Hacker News Reader

Comment by gopowerranger | Hacker News Reader

People aren't paying attention to what this is for. It's to prevent people from modifying RF devices, transmitters, from operating at power levels and frequencies, or other technical things, from interfering with other devices and outside the bounds of their intention. It's to protect devices from interference that was prevalent in the days when radio was first invented, and for decades afterwards.

This is a good thing.

View on HN

Back here in the real world we understand that the most cost effective manner for device manufactuers to comply with this moronic regluation is to lock down the entire device.

Phone, Routers, Tablets and even some Laptops will be shipped with out the ability to change the operating system, or firmware. As that will be the cheapest way to comply with the regulation and since less than 1% of the customers ever bother changing there will be no backlash

This will be a nightmare for Computer security, and innovation.

I hope you love botnets becuase the FCC is creating the next wave of them with this moronic regulation.

Back here in the real world we understand that the most cost effective manner for device manufactuers to comply with this moronic regluation is to lock down the entire device.

What real world do you live in? Because in the real world I live in, the kind of component and firmware we're really talking about here are probably bought in as a ready-made part by whoever makes that phone, laptop, router, etc. They might not have any access to reprogram it at all.

What these kinds of regulations are effectively saying that if they do have that ability then for the device to remain properly certified for use as required by the regulatory authority there must be safeguards to prevent certain types of potentially harmful change or you "void the warranty" from a regulatory point of view. In that case, you revert to the default position you were in anyway, which is that you have to get the device and its software properly approved before you can lawfully use it.

There is nothing in that FCC paper that says you won't be able to change the operating system and that is specifically mentioned in section 24.

The FCC doesn't require it directly, but since it's the cheapest way to achieve what the FCC does require it's what the manufacturers will all do. That's the point.

Out of interest, can you name any botnets that have exploited mcu's with locked firmware, like mobile basebands?

And at the same time it locks down devices where radio and processor are connected – you won’t be able to modify the software on some Android devices at all anymore.

This is NOT acceptable.

Punishing people for sending on illegal frequencies: Okay.

Restricting what people can do with devices they paid for: NOT OKAY.

Why is restricting a device from interfering with emergency bands or bands used for medical devices unacceptable?

Because in some devices, to restrict the baseband from interfering, you also have to lock down the OS to the point where the user can’t modify it.

Also, another example: You recently saw the post on Hacker News how someone collected images from weather satellites by modifying a normal satellite receiver? This would not be possible under the new rules.

People should be punished for doing illegal stuff, not for being able to do illegal stuff.

>>Because in some devices, to restrict the baseband from interfering, you also have to lock down the OS to the point where the user can’t modify it.

Can you please give an example ? also couldn't be solved by designing the correct API for the communications processor ?

I'm fairly sure this applies only to SDR transmitters, not receivers.

There are just as many legal and legitimate applications of fully reprogrammable SDR transmitters.

What are these?

Every transmitter I can think of requires FCC type certification. If you're a police department, you can't build your own radio to use on the spectrum you've obtained from the FCC, you have to buy something that's type-certified. Similarly, you can't just go ahead and transmit on the unlicensed bands, that also requires type certification. SDRs are sold as test equipment; using them as radio transmitters is generally illegal.

Now if you're a licensed amateur radio operator, SDRs are great. You are completely allowed to use them to transmit on any frequency for which your license class allows. But there is a special provision in the laws for amateur radio operators; we are specifically licensed to build our own equipment. (And keep in mind that the use case for amateur radio service is limited by the same law; if you're not using it to establish 2-way communication with another amateur, you're doing something illegal. There are exceptions, though, and the FCC doesn't seem too picky. WSPR and other propagation beacons are strictly one-way affairs, but are tolerated.)

Nowadays every integrated transceiver is an SDR. It's much cheaper to build hardware that works over a range of frequencies and limit those frequencies via software than it is to make multiple SKUs of hardware that are hardware limited to a specific frequency.

Because in some devices, to restrict the baseband from interfering, you also have to lock down the OS to the point where the user can’t modify it.

I don't think "My decisions have made it hard for me to comply with this reasonable regulation so I shouldn't have to" is a very good argument.

