I understand having a locked-down version of mass-produced tech that is for the general consumer. However, I also firmly believe that anyone who wants to should be able to bypass that. Make it void the warranty, don't allow customer support, whatever you want. But make it a free option to have complete control. If I buy a piece of hardware, I should own it.
(I say this as I am currently charging a cheap doorbell camera I bought, so that I can start dumping data from it and reverse engineer its protocol, solely so that I can lock it down to my local network...We needed something at the time, or I would have bought something more expensive that had an open protocol...)
> Make it void the warranty, don't allow customer support, whatever you want.
That would be a regression in terms of consumer rights, in my opinion. I don’t see why the EU shouldn’t legislate such that the right to control owned hardware becomes fundamental.
In large, powerful and potentially dangerous machines, the firmware may embody safety features. Like preventing dangerous doses of radiation or overspeed conditions on motors or overcharging of batteries.
Why should you have a right to operate the device in a dangerous way that the manufacturer engineered to prevent?
You may be happy to injure yourself and to disclaim your right to warranty but the manufacturer could suffer immense reputational damage if, say, your hacking the firmware of their humanoid robot results in it tearing off your head.
There are a lot of devices where I agree with you, but smart glasses are an appliance category where this categorically needs to be considered a vulnerability. The main thing that currently stands between them being truly amazing and useful (e.g. for people with impaired vision or hearing) and a spying/peeping tool is the firmware on the device.
Smartglasses have an amazing potential to improve lives when done right and used right. But that relies on a sophisticated sensor hardware package that is barely distinguishable from a covert recording tool, simply because of the reality of the problem space. I don't want to see that tech totally canned as assistive technology because of backlash.
I think this is an area where society is going to adjust to fit the technology. The idea that you can not be recorded in places where eyeballs can see you is going to seem pretty quaint before long. Uncomfortable to our sensibilities, but probably not uncomfortable after the next few generations.
I think the current line in the sand is surreptitious recording with their glasses and the light. They have anti-tamper stuff around that, and I can see why they intentionally make root difficult.
People don't like recording, and there's usually a patchwork of laws that give many people some protection against it.
Not related to meta but the privacy angle is stuff like the Plaud recorders. Normal non-techies have had these, including personal run-ins of my own with people using these dedicated smartphone recorders where it was neither acceptable nor consensual.
I'm curious if ubiquitous and hidden personal recording collapses the technical or social lines set up against it first.
I am a proponent of giving users an option to root their devices through a one-way process that would surrender warranty coverage and print a big safety warning at every boot.
The problem with allowing easy shell access is that any arbitrary command execution would be used by exploits, too. There are also a lot of social engineering schemes that bait users into running commands on devices. We don’t think about these risks as engineers, but you wouldn’t believe how many people will follow instructions to open something like the Chrome developer console and paste code blocks into it because some website told them to.
So a secure device is also a locked down device. Getting a shell is a vulnerability because it’s the avenue that exploits would use.
That’s why I’d be in favor of a one-way developer mode, which provided shell access but printed a big unavoidable warning screen in the boot loader.
Unfortunately, it’s not that simple. Asking people to give up their warrant isn’t legal, generally, so outside of very specific circumstances you have to cover warranty claims on the device, too. I have worked at companies that tried to offer some developer access like this, and the number of people abusing customer service because they got in over their heads with modifications and wanted a fresh unit or want CS to support their mods would make you face palm. So the compromises end up being things like only allowing ephemeral mods that get wiped at each reboot and other default-safe behaviors, but that doesn’t make people happy either.
Following the reasoning behind the magnuson-moss act, it should not automatically void the warranty. That puts users who want software freedom at a financial disadvantage.
I agree some big obvious scare message is good but actually voiding the warranty restricts user freedom.
Do I really "have root" if you've put an "unavoidable warning" in the bootloader that I can't remove? It sounds like you're still asserting primary control over the device and giving me a limited playground. Your suggestion is "we're going to run code you don't control on your device, and it's going to run before any of your code, and there's nothing you can do about it, but don't worry we pinky promise to only make it print a warning scolding you for wanting root."
Nah. No thanks. I want actual freedom, not a longer chain.
This is a very neat and approachable write-up, but is the final result really a hack? If I understand correctly, you're getting an unprivileged shell on a device you own and have paired with your smartphone; more or less equivalent to installing a terminal emulator on your smartphone, except it's smart glasses in this case.
Should the ability to own one's devices really be treated as a vulnerability and reported to security teams?
How is it not a hack to do something on a device that doesn't let you do it? Yes, the device should let you do whatever you want, but it doesn't, therefore this is a hack.
I'm going to try it on my glasses, and disable autoupdating ASAP.
"hack" on hackernews essentially means "I used my computer to do something". The overwhelming population of people here aren't security engineers, but rather developers.
Sure but nowadays can you get unprivileged shell to your smartphone no matter the brand and model? Your TV? Your car infotainment system? Your smart fridge? I can continue on and on.
