For instance it could have a separate cover just for closing the lid. Like mechanical keyboards do but this one would be for closing.
It could be 3d-printed or something.
This one is for keyboard itself but I expect it to cover all of the bottom chassis so that the pressure would be dispersed to the frame of the closed screen:
Like this (this is really sloppy sketch but I don’t care):
That looks ever so comfy to type on. Sorry, couldn’t resist. 
Might I suggest you attach it to the bezel instead.
Not attach it permanently.
You’d open the laptop and remove it. So you’d be carrying the laptop with it but open the laptop and remove this green part.
Maybe it could work idk.
Also it could be attached to bezel. It would work the same.
All good and well, but I still feel it a missed opportunity not to accommodate discrete laptop switches from the get-go.
As you say, altering the lid with a deep bezel to recess the screen may do the trick. The frustrating thing is this only needs to be very few mm from the standard design, you are going to the trouble of commissioning an aftermarket part for something that should have been standard.
Probably a dumb idea, but with enough screen protection (e.g. with a keyboard cover as mentioned here: Input Module Keyboard Switches - #21 by Martynas), could the hinge’s force allow the laptop to be closed while depressing Cherry ULP keys (3.5 mm tall, up to 1.8 mm travel)?
The actuation curve of the switch (MX ULP TACTILE - MX ULTRA LOW PROFILE - MX SWITCHES - CherryMX) suggests that at 1.5 mm depression, the force per key is 30 cN, so even with 50 keys, the total force would be 15 N. I have no idea what the hinge’s force is as a function of angle, or even what the right order of magnitude is there. Could that be altered by e.g. magnetic strips on the keyboard edge that attract the bezels to shut the laptop stronger?
Do laptops have to close, anyway? 
Ideally, the single-key PCBs mentioned by @nrp would be great, but if those don’t work out, it would be nice to be able to use off-the-shelf parts to make a split-ish ortho input module.
What I would love would be something that would adapt a T-class Thinkpad keyboard and track pad to the framework 16.
Given that they are plentiful and inexpensive in the aftermarket this doesn’t seem entirely impossible. Might have to do a bit of soldering to replace the existing flex cables with a new design that handles the signaling and connects to the right place, and I haven’t measured whether there is enough space, but I can dream.
I wonder if any of the aftermarket clone vendors would do a run of a thinkpad-inspired keyboard and track pad for the 16?
A bunch of people are working on thinkpad-based keyboards over in Thinkpad keyboard mod (super early stages)! – some of the discussion has touched on the feasibility of expanding the hinges somewhat to allow space for slightly larger keyboards.
I wouldn’t mind a kludge. Because I DO want a thin laptop. The kludge would go like this:
In place of the Framework keyboard, there is a flat module. This module fits into the same space (is perhaps thinner), but is essentially an adapter. When you sit down to use your laptop, you open it up and plunk your fancy ortholinear or-what-have-you keyboard on top of this adapter, which contains the electrical connections to emulate the Framework keyboard. . And you use your custom keyboard instead of the Framework provided one.
Your keyboard would be carried separately.
Sorry, but that’s the best I got…
Which will work great until you need to run for the bus or something and slam the lid down on yer thick-ass keyboard. Crunch!
Being able to have a custom keyboard layout is what I’ve been dreaming off since seeing the first Framework announcement.
Has there been any news regarding fully custom keyboard input modules, ideally with low profile mecanical switches? What is missing for it to work? Is it just a vertical clearance issue?
If so the Kailh PG1316M could fit:
The PG1424 series could also be short enough:
Knowing that and the fact that the connection is a simple USB interface that’s simply laid flat, how hard is it to make a fully custom keyboard input module?
The math on those switch heights doesn’t quite work out. The total “hard” thickness available for an input module is only 3.1mm (aka the “key down” thickness), with a soft/pressable (“key up”) thickness limit of 3.7mm. The PG1316 switches are 5mm tall (including keycap and 0.8mm PCB, according to the write-up on the mikefive keyboard which uses them) so you’re ~1.3mm over the available soft space. In the keydown position, they are 3.1mm so technically compliant, but the LCD is going to be pressing all the keys down by about ~1.3mm when the lid is closed, which might leave a mark.
The PG1425 are similar, with a 6.4mm “key up” height and 4.40mm “key down” height (the measurements in the datasheet are a little misleading as they don’t include the 0.5mm switch legs).
That said, I think you could add a ~1.5mm shim to the laptop hinge without /too/ much difficulty, and I’ve got some parts I’m waiting to get back from being printed to validate that. If shimming is reasonable then a lot of options open up. But any keyboard build is going to have to be super height-conscious.
Thanks a lot for the detailed response!
As I can now see 3.7mm is really, really short indeed. My low profile keyboard wich is “short” by most standards is indeed way over 1cm (1.6mm hotswap sockets + 1.6mm PCB + Kailh v1s + keycaps).
And so, so far, there has not been a single fully custom Framework keyboard, right? Not even with more traditional silicone designs?
Holy…
You guys are fabulous. I know you said you were trying to see if single key modules were possible. But I feared it would have to be shelved, just have too many other more important things to deal with. Great to see it’s progressing!
Yeah, this is amazing. I’m very excited to learn more about these modules and how they will work. Thanks for the update!
Here is the idea, of how you’d you them
You’d attach them to a plate, similar looking to this, which is used with custom mechanical keyboards. Except, it would need to provide the key matrix traces in addition to providing structure.
Ah great, that looks achievable. Is the MCU going to be on the PCB lattice then too? Will that be something Framework can provide, or maybe at least provide some guidance for those that are entry-level hardware designers?
Since it’s meant for custom key arrangements, I don’t think Framework would be making the lattice plates. It’s possible they might do one. I think they would likely create a pcb example file at least. And guidelines for material choice to ensure it’s strong and rigid enough. Maybe polyimide with stainless steel stiffener.
I’m sure a lot of keyboard fans will jump on this & share their designs. Maybe someone will share an ordering step-by-step for those who aren’t experienced with custom pcb design and sending them to be made. Take these files, upload to jlcpcb, select these options, etc.
All assuming that Framework’s single-key module makes it to production. Everthing is in a really limited space, and something like this has never been done before.
There’s already a microcontroller sample file in the InputModules git repo, under “MicrocontrollerInputModule”. It uses an RP2040. You’ll probably need an I/O expander, though. I’ve got some designs open sourced already for my X1 Carbon Gen 10 keyboard that can be generalized. Maybe make a blank PCB with the right shape for an input module and the microcontroller (and I/O expander?) in the corner, then folks would just need to “add their keys”.
