Hacker News Reader

Comment by jancsika | Hacker News Reader

I wish they had gone with a single kill switch. The most common use case is probably "Nothing gets in or out right now." But instead of flipping one switch one direction, the user must now flip three switches the same direction to get that common behavior.

Also, what's the use case for switches flipped in different directions that wouldn't be covered by software buttons? We can't possibly be talking about targeted attacks-- in those cases you're screwed no matter what.

That only leaves indiscriminate software attacks. Now take the hardware switch setting where baseband=off and wifi=on. I guess it's nice that evil software cannot pretend that baseband is off when it's actually on. But evil software attempting to "get out" is probably going to use a high level call out to the net. So who cares that it must now leave through the wifi radio instead of the baseband radio?

View on HN

I'm happy about multiple switches. I don't really understand how cell basebands work and if I can trust a software kill. My biggest issue with my phone is cell phone companies selling real time location data, so I will mainly have that switch off while using wifi for communication. If the software kill means the baseband transmits no data, great, I'd be fine with just that. If it means the basebands transmits no data from the phone, but still beacons out to cell towers, I absolutely will want to kill it while remaining on wifi.

I also expect to have the baseband switched off most of the time and wifi on most of the time.

I basically want a pocket computer. Access to the cell network is a useful feature, but one I don't need most of the time, so why be connected to it all the time, with all the privacy concerns that comes with.

It will be interesting to see how carriers react to the option, though. As it stands, they are happy to help themselves to your location and usage data (much like ISPs do), having a handset that could turn that money-making flow on and off might cause at least some to refuse to take you on as a customer.

You cannot do that in the EU. Also, phones communicate imei and other sim information, not OS information on the header plane during handshakes with the towers.

What kind of crazy country do you live in where the carriers can choose whether or not to have you as a customer based on the device you use?

There's plenty that are "bring your own device". From a carriers standpoint, switching off the basband is the same as turning it off, so I would rralpy doubt they could figure out the difference, much less do anything about it.

Like an iPod Touch? https://www.apple.com/ipod-touch/

Despite the name, it's basically just an iPhone without a cellular radio (though I'm not sure if it's actually removed, or just disabled/neutralized). So you can use it to do anything an iPhone can do, it's just dependent on having access to a WiFi connection if you want to do anything over the network.

Not sure why you're downvoted. The iPod Touch is a great solution and I've bought one before. I don't love the Apple (walled garden) ecosystem though, which is another reason I'm excited about the Librem 5.

I have and enjoy my iPod Touch for a whole host of reasons, particularly its thinness when compared to a iPhone.

That said, sometimes you need/want cellular in a pinch, but don’t want it always on, hence the kill switch.

> I basically want a pocket computer.

If you have hobbyist-level electronics assembly skills, a weekend, and a couple of hundred dollars, you can pretty easily assemble such a thing using a Raspberry PI with a touch screen, battery, and any of several commercially available cases (or 3D print your own).

I have one that I put together to use for pentesting. It works quite well.

I have several Pi's. To be more specific, I want a pocket computer that will fit in my pocket, and have more than a few hours of battery life.

Mine fits in my pocket (it's about 1cm thick) and the battery lasts about 4 hours if I keep the display on. The display is the big power draw, and I could have used a more expensive but less thirsty one.

The easiest way to meet the size constraint is to use the Pi Zero. I wanted more grunt than that, though, so I used a normal Pi and stripped the board of all the connectors except the micro USB to achieve the profile I needed.

> if I can trust a software kill

You can't. If the functionality can be enabled/disabled via software, then there's always the possibility that malware will enable it.

Sure. But to justify the three switches you must tell me more about this malware that would succeed with the baseband disabled in hardware while you continue to have wifi enabled in hardware.

I can only think of the special case where a carrier wants send a specially crafted message to exploit a flaw in FLOSS in order to implant malware that tricks the user about the status of the baseband. Still, if the carrier wanted to take that risk the only users who are protected are the ones who never switch on the baseband OS.

I don't think it makes sense to decrease usability for the off chance that some user has an idiosyncratic threat high enough to never use the baseband os, yet low enough to toggle wifi on and off.

> to justify the three switches

I'm not speaking to the number of switches, merely the importance of actual physical switches that physically disconnect the subsystems.

Reply on news.ycombinator.com