Actually having unprivileged access to something is quite rare nowadays in non-strictly computer products nowadays.
Maybe I’ve still got the old PS3/PSP-exploit mentality, but reporting an arbitrary code execution vuln to the OEM is kind of sad… that deliberately closes a door for people to use their devices as they see fit.
Of course, the “don’t update after this firmware version” bit would always be the case (since any widely-distributed such vuln will be patched eventually), but deliberately moving the timeline to the left is sad.
So many things are done with the end of making some profit, either directly monetary (bounties) or career.
People like these hope to get the recognition by the big companies to get the fat paychecks they dream to also pull in.
On the other hand, many others hack stuff and at most tell their closest friends.
Of course, it's availability bias: if this person wasn't disclosing it somehow, you wouldn't know. And even just a random article online is basically a disclosure minus the other benefits, so people contact the OEM while they are at it.
Regardless of the exploit, I wonder if there's a market for similar looking smartglasses that let you do whatever you want. Some competition in the space would be good for inviting regulation.
The non-displaying kind is just Bluetooth headsets with cameras and some hype. There are plenty of cheap knockoffs for those.
The display kind is full of exotic parts, materials and manufacturing techniques. Building those is not a product design problem but funding and engineering problems.
I haven't even seen a market for these, much less something that takes more work. It's probably hobby level at best to a very select hacker-focused crowd, but even they have historically been anti-sousveillance.
> "I used android.media.ToneGenerator in my DAT APK to signal which part of the code was executing:"
I've heard that morse code allows for easier debugging than beep counting, if you have to use audio. Not had a chance to put it into practice myself, but since digits especially are simple in morse code, it might be useful to someone
It's really sad that getting a shell on a device you own is now considered a vulnerability.
I understand having a locked-down version of mass-produced tech that is for the general consumer. However, I also firmly believe that anyone who wants to should be able to bypass that. Make it void the warranty, don't allow customer support, whatever you want. But make it a free option to have complete control. If I buy a piece of hardware, I should own it.
(I say this as I am currently charging a cheap doorbell camera I bought, so that I can start dumping data from it and reverse engineer its protocol, solely so that I can lock it down to my local network...We needed something at the time, or I would have bought something more expensive that had an open protocol...)
> Make it void the warranty, don't allow customer support, whatever you want.
That would be a regression in terms of consumer rights, in my opinion. I don’t see why the EU shouldn’t legislate such that the right to control owned hardware becomes fundamental.
3 replies →
In large, powerful and potentially dangerous machines, the firmware may embody safety features. Like preventing dangerous doses of radiation or overspeed conditions on motors or overcharging of batteries.
Why should you have a right to operate the device in a dangerous way that the manufacturer engineered to prevent?
You may be happy to injure yourself and to disclaim your right to warranty but the manufacturer could suffer immense reputational damage if, say, your hacking the firmware of their humanoid robot results in it tearing off your head.
9 replies →
[flagged]
There are a lot of devices where I agree with you, but smart glasses are an appliance category where this categorically needs to be considered a vulnerability. The main thing that currently stands between them being truly amazing and useful (e.g. for people with impaired vision or hearing) and a spying/peeping tool is the firmware on the device.
Smartglasses have an amazing potential to improve lives when done right and used right. But that relies on a sophisticated sensor hardware package that is barely distinguishable from a covert recording tool, simply because of the reality of the problem space. I don't want to see that tech totally canned as assistive technology because of backlash.
I think this is an area where society is going to adjust to fit the technology. The idea that you can not be recorded in places where eyeballs can see you is going to seem pretty quaint before long. Uncomfortable to our sensibilities, but probably not uncomfortable after the next few generations.
4 replies →
I think the difference between improving lives and spying is having the glasses be completely obvious about their purpose.
Would people using them because of visual impairment be OK with them being bright green?
2 replies →
I think the current line in the sand is surreptitious recording with their glasses and the light. They have anti-tamper stuff around that, and I can see why they intentionally make root difficult.
https://www.theverge.com/gadgets/962514/meta-privacy-light-t...
People don't like recording, and there's usually a patchwork of laws that give many people some protection against it.
Not related to meta but the privacy angle is stuff like the Plaud recorders. Normal non-techies have had these, including personal run-ins of my own with people using these dedicated smartphone recorders where it was neither acceptable nor consensual.
I'm curious if ubiquitous and hidden personal recording collapses the technical or social lines set up against it first.
I am a proponent of giving users an option to root their devices through a one-way process that would surrender warranty coverage and print a big safety warning at every boot.
The problem with allowing easy shell access is that any arbitrary command execution would be used by exploits, too. There are also a lot of social engineering schemes that bait users into running commands on devices. We don’t think about these risks as engineers, but you wouldn’t believe how many people will follow instructions to open something like the Chrome developer console and paste code blocks into it because some website told them to.
So a secure device is also a locked down device. Getting a shell is a vulnerability because it’s the avenue that exploits would use.