People should be punished for doing illegal stuff, not for being able to do illegal stuff.

There is not much constructive use for something that would breach these regulations. Contrary to some of the FUD being posted in this thread, the regulations don't appear to be about custom firmware for the UIs or even signalling on a device. They're about the control software for the radio itself, to ensure it plays nicely with everyone else. That software is typically extensively tested and certified before being licensed for use, and any supplier whose software didn't work properly in the field could expect a very rapid and unsympathetic response from the authorities.

What a lot of people fail to understand is how disruptive a radio device that doesn't follow the standards can be. I worked for a radio network operator for a while. One day I got into work to find there were widespread reports of radios not operating properly across a major city. Someone was transmitting something out of spec and causing sufficient interference to effectively bring down the whole network. Do you know what happens at that point?

This is what happened that day. Real people got into real vehicles and started moving around a search pattern with relatively crude sensors, trying to locate the source of the interference. And then when they found it, they had to figure out how to shut it off. That involved gaining access to premises where the rogue device was located. Ideally you do that with the consent of the residents, because the operator of the rogue device may not even realise that it's not behaving correctly. Worst case, you're looking at forcing entry, which means bringing in other public services who are allowed to do that sort of thing with the appropriate authorisations, which in turn have to be sought from the appropriate authorities before anyone can move. And then once you've got into the premises you still have to find the device itself, and then you have to figure out how to turn it off.

Contrary to any movies you've seen, this process can take a considerable time, probably at least several hours unless you're very lucky. During that time the entire system will be operating at reduced capacity, if at all. Think about the range of a modern cell phone, and consider that any device that can transmit over that range can also interfere with the control channel for any signalling system within its range. Also consider that we're talking about changes that might increase the power and therefore the range of the device, or might change the device so it's using different frequencies and disrupting other forms of transmission/communication, not just whatever it was designed to work with originally.

Bottom line: even one small rogue device on a radio network can cause serious disruption over a wide area, and not all devices covered by these regulations will be small or have relatively low-powered transmitters. Imagine a whole city block or business park where wireless devices didn't work all day. Imagine a whole city where mobile/cell phones weren't reliable. Imagine a major incident where emergency services couldn't communicate properly to co-ordinate their response because their radios didn't work, or where the entire response was delayed by several minutes because no-one could find a landline to even call the emergency services in the first place. You get the idea. And just to be clear, this is not some paranoid fear of a hypothetical problem. It actually happens, just fortunately not very often, thanks in large part to the kind of regulatory regime we're talking about here.

It's true that with any sort of government regulation you have limitations on individual freedom. Laws do that, and should never be made lightly. But at the risk of ending on an inflammatory note, consider that while people can hold strong views on whether private individuals should be allowed by law to own and shoot guns, and reasonable people can debate the pros and cons of that, no reasonable person thinks it's a good idea for anyone who wants one to be allowed a WMD where they can push a button and kill everyone in a city. There simply aren't any good uses for that kind of technology, and the damage one person can cause to others is wildly disproportionate to any personal freedom given up by someone who for some reason wants such a device even though they can never use it. Obviously I'm not suggesting the threat from a rogue radio transmitter is comparable to the threat from a nuclear bomb, but it really is a significant risk and even potentially a matter of life and death.

Manufacturers don't care about whether end-users have the ability to install stuff like OpenWRT on their products, so if the easiest way to comply with the regulations is to lock them down altogether they'll do it. It's the users who lose out, and they're not the ones who are designing the products in the first place.

Why do people keep bringing thinks like OpenWRT into this? The regulations here are about the firmware in the radio components themselves, not any particular firmware running on a specific end user device that happens to include such a component.

The firmware running on the end user device and provided by its manufacturer probably doesn't even have access to the firmware running on the radio component in many cases, in which case the kind of regulations we're talking about would make no difference at all to that device.

If the main device firmware does have access to reconfigure the radio component, these regulations are a good thing, because it means you can afford to let end users swap out the main device firmware without risking them breaking the radio with adverse consequences for everyone else nearby. If you didn't have that safeguard, the only safe thing to do would be to lock down the entire device in exactly the way you're describing.

You are arguing against the regulatory framework that makes your desired outcome practical to achieve.

