A bit ago I accidentally bought a Keychron keyboard that didn't have QMK support, but then found that someone had ported it anyway, and I was able to flash it and fix some problems I had (the main one being that it seems like the default firmware only scans a single row of keys while asleep, possibly to save energy, but I want any keypress to wake the computer.) Maybe it's not perfect, but the community around open source keyboard firmware is great, it's a good example of how open source is beneficial for end users, especially power users.
And, in an era where vendor software and firmware for all kinds of devices is constantly having vulnerabilities, it is possible the liability of doing custom shit will eventually outweigh whatever benefits hardware companies seem to think it has, versus just having industry standards and free software instead. But, maybe that is wishful thinking.
My only complaint with Keychron is that of 5 or so keyboards I've owned from them I wound up having to resolder 3 different hotswap sockets on 2 different keyboards right out of the box. That is... Not a fantastic failure rate. OTOH, they are otherwise an unambiguously great value, with materials and build quality that feels pretty impressive for the cost, QC issues notwithstanding.
My only complaint is that the backlight behavior has like a dozen different settings, none of which are "off". There's "on", "flicker", "blink when you hit the key", "ripple out when you hit the key", "move in waves regardless of input", etc., but there's no "don't be light" option.
I bought a K4 Nordic version, only to learn none of their software supports it. So i can't reconfigure the firmware.
The keyboard has swapped Super/Alt (Windows/Alt) on the windows layout which is so annoying. I hate spending money on hardware so i'm living with it, but i curse the engineer that burned the prom every day.
That sucks, the main reason I like KeyChron is the little switch on the side of mine that toggles between Mac and Windows! I wouldn't buy another keyboard without this feature TBH.
Yeah on the mac it works fine. The keyboard just shipped with the Left Alt and Left Windows key in the correct positions, but they are mapped opposite in the firmware. Imagine how maddening it was in the beginning.
GP’s point is that many keyboards these days come with a Mac/Windows switch, and one of the things that switch does is swap the keycodes sent by the Alt and Win keys—the “Mac” behaviour is exactly what you’re describing. (The other thing it does is toggle FnLock for the F-keys.) So if I had a problem like yours with my keyboard I’d think that the Mac/Win switch were in the wrong position. If I’m reading page 9 of the manual[1] right, on the K4 the switch in question is right next to the USB-C port.
If you're not aware, you can remap modifier keys in System Settings on macOS. It doesn't have individual settings for left/right versions of a modifier key, so keep it in whichever mode has the modifiers on left/right paired with each other correctly, and then swap their functions to the right ones.
> "the little switch on the side of mine that toggles between Mac and Windows"
The physical slide switch on the Keychrons is a net negative for power users who use a KVM between Windows/Mac/Linux because it would wear out quickly and isn't replaceable. It's more sensible to have switching between layouts by keypress combination.
At least on the keychrons i've got, you can absolutely configure layout changes via key combination, and that slide switch is just an input like every button is an input.
I've got 2 that both run QMK firmware, so maybe the non-QMK ones can't do that, but at least some can.
Have been using one of their keyboards for two years switching between windows and macos and that key was not the one that wore out first.
Plus on most keyboards you could configure the layout switch with qmk anyways.
my issue with the one i have (k2, iirc) is that if i use it wirelessly and the battery dies, it becomes a brick until its been fully recharged. it won't even work as a wired keyboard until then. so ever since then i only use it wired.
Hmm, I have a mouse and trackballs from ploopy, they run QMK, so why was a new project needed? What is the added value?
To me, the missing feature in the QMK ecosystem is a devices too device communication channel. For now there is only caps lock & others: it would be nice to activate a mouse layer (for instance: increase DPI) to a keyboard key.
It is probably solvable if a heavyweight was to push for new HID descriptors or something similar.
I recently bought a Keychron TKL keyboard that uses Hall effect switches. It is, in my opinion, a lovely board to type on and so far I like it quite a lot. (Hopefully, this board with its socketed HE switches will serve longer than the previous board that had soldered Cherry MX Blues.)
