[Issue]: Sony Bluetooth keyboard with PlayStation keys shows up on the host as a second DualSense
Nobody has claimed this yet.
Assessment
- Difficulty
- 4/5
- Estimated time
- 3-5 days
- Newbie friendliness
- 48/100
- Issue type
- Bug
- Clarity
- Mostly clear
- Activity status
- Active
- Tech stack
- android, java
- Domain
- backend-api-design, mobile
Research direction
Start by reading ControllerHandler.sendControllerArrival() and tracing how Android input devices are classified and mapped to controller buttons. Reproduce with a Sony N1158 keyboard and a DualSense paired to a Shield, then check whether the keyboard is announced as a controller and whether its PlayStation keys map as described. Done means the keyboard no longer creates an unwanted controller, or its buttons map correctly if it is supported as one.
Written by the indexing model from the issue text.
Description
Describe the bug
I have a Sony wireless keyboard (model N1158) paired to my Shield. Besides the normal keys it has the PlayStation face buttons and a d-pad printed on some keys. The first time I press one of those keys during a stream, Moonlight announces a new controller of type PS with capabilities 0x00. The host creates a virtual DualSense for it, so Steam lists two DualSense controllers with the same artwork and there is no way to tell which one is the real pad. I only use the keyboard as a keyboard, so this extra pad just causes confusion.
Most of the PlayStation keys on that pad also arrive as a different button. This is what Steam showed for each key.
| Key on the keyboard | Button seen on the host |
|---|---|
| Triangle | Square |
| Square | Cross |
| Cross | Circle |
| Circle | Cross |
| L1 | Triangle |
| L2 | L1 |
| R1 | Cross |
| R2 | R1 |
| L3 | Create (Select) |
| R3 | Options (Start) |
| Select | left trigger |
| Start | right trigger |
The keyboard's normal arrow keys do nothing on the pad, while its d-pad keys move it.
As far as I can tell from ControllerHandler.sendControllerArrival(), the controller type is picked from the vendor ID alone, so any Sony device that Android reports with gamepad buttons becomes a PS controller. The capabilities are 0x00 because the keyboard has no trigger axes.
It would help if Moonlight either ignored devices like this one, which are keyboards first and have no analog sticks, or offered a way to exclude a device from being sent as a controller. If they are meant to work as controllers, the face buttons should map to the buttons printed on the keys.
Steps to reproduce
- Pair a Sony wireless keyboard with PlayStation keys (model N1158) and a real controller (I use a DualSense) to the Shield.
- Start a stream with Steam Big Picture.
- Press one of the keyboard's PlayStation face-button or d-pad keys.
- Open Steam's controller settings. A second DualSense is listed, and it reacts only to the keyboard.
Affected games
Steam Big Picture. The extra controller is created on the host as soon as the key is pressed, so it is not specific to a game.
Other Moonlight clients
Not tested. I only have the Shield as a client.
Moonlight adjusted settings
Yes
Moonlight adjusted settings (please complete the following information)
Mostly video and audio settings. The built-in USB gamepad driver is off. The keyboard and the DualSense are both connected over Bluetooth.
Moonlight default settings
Not tested.
Gamepad-related connection issue
No. No gamepads are connected to the host directly.
Gamepad-related input issue
Not applicable. The real gamepad works, the problem is the extra controller.
Gamepad-related streaming issue
Not tested with html5gamepad.com, but the host log shows the extra controller arriving from the client (below).
Android version
Android 11
Device model
NVIDIA Shield TV Pro
Server PC OS version
Debian 13, streaming host Wolf (games-on-whales/wolf) in Docker
Server PC GeForce Experience version
Not applicable, the host is Wolf on Linux.
Server PC Nvidia GPU driver version
595.91.07
Server PC antivirus and firewall software
None
Relevant log output
Host log at the moment the extra controller appeared. Controller 0 is the real DualSense and controller 1 an Xbox pad.
16:29:52 [INPUT] CONTROLLER_ARRIVAL controller 0 type 2 capabilities 0x01 button flags 0x00f7ff00 (held pads: [])
16:31:24 [INPUT] CONTROLLER_ARRIVAL controller 1 type 1 capabilities 0x03 button flags 0x00f7ff00 (held pads: [0])
16:40:18 [INPUT] CONTROLLER_ARRIVAL controller 2 type 2 capabilities 0x00 button flags 0x00f7ff00 (held pads: [0, 1])
16:40:18 [INPUT] active_gamepad_mask changed 0x0003 -> 0x0007 (packet for controller 2, held pads: [0, 1, 2])
To be sure it is the keyboard, I counted the bytes read from each virtual pad's input device on the host. With the real DualSense switched off and only the keyboard's PlayStation keys pressed, only the pad created for controller 2 received input. While I used only the real DualSense, controller 2 received nothing.
Additional context
Moonlight for Android v12.2. The host is Wolf, which creates a virtual DualSense for every PS controller the client announces, so the extra controller is fully visible to games.
- Dominant language
- C
- Stars
- 7.5k
- Forks
- 1.8k
- PR merge metrics
- No merged PRs in 30d
Getting set up
This project ships no dev container, Dockerfile or contributing guide, so setting up is up to you: start from its README, and see our first-contribution guide for the general steps.
First steps
- Read the whole issue, then the project's contributing guide.
- Comment on the issue to say you are picking it up — it saves two people doing the same work.
- Fork the repository and make your change on a branch.
- Open a pull request that references the issue number.
More from moonlight-stream/moonlight-android
-
Difficulty 2/5 1-3 hours Newbie friendliness 82/100
-
bug
Difficulty 2/5 1-3 hours Newbie friendliness 82/100
-
bug
Difficulty 2/5 1-3 hours Newbie friendliness 74/100
moonlight-stream/moonlight-android#1612 · 2 comments ·
-
[Issue]: HDMI-CEC OK/Select button does not work on Google TVPossibly taken @mrsric claimed this 29 days ago. Openbug
Difficulty 2/5 1-3 hours Newbie friendliness 82/100
moonlight-stream/moonlight-android#1603 · 2 comments ·
-
bug
Difficulty 2/5 1-3 hours Newbie friendliness 85/100
All issues in moonlight-stream/moonlight-android
Similar issues
-
Difficulty 2/5 1-3 hours Newbie friendliness 68/100
arduino/ArduinoCore-mbed#1126 ·
-
Difficulty 1/5 Under an hour Newbie friendliness 88/100
-
security
Difficulty 1/5 Under an hour Newbie friendliness 93/100
btclib-org/btclib-secp256k1#1427 ·
Maintainers usually reply within 1 day
-
Difficulty 2/5 1-3 hours Newbie friendliness 65/100
esmuellert/codediff.nvim#594 ·
-
Difficulty 1/5 1-3 hours Newbie friendliness 67/100