[Bug]: Sensitivity slider value is not wired to the trackpad, setting is ignored and double-applied

[Rozerxshashank]

Bug Description

Related to #42 which requested the sensitivity slider. The slider now exists in Settings, but it's not correctly connected to the trackpad. There are two problems:

Problem 1: The trackpad ignores the user's sensitivity setting.
The trackpad page never passes the configured sensitivity to the gesture hook. Instead, the hook falls back to a hardcoded default of 1.5, so no matter what the user sets in Settings, the client always uses 1.5.

Problem 2: Sensitivity is applied twice once on the client, once on the server.
The client multiplies all touch movement by its hardcoded 1.5 before sending it to the server. The server then multiplies the same values again by the configured sensitivity from server-config.json. So if the user sets sensitivity to 2.0, the actual result is 1.5 × 2.0 = 3.0 not what the user intended.

Proposed Solution:
Apply sensitivity in one place only the server since it already reads the saved setting.

Remove the hardcoded 1.5 multiplier from the client-side gesture hook and let the server handle scaling. This way the Settings slider directly controls cursor speed with no double-application.

Steps to Reproduce

1. Start the server with npm run dev
2. Open Settings and set Mouse Sensitivity to 0.5x, then click Save Config
3. Connect from a mobile device and use the trackpad
4. Notice: the cursor still moves faster than expected (0.75x instead of 0.5x)
5. Now set sensitivity to 3.0x and save — the cursor moves at 4.5x instead of 3.0x

Expected Behavior

The sensitivity slider should directly control cursor speed. Setting it to 0.5x should give half speed, and 3.0x should give triple speed without any hidden multiplier inflating the value.

Desktop Operating System

Windows 11

Mobile Device (Client)

Samsung Galaxy A56

Rein Version

1.0.0

Console Logs / Error Messages

No errors. This is a silent logic bug.

Screenshots / Recordings

No response

Checklist

• I have searched for existing issues to avoid duplicates

[Rozerxshashank]

Hey @imxade, I want to work on this issue. Can you assign it to me?

[Nakshatra480]

@imxade pls assign this issue to me

[imxade]

@Rozerxshashank go on and keep the discussions in discord

[Nakshatra480]

@imxade I think we should actually move the sensitivity logic to the client instead of the server. If we handle it on the frontend using window.devicePixelRatio , the cursor will feel consistent across different phones and updates will be instant without needing a server reload. It simplifies the backend a lot since it just becomes a "dumb" driver that executes whatever coordinates the client sends. Can i raise a PR with this approach?

[Nakshatra480]

Hi @Rozerxshashank @imxade I have reviewed the changes in trackpad.tsx and identified a few items that need to be addressed before merging:

• The event handlers (specifically handleClick , sendKey , and handleToggleKeyboard ) are currently being redefined on every render. Please wrap these in useCallback to maintain referential stability and prevent unnecessary re-renders in the child components.
• We should replace the inline style props for grid positioning with standard Tailwind utility classes (e.g., col-span-full , col-span-3 ). This ensures our styling approach remains consistent and easier to maintain.
• The logic at line 44, which automatically appends a space to any input longer than one character, seems problematic for general use cases (such as entering passwords or specific commands). Unless this is a strict requirement, I recommend removing it to prevent unexpected behavior.
• Please extract the hardcoded timeout values ( 50ms , 10ms ) into named constants at the top of the file to improve readability and configuration."

[Rozerxshashank]

Thanks @Nakshatra480 for the suggestions but the changes I did were only in useTrackpadGesture.ts and InputHandle.ts . So your first point of the "reviewed changes in trackpad.tsx" is irrelevant and seems you are using LLM without reviewing yourself as a human.

But your points are good and as @imxade said you to review my PR. I will address the things that you mentioned.

[Nakshatra480]

Hey @Rozerxshashank, you are doing good implementation about the issue but i think you are using AI based coding LLMs for code generation bcz you can't even know what files of code you make PR on.

I know writing code by LLM is easy and efficient but i don't prefer it for long term contribution. Considering you cannot do it anymore, but the dedication you showed by opening this issue is genuine.

[Rozerxshashank]

Sorry to inform you @Nakshatra480, but I updated this file after your comment that that I am attaching. Please see the commit history before raising any allegations. Because raising these types of allegations on a contributor waste the contributor's as well as thew maintainer's time. @imxade Kindly look into this situation as this guy is just finding ways to disrepect me instead of actually contributing.

The commit which showed the change after your comment suggestion -

[Nakshatra480]

Sorry @Rozerxshashank i misunderstood the thing that you try to explain, my mistake

[Rozerxshashank]

Sure no worries! I hope you may learn something from this.