It was advertised as being open-source and supporting QMK. Not at some time in the mysterious future, but right then. (And that's still true right now -- the web page is unchanged.)
But it does not have source available. The link for source code on the website lands at a Github page for a completely-unrelated keyboard, and I can't do anything at all with it using QMK. Extensive digging to find the source has resulted in zero positive results.
In the case of this Keychron keyboard, the promise of source code consists of resolute lies more than it does vaporware: The source is stated to openly exist, but it does not exist.
A question about this situation to their support email resulted in assignment of a ticket number, followed by silence.
Keychron is a real company with a good history so it isn't unreasonable to expect it and thus you shouldn't be too cynical. However don't confuse that with don't be cynical at all - many companies with a good history have canceled projects, and this is still vaporware.
Or it's (again) a brilliant idea from those marketing people, something along the lines of "let's announce something we don't have and see how it is received, so that we will see if there is interest or not".
It's not even made up, I have seen marketing people actually do that.
This would be more exciting if Keychron had mice of more interesting / experimental form factors and functionality. Just like their keyboards, their mice are mostly differentiated by obscure technical features (polling rates!).
I'm a bit worried that a decent, modern replacement for the extremely obsolete multi-button mouse I have now (Logitech G602) will never be available.
it uses this board: https://github.com/jfedor2/rp2040-pmw3360 and i can dig into my .stl files for the mouse if you want... made it on Blender and https://ryanis.cool/cosmos/beta - it's not the most lightweight thing in the world (would say around 100 grams) but it slides nicely and maybe by cycling/exercising often i don't think this one would give me RSI at any point in time... it's angle based on the most muscle neutral position the human arm anatomy has, between 20° - 30° [0] https://www.jospt.org/doi/pdf/10.2519/jospt.2004.34.10.638
I miss my G602 so much. Replaced it with a G604 and then replaced that one with another G604 and then a third G604 before realizing that Logi's build quality has seemingly dived off a cliff for the premium they charge for their product and swapped to the 50% as expensive Keychron mice and keyboards and have been satisfactorily living with the decrease in available buttons (though quite an adjustment was needed). Onset of carpal tunnel has seen me switch to a vertical mouse anyways, so a change was inevitable.
My G602 died due to Logitech's infamous double-click issue. :(
I've been fine with the G502 X Lightspeed (which should avoid said double-click issue due to optical switches) and some more efficient button use to compensate for the fewer side buttons.
I looked it up (because I had already replaced switches on previous mice) and the fix required carefully removing a rubber cover that is glued in place and tears if you look at it wrong. It also requires removing the skates at the base, which means buying new ones. Even if you do that, the switch is very hard to access. Easily a 60+ minutes job.
I bought the switch thinking it was a 2 minute job and it’s still sitting on a closet for when I have the patience to do it.
What's wrong with the G502? I like the wired G502 pretty well. Ideally I'd tweak the shape to be a little more like the M500s shape, but I'm happy enough with it overall.
I had a G502 that lasted years, but the last two I bought lasted me 6 months each (the wheel got completely gummed up somehow). I acknowledge that cat hair and sweaty hands are a bad combination for a mouse wheel, but that's insanely poor durability, and they're annoying enough to disassemble that they're effectively not cleanable.
It took me three hours to figure out how to blacklist my Keychron keyboard from showing up as a joystick - and screwing up several games that thought they were receiving constant joystick input - in Linux, so I'm not sure how keen I am on using their stuff in the future. Is that a common issue in Linux?
> This repository is currently in its early setup phase.
> At the moment, the repository mainly contains project scaffolding, policy files, and the initial project direction. Firmware sources, board support, build instructions, and flashing guidance are still being prepared.
> That means zgm is early, but it is intentionally early in public. The goal is not to wait until everything is finished before opening the door. The goal is to make the long-term direction visible and let the project grow into a useful open firmware platform in the open.
> That means zgm is early, but it is intentionally early in public.
"Intentionally early" is one of those phrases that means they vibed this README out in an hour with AI, and now hope that they can create firmware for it in the next 6 months.
Hopefully better than their QMK fork. They force push over their branch in their fork each new release effectively erasing history of the last version...
