Those switches are software under the hood though, at least as far as I can tell. I installed freebsd on my purism laptop, and it no longer shuts off the camera/microphone.

Really? The marketing says they are hardware switches: https://puri.sm/products/librem-13/

> Two hardware kill switches, microphone/camera and wireless/bluetooth

> Now with a physical toggle switch, when your camera and microphone are switched off, you know they are off. Wireless and Bluetooth are combined in a second hardware switch to control all your radio signals inbound and outbound.

That would be a flat out lie if it wasn't actually a hardware switch which is hard to believe for a company whose entire identity and reputation is tied to these exact features.

I don't have video/audio working on this install, so I'll try to get that set up later and make a video.

I do have a video of the wifi in action though: https://tmp.thekyel.com/month/wifi.mp4

I reached out to them a couple of times, and they asked if I had rfkill installed, I told them no, but it shouldn't matter if it was a hardware switch.

They told me that they would look into it, and then never responded after that.

Interesting indeed! Here [1] is how it allegedly works. Is the laptop easy enough to open and check if the board you have looks anything like those photos?

On the other hand, putting a signal onto disable pins is not exactly a kill switch. The description sounds like a hardware switch in front of a soft kill, which would make it some sort of quasi-HKS. A real kill switch would need to totally cut power to that controller rather than rely on it playing nice.

On the gripping hand, I find it hard to believe they would overlook something that crucial. Eg, if it didn't work, it would /have/ to show up during testing.

I wonder if the chip needs to be tightly integrated into the board and a proper HKS would require cutting signal to dozens of those pins -- something that is perhaps impractical -- so they went for a lesser option.

FWIW, there is at least one more person alleging the same broken functionality [2]

[1] https://puri.sm/posts/hard-not-soft-kill-switches/

[2] https://forums.puri.sm/t/wifi-killswitch-didnt-work-when-i-n...

Maybe they mean that the switches themselves are hardware, but the deactivation is implemented in software, only triggered by the (harware) switch.

That's a major accusation. Can you do a video or blog post where you show this happening? Eg, start recording something and hit the switch and show it continuing to record.

Switches on one device don't necessarily indicate how it's done on a different device.

As far as I remember Purism has always claimed that it's hardware switches on the phone, and I don't see why they'd lie because it would be very easy to verify.

I've never installed purism's version of linux, and they seem to work - I recall the camera/microphone disappear from the usb bus, shown in syslog. I don't remember what happened with the wifi/bt as I leave them off.

I could go check this behavior.

I haven't tried freebsd - could it be freebsd doesn't check for a hotplug?