Keep an eye on them, they drop new stuff semi-regularly. It’s all handmade so not constantly fully stocked.

https://www.asymplex.xyz/

so i had a crack at fixing it and got it to compile, obviously unable to test it functionally. I've pushed the code here.

I noticed you're not actually using the joystick features since you're reading the analog pins manually (btw, you should read them once per scan instead of in every if statement). so i just removed all the joystick code and the JOYSTICK_ENABLE = yes. i'm not sure what exactly you're trying to do but what I've got builds and doesn't remove any of the functional code, so see how you go from there. I don't know much about using the joystick feature in QMK so I can't really advise on how to use it instead of what you're doing, but as it stands you seem to not be utilising it anyway.

Good luck. if you're stuck, maybe try going to the QMK discord and explaining what you are trying to achieve and see what people say there. Be careful not to assume the implementation and ask the wrong question, describe what you want to achieve at a high level and see what they advise to implement it (see the XY problem)

i don't think you need to include config_common.h, that's just a convention for when there are multiple revisions of a keyboard but most of the stuff is common.

as for the other code, try removing things until you get it to compile and start adding it back. it looks like the problem could be happening inside some macro expansion or something. i think it would help to format your code a bit to make it easier to follow, there's not really a need for the switch statement with only one case, etc. use consistent indentation, and such.

your compilation errors now just seem to be because of mismatched braces. you have an extra closing brace in keymap.c before the xPos = analogReadPin(B4); as well as after the intended close of the matrix_scan_user method.

For the joystick part, do you have JOYSTICK_ENABLE = yes in your rules.mk?

For this you need an analogue sensing mechanism such as hall-effect sensors (as Wooting use). You can see some open-source reference designs for PCBs, Firmware, and switches using hall-effect sensors on riskable's github to get you started. Technically you should be able to make a PCB like this and then use the same "switches" that Wooting use, though idk where they source them.

I'm not sure but I don't think optical switches have that ability either, you're either breaking the beam or you aren't.

Wooting uses hall-effect sensors and MX-compatible "switches" that have a magnet in the plunger. The set and reset thresholds can be set in software/firmware.

So it can be done with hall-effect switches but not with mechanical switches as you say.

Ergos can be actually relatively cheap. That's why I've spent god-knows how much and have like 30 of them.

I'm pretty sure that's a claw44

Does your keyboard have asymmetric halves? The left having 7 columns and the right having 6? As the other commenter pointed out you have defined 6 columns but your layout macro in eiris.h has 7 elements for the left hand. I think the way to solve it is to define it as 7 columns, add NO_PIN to the end of MATRIX_COL_PINS_RIGHT and then just add XXX to the end of all the rows for the right half in your layout macro in eiris.h.

This is outlined in the documentation

Even with AI models that can identify that there are birds in the picture. Having it decide with accuracy that the picture is of a bird is still a hard problem.

Ideally you want to use QMK’s tri_layer feature as I outlined in this comment.

Also you can just use the _______ keycode instead of restating MO(...) on every layer as that is the an alias for KC_TRANSPARENT meaning it maintains the function of the layer below.

This doesn’t work perfectly, IIRC if OP presses MO(1), then MO(2), then releases MO(1) it’ll stay on layer 3. The better way to do it is defining the layer_state_set_user function and calling update_tri_layer_state as documented here

I know this doesn’t answer the question but I want to offer some advice instead.

In my opinion just don’t. If the company want you to have access to emails on the go then they should give you a company phone. If they don’t, why are you trying to? Don’t put work things on your personal phone.

This is the correct answer

/thread

Yeah, it does

The main difference that has me using LibreTube rather than NewPipe is because my subscriptions are on my piped account so they are synced between phone and desktop (browser). Piped is built on top of NewPipe’s extractor library anyway.

You just use three backticks to start and end a code block, it’s just markdown.

e.g.

version: '3.4' 
  
services:
   vaultwarden: 
     image: vaultwarden/server:latest 
     restart: always 
     # environment: 
     #   SIGNUPS_ALLOWED: 'false' 
     #   ADMIN_TOKEN: 'your authentication token' 
     ports: 
       - '127.0.0.1:8200:80' 
     volumes: 
       - vaultwarden-data:/data/ 
...

To speed up your PCB creation this board would be really quite easy to make in ergogen and that can spit out a PCB for you and all you’ll need to do is route it. Either way, it shouldn’t be too hard to make anyway.

One tip if you make it in kicad yourself from scratch. Set your grid size to a fraction of 19.05 (MX Spacing), start by laying out one half of the switches and then rotate them as a block, then do the same for the other half. that’ll make it much easier to work with than trying to place everything on an angle.

Good luck.