Of course there's no change log anywhere either...
I bought a Q6 wired keyboard. Most solid keyboard I've ever owned.
Only problem is the backlight didn't actually light up the keys.
It wasn't fixable. first, the keys that came with the keyboard were opaque. And after the arduous task of replacing the keys with shine-through keycaps -- all the LEDs were positioned below the keystem instead of above, so the lettering wasn't illuminated.
I would have loved to use that keyboard. Switched to steelseries apex pro.
I really wanted to like Keychron, but after spending 200$ on one of their high end keyboards, certain keys weren't detecting presses less than a month later. I couldn't end up fixing it so I tossed it out. Felt like a waste
Same. I ended up resoldering nearly all of the contact points (i.e. not the hot swap contact points but where those connect to the main board) because I liked the keyboard and firmware so much but it was such a stupid problem (way too little solder from the factory) for such an otherwise good and high end keyboard.
And all of us have owned multiple keyboards that have lasted for years of daily heavy use without a single key failing. Even the $30 membranes you could buy at $local_retailer didn't fail in a month.
I’ve had positive experiences with Keychron products in the past, but I’ve moved past them as I now prefer orthogonal layouts. That said, I loved their support of QMK on my Q1 Pro and I really hope this ends up on that same level.
Hmm that opens up a way for cheating in games - have a Screengraber and image processing running directly communicate with mouse and let it calculate the distance to region of interest internally and then give out delta position. For the system, it looks like the mouse has been moved very fast and far
There's a ton of low-latency proxy hardware meant for cheating that does exactly that, out of the box, for those too lazy to do it themselves on any devboard. It's not an excuse for closed source mouse firmware
It's already being done. The fanciest game hacks use a custom PCI card to DMA game memory, read it on another computer, where it can do wallhacks and such.
But also, the second computer can drive a virtual mouse and keyboard which is connected to the 'legit' PC to even aim/move/shoot for you.
I stumbled upon this brand when I was looking for a hall effect QMK keyboard.
Glad to see them introducing this - I might make a purchase as while my current mouse is great, I can't customise it because the associated bloatware runs only on Windows.
Very interesting. I've been eyeballing a mouse/trackball project for a while based on some of the open source designs like Ploopy and the like. Slap a Nordic chip on it and a joycon battery. Even better if you can cook up a wireless charging mousepad like Logitech offers!
Never have I thought "I want a mouse with modifiable firmware", IMHO this is only going to lead to more bugs arising from the associated increase in complexity and the "we can always fix it later" mentality. All the mouses and keyboards I've used have a mask-programmed MCU, and essentially no bugs to speak of.
I think the point is less to add features and more to lower the barrier to entry for people making mice, since, as you said, in large part keyboards and mice are appliances to the degree that it should rapidly converge to a stable set of features.
Plus, on a human level, I wouldn't be surprised if the teams working on mouse development were annoyed they couldn't show their past work in the same way...
I suspect that a lot of hardware will start shipping open source firmware soon. Routers, IP cameras, etc.
It'll be a selling point for everyone tinkering with vibe coding, and it's no longer a moat.
Companies that don't do this will have their firmwares dumped with open source models and tools. Companies that do this will have community goodwill and extra distribution that comes from having people talking about your product.
That might happen in niche enthusiast verticals like PC gaming. It's not going to happen in the broader market because there is no broader interest in it.
Most people want open source firmware for their router like they want open source sewing patterns for their pants. Primarily these are consumer products where people buy the product as a way to intentionally avoid spending time dealing with the problem they solve.
Open source pants might sell well in a Hobby Lobby and an open source IP camera might sell well in a Micro Center, but neither would sell well enough to get shelf space in a Walmart... and the best selling one in Walmart sells a tiny fraction compared to whoever can land an Xfinity contract.
Of course they are, as are open-source IP cameras and routers. My point is not that they don't exist, my point is that these things have interest only within relatively tiny bubbles.
Do you have any cameras that support openipc you’d recommend?
I need surveillance cameras (outdoor mostly) but the cheap ones I found are a horrid proprietary mess. I even bought and returned one after the setup was complex and invasive.
