.
That's... not as high as I thought it might be, considering how expensive mechanical keyboards can get and how esoteric of a product this is.
I would claim that there’s nothing expensive about mechanical keyboards, there’s healthy profit margins, and passionate people.
i.e. no, it can't be run without a Windows or MacOS machine to set it up first. Oh well, $300-450 was a lot for a keyboard anyways. (Well, not bad for what this is, but I have to tell myself something to feel better)
Should be much cheaper to produce, and cost is always the problem with projects like this.
Each key is magnetically suspended with rare earth magnets allowing them to be 97% transparent and ultra low friction. Each switch has 4mm of travel and will be available in linear and tactile variants. They also feature a software adjustable actuation point in 0.1mm increments and rapid trigger functionality through analog hall effect sensing.
The keys contain magnets along their perimeter which are attracted by magnets in the frame which surrounds them. This magnetic attraction suspends them in place and provides the return force which makes the key bounce back after depressing - similar to a spring.
I got the same implants after talking to him about it on IG. It barely worked. An array of magnets on the skull would probably work. I had to get the magnets in my ears removed when the casings failed.
My guess is that if this keyboard is released and if you use your fingertips rather than fingerpads to type, you might not have a problem unless you are ticklish.
If they are really with the times, then they will use printed polymagnets:
Programmed magnets, or polymagnets are magnetic structures that incorporate correlated patterns of magnets with alternating polarity, designed to achieve a desired behavior and deliver stronger local force. By varying the magnetic fields and strengths, different mechanical behaviors can be controlled.
I already mailed them and I hope they can consider this.
and then there's maltron keyboards, a bit older-school.
I have a kinesis edge and a keyboard.io, I like the kinesis edge rgb variant. the keyboard.io is very hard for me.
Can someone just please make a high-quality MS Natural Ergonomic 4000? Mine is getting worn out.
edit- holy shit it looks like MS reintroduced it! https://www.microsoft.com/en-us/d/microsoft-ergonomic-keyboa...
I also miss the MS Natural 4K. There's a cheap knockoff on Amazon, but I didn't bother with it.
> Nevertheless, if there were a world where amazing products suddenly snapped out of existence and were replaced with husks of their former selves, it would be this keyboard. Oh, right. That is this world.
Amazon review of the latest gen ergo keyboard.
If ms does a MS Natural 4001 or whatever, a genuine successor, I'll probably buy 3. It was _really_ nice.
Best of all worlds -- Ergodox thumb cluster with traditional staggered keys. Upgraded from a Kinesis Freestyle RGB and I will tell you, the Dygma is worth every penny!
Also fwiw I don't think there's any real evidence to suggest that ergonomic keyboard layouts help with RSIs.
Rough outline to get you hacking: Run `setxkbmap -print | xkbcomp -xkb - -` and edit super/hyper/meta keys to your liking. Install result with `xkbcomp my-cool-layout.xkb $DISPLAY`.
IMO APL is best implemented as a new input method rather than a layout, add a new .mim file into /usr/share/m17n.
The UHK totally transformed how I use my desktop. I literally don't even have a mouse to plug in now if I wanted to because it's been so many years since I needed it. If the UHK ever broke or went out of business, I'd be in a world of hurt. Shipping times for the UHK are abysmal. I think it took almost a year to get my UHKv2. I spent some time searching for a similar product but couldn't find it. I don't think implementing that in software would be too hard, although it would not be cross platform like the UHK is.
Anyone know of "mouse with the keyboard" solutions?
this just seems to be a new attempt at the ol' Optimus Maximus
> Where is Flux based?
> Flux is based in Sydney, Australia. We are from the future, due to time zone differences.
- What are your plans for TouchID?
I would spend the price of the custom keyboard plus the price of an Apple keyboard to tear down, just to have that feature included. Offer a primary price point without it, add $149 for a TouchID "module", see what happens.
I pretty much came to say the same thing.
I want a repairable keyboard that works reliably across OS for a reasonable price. I'm pretty sure I am not the target market.
I would appreciate something which could understand both my keyboard firmware and whichever app has focus, and summon a hotkey helper on screen which mirrored my key layout.
Although it feels a little wasteful to always be using 4W on a keyboard. But it looks so !much better than any RGB setup I've seen, and it's got so much potential on things like a Raspberry Pi, it could be the terminal display and keyboard in one!
E-ink might be even cooler!
What's the hold up?
For me it's a simple affair: Apple Wired Keyboard. I always liked thin keyboards with minimal travel / low profile keyboards.
I love it. I’ve no use for something like this, but I’m in absolute awe at the design on display. Marvelous. I hope Apple or somebody makes this person obscenely wealthy.
People hated on Apple for the touch bar, but this is the natural evolution of the keyboard, why stick to fixed physical keys when you can have them digitalized to be context aware
The problem will be the price
Mechanical keyboard market already is a scam where 90% of products are greatly overpriced, that will not encourage them to offer a reasonable price unfortunately
EDIT:
The keyboard will cost between US$299 - $350 after discounts for pre-orders with a retail price of US$450.
Yeah, the price of a console-tier GPU, way too overpriced to be a successful product
With this I can I can keep it clean.