That’s why I’d be in favor of a one-way developer mode, which provided shell access but printed a big unavoidable warning screen in the boot loader.
Unfortunately, it’s not that simple. Asking people to give up their warrant isn’t legal, generally, so outside of very specific circumstances you have to cover warranty claims on the device, too. I have worked at companies that tried to offer some developer access like this, and the number of people abusing customer service because they got in over their heads with modifications and wanted a fresh unit or want CS to support their mods would make you face palm. So the compromises end up being things like only allowing ephemeral mods that get wiped at each reboot and other default-safe behaviors, but that doesn’t make people happy either.
Following the reasoning behind the magnuson-moss act, it should not automatically void the warranty. That puts users who want software freedom at a financial disadvantage.
I agree some big obvious scare message is good but actually voiding the warranty restricts user freedom.
2 replies →
"They might do something stupid with it" is just authoritarian. You would take our sticks
3 replies →
Do I really "have root" if you've put an "unavoidable warning" in the bootloader that I can't remove? It sounds like you're still asserting primary control over the device and giving me a limited playground. Your suggestion is "we're going to run code you don't control on your device, and it's going to run before any of your code, and there's nothing you can do about it, but don't worry we pinky promise to only make it print a warning scolding you for wanting root."
Nah. No thanks. I want actual freedom, not a longer chain.
> bait users into running commands on devices. We don’t think about these risks as engineers
As an engineer, you wouldn't believe how many times I've recently used `curl | bash` to install a piece of software
[flagged]
This is a very neat and approachable write-up, but is the final result really a hack? If I understand correctly, you're getting an unprivileged shell on a device you own and have paired with your smartphone; more or less equivalent to installing a terminal emulator on your smartphone, except it's smart glasses in this case.
Should the ability to own one's devices really be treated as a vulnerability and reported to security teams?
Manipulating a system to get unintended access is a hack under any common definition.
The intent of the owner of the artifact is the only consideration. The wish of some previous owner is irrelevant.
3 replies →
How is it not a hack to do something on a device that doesn't let you do it? Yes, the device should let you do whatever you want, but it doesn't, therefore this is a hack.
I'm going to try it on my glasses, and disable autoupdating ASAP.
"hack" on hackernews essentially means "I used my computer to do something". The overwhelming population of people here aren't security engineers, but rather developers.
I considered that, but the mentions of pwn2own and the bug bounty program have me thinking "hack" is used in the security/mainstream sense.
Hack as a term predates the concept of cyber security. It was originally "a quick and dirty solution that is clever"
https://en.wiktionary.org/wiki/hack
1 reply →
[dead]
Sure but nowadays can you get unprivileged shell to your smartphone no matter the brand and model? Your TV? Your car infotainment system? Your smart fridge? I can continue on and on.
Actually having unprivileged access to something is quite rare nowadays in non-strictly computer products nowadays.
Maybe I’ve still got the old PS3/PSP-exploit mentality, but reporting an arbitrary code execution vuln to the OEM is kind of sad… that deliberately closes a door for people to use their devices as they see fit.
Of course, the “don’t update after this firmware version” bit would always be the case (since any widely-distributed such vuln will be patched eventually), but deliberately moving the timeline to the left is sad.
So many things are done with the end of making some profit, either directly monetary (bounties) or career.
People like these hope to get the recognition by the big companies to get the fat paychecks they dream to also pull in.
On the other hand, many others hack stuff and at most tell their closest friends.
Of course, it's availability bias: if this person wasn't disclosing it somehow, you wouldn't know. And even just a random article online is basically a disclosure minus the other benefits, so people contact the OEM while they are at it.
Regardless of the exploit, I wonder if there's a market for similar looking smartglasses that let you do whatever you want. Some competition in the space would be good for inviting regulation.
The non-displaying kind is just Bluetooth headsets with cameras and some hype. There are plenty of cheap knockoffs for those.
The display kind is full of exotic parts, materials and manufacturing techniques. Building those is not a product design problem but funding and engineering problems.
I haven't even seen a market for these, much less something that takes more work. It's probably hobby level at best to a very select hacker-focused crowd, but even they have historically been anti-sousveillance.
> "I used android.media.ToneGenerator in my DAT APK to signal which part of the code was executing:"
I've heard that morse code allows for easier debugging than beep counting, if you have to use audio. Not had a chance to put it into practice myself, but since digits especially are simple in morse code, it might be useful to someone
> Beep is the new printf
Not for anyone that put PCs together in the 90's. Beep is also the old printf.
Nice. How robust is this with software upgrades though? Also looks like arbitrary code execution in the app sandbox, so no persistance?
Imagine shipping a firmware update to enforce APK signatures, only for the enforcement itself to be controlled by an unauthenticated config value
(dupe) https://news.ycombinator.com/item?id=50406224
That's this thread. Did you mean to link another?
HN does post merges sometimes. Could be that?
[flagged]