Basically all I need is access to the live stream from a raspberry pi where I can store and process the video.
> Do you have any cameras that support openipc you’d recommend?
I don't, unfortunately. My suggestion would be to use cameras that stream over RTSP, an open protocol, so you don't need to use any proprietary tools/apps. Make sure the cameras are air-gapped.
thats not the blocker you think it is. they can "leak" the info, wait until broadcom lawyers notice, agree to investigate themselves and fire a random contractor (who gets paid on the side). they just need to do it outside of america to avoid discovery if anyone decides to sue them.
And, in an era where vendor software and firmware for all kinds of devices is constantly having vulnerabilities, it is possible the liability of doing custom shit will eventually outweigh whatever benefits hardware companies seem to think it has, versus just having industry standards and free software instead. But, maybe that is wishful thinking.
My only complaint with Keychron is that of 5 or so keyboards I've owned from them I wound up having to resolder 3 different hotswap sockets on 2 different keyboards right out of the box. That is... Not a fantastic failure rate. OTOH, they are otherwise an unambiguously great value, with materials and build quality that feels pretty impressive for the cost, QC issues notwithstanding.
The keyboard has swapped Super/Alt (Windows/Alt) on the windows layout which is so annoying. I hate spending money on hardware so i'm living with it, but i curse the engineer that burned the prom every day.
[1] https://www.keychron.com/pages/k4-keyboard-user-manual
https://www.keychron.com/pages/k4-user-manual
--
Edit and if i switch to the macOS mode. The right alt-gr key is not mapped correctly.
https://i.imgur.com/FXkmZRV.png
Settings are per keyboard so fixing your external keyboard won't mix up any other keyboards you use (particularly for laptops).
I use my Keychron C3 in Windows mode because the macOS mode ain't right with the bottom left corner keys.
Maybe placement of Control/Option/Command aren't similar enough to the MacBook Pro keyboard for me.
Been meaning to figure this out and fix it... swapping physical keycaps or remapping in software or replacing the keyboard...
Hope your next buy's perfect.
The physical slide switch on the Keychrons is a net negative for power users who use a KVM between Windows/Mac/Linux because it would wear out quickly and isn't replaceable. It's more sensible to have switching between layouts by keypress combination.
I've got 2 that both run QMK firmware, so maybe the non-QMK ones can't do that, but at least some can.
To me, the missing feature in the QMK ecosystem is a devices too device communication channel. For now there is only caps lock & others: it would be nice to activate a mouse layer (for instance: increase DPI) to a keyboard key.
It is probably solvable if a heavyweight was to push for new HID descriptors or something similar.
They claim 2027 Q1 will be the release date.
Links below purporting to be “the repository” lead to a repo with no source code.
Sorry, I’m cynical. An announcement 6-9 months ahead of release? Why? Right now it’s vaporware.
I recently bought a Keychron TKL keyboard that uses Hall effect switches. It is, in my opinion, a lovely board to type on and so far I like it quite a lot. (Hopefully, this board with its socketed HE switches will serve longer than the previous board that had soldered Cherry MX Blues.)
It was advertised as being open-source and supporting QMK. Not at some time in the mysterious future, but right then. (And that's still true right now -- the web page is unchanged.)
But it does not have source available. The link for source code on the website lands at a Github page for a completely-unrelated keyboard, and I can't do anything at all with it using QMK. Extensive digging to find the source has resulted in zero positive results.
In the case of this Keychron keyboard, the promise of source code consists of resolute lies more than it does vaporware: The source is stated to openly exist, but it does not exist.
A question about this situation to their support email resulted in assignment of a ticket number, followed by silence.
So yeah, anyway: The cynicism is valid.
It's not even made up, I have seen marketing people actually do that.