Those who want to run exclusively open source software and firmware will not like such an approach.

As far as I am aware, they (or the developers of the software they use) are free to seek the necessary regulatory approval to operate a radio running their choice of software instead. They just have to play by the same rules as everyone else. Running open source software is completely orthogonal to having the proper regulatory approvals to transmit radio signals.

The issue, as I see it, is that we ban owning something, instead of banning bad behaviour.

It’s similar to the gun debate or the "speed limits on the autobahn" debate.

Funnily, I’m on a different side in each of these 3 debates, but the issue is still there.

I understand the ethical conflict here. I'm generally liberal in my views and tend to favour lighter regulation by default, so personally none of these issues sit easily with me. I believe technology is neutral and how it is employed is what counts, so my usual stance is that freedom should be the default and with that freedom must come responsibility. But I'm also a pragmatist.

You can debate the merits of useful technologies that are also dangerous. Many technologies have both negative and positive applications. Sometimes in reality those technologies are used more often for negative purposes. We then get conflicts where the same tool that can protect an individual from harm can also harm others and/or protect the individual from the consequences of their harmful actions. We also get rather different kinds of conflict where a technology can be a net benefit to society because it helps many people but that comes at a high cost a few innocent people who will lose out in some way.

However, when there is nothing to balance in how a technology is employed, when realistically it can only be used for harmful things or it causes wildly disproportionate harm, I think the ethical stance must change as well. Some things can't be fixed afterwards, and in that case mere deterrence isn't always enough and I do think formal regulation can be justified.

The thing is, I think the balance here is at some different point.

In Germany, we ban guns, but don’t really restrict people modifying routers, as long as it doesn't interfere with anything.

> However, when there is nothing to balance in how a technology is employed

But in this case there is:

- open firmware

- security research

- running radios outside their regulated envelope without interfering e.g. because you're building a directed link between two sites in a rural area. With locked-down device you can't do that even if you got a permit. You'll have to buy more specialized, more expensive hardware

Bad behavior is already banned. People behave badly nonetheless.

Reading some of the other comments in the thread; two things become clear to me. People causing their devices to behave badly have no idea what the ramifications are, lacking knowledge in both regulatory issues ("what's DFS? what's TDWR?") and RF-related issues ("jacking up the power can't cause spurious emissions!") [note]. These people should not be modifying their radio equipment, they don't have the understanding required to do so safely.

That said, you can be licensed by the government to prove that you understand these things, and then the government gives you free reign to build your own equipment and basically do whatever you want. If I were modifying my WiFi equipment for Part 97 operation (questionably legal for various reasons I won't go into), I'd know to look for things like: interference with primary users of the frequencies, spurious emissions, permissible RF exposure limits, and so on.

"Unlicensed" doesn't mean you can do whatever you want. People seem to think it does, however, and the FCC has a very tiny enforcement budget, so it seems natural to go after the device manufacturers rather than the end users. The device manufacturers already pay to have their devices certified, a few more hoops don't matter to them (and save the taxpayers money), and a well-locked device will prevent even the most ignorant hobbyist from causing plane crashes, so I understand their logic.

I am probably going to write them a letter opposing the changes, but mostly from the perspective of amateur radio operators rather than "I can do whatever I want in my own home".

[note] Let me share some choice quotes from elsewhere in the thread:

"Sounds like those TDWR systems are in dire need of an upgrade if that's enough to cause fatalities."

"The hardware should know the amplitude of the input signal and what maximum gain can be applied to it before these detrimental effects set in. [...] I'm not going get a multi-gigahertz-capable scope costing tens of thousands of dollars just to tweak my home router."

It feels very much like a knee-jerk reaction. Oh no, they're taking our wifi away, instead of figuring out why the regulations are proposed, and why there are unintended consequences far beyond a little noise at the edges of the wifi spectrum. Additionally, they don't want to realize that like you described, the cheapest way is to put in a radio that doesn't need a binary blob to operate, because router makers are sure as hell not going to pay for the expensive re-certification process after every software update.

https://apps.fcc.gov/kdb/GetAttachment.html?id=1UiSJRK869Rsy...

Here's a draft of the rules where they specifically ask how You will