I'm a bit worried that a decent, modern replacement for the extremely obsolete multi-button mouse I have now (Logitech G602) will never be available.
it uses this board: https://github.com/jfedor2/rp2040-pmw3360 and i can dig into my .stl files for the mouse if you want... made it on Blender and https://ryanis.cool/cosmos/beta - it's not the most lightweight thing in the world (would say around 100 grams) but it slides nicely and maybe by cycling/exercising often i don't think this one would give me RSI at any point in time... it's angle based on the most muscle neutral position the human arm anatomy has, between 20° - 30° [0] https://www.jospt.org/doi/pdf/10.2519/jospt.2004.34.10.638
I've been fine with the G502 X Lightspeed (which should avoid said double-click issue due to optical switches) and some more efficient button use to compensate for the fewer side buttons.
Isn't that just the tact switch failing? Two minute job to replace...
I looked it up (because I had already replaced switches on previous mice) and the fix required carefully removing a rubber cover that is glued in place and tears if you look at it wrong. It also requires removing the skates at the base, which means buying new ones. Even if you do that, the switch is very hard to access. Easily a 60+ minutes job.
I bought the switch thinking it was a 2 minute job and it’s still sitting on a closet for when I have the patience to do it.
What's wrong with the G502? I like the wired G502 pretty well. Ideally I'd tweak the shape to be a little more like the M500s shape, but I'm happy enough with it overall.
> This repository is a lightweight analysis and report-generation workspace.
And "please contribute to our blank open-source project" can be taken with a grain of salt.
> This repository is currently in its early setup phase.
> At the moment, the repository mainly contains project scaffolding, policy files, and the initial project direction. Firmware sources, board support, build instructions, and flashing guidance are still being prepared.
> That means zgm is early, but it is intentionally early in public. The goal is not to wait until everything is finished before opening the door. The goal is to make the long-term direction visible and let the project grow into a useful open firmware platform in the open.
"Intentionally early" is one of those phrases that means they vibed this README out in an hour with AI, and now hope that they can create firmware for it in the next 6 months.
Of course there's no change log anywhere either...
Does that mean they remove any release tags as well?
Only problem is the backlight didn't actually light up the keys.
It wasn't fixable. first, the keys that came with the keyboard were opaque. And after the arduous task of replacing the keys with shine-through keycaps -- all the LEDs were positioned below the keystem instead of above, so the lettering wasn't illuminated.
I would have loved to use that keyboard. Switched to steelseries apex pro.
I'm on a Mac keyboard bent for the very reason, actually,
Mouse firmware sounds very cool, though. If you could access a layer by holding down a central button, that could open up neat possibilities.
Which one do you use/have now?
and yes, hdmi/displayport capture -> ML object detection -> fake HID mouse to cheat is absolutely already a thing.
But also, the second computer can drive a virtual mouse and keyboard which is connected to the 'legit' PC to even aim/move/shoot for you.
Glad to see them introducing this - I might make a purchase as while my current mouse is great, I can't customise it because the associated bloatware runs only on Windows.
Plus, on a human level, I wouldn't be surprised if the teams working on mouse development were annoyed they couldn't show their past work in the same way...
It'll be a selling point for everyone tinkering with vibe coding, and it's no longer a moat.
Companies that don't do this will have their firmwares dumped with open source models and tools. Companies that do this will have community goodwill and extra distribution that comes from having people talking about your product.
Most people want open source firmware for their router like they want open source sewing patterns for their pants. Primarily these are consumer products where people buy the product as a way to intentionally avoid spending time dealing with the problem they solve.
Open source pants might sell well in a Hobby Lobby and an open source IP camera might sell well in a Micro Center, but neither would sell well enough to get shelf space in a Walmart... and the best selling one in Walmart sells a tiny fraction compared to whoever can land an Xfinity contract.
Those already exist: OpenWRT, OpenIPC. I wish more hardware was available though, especially for cameras.
I need surveillance cameras (outdoor mostly) but the cheap ones I found are a horrid proprietary mess. I even bought and returned one after the setup was complex and invasive.
Basically all I need is access to the live stream from a raspberry pi where I can store and process the video.
I don't, unfortunately. My suggestion would be to use cameras that stream over RTSP, an open protocol, so you don't need to use any proprietary tools/apps. Make sure the cameras are air-gapped.
This is not as "one and done" as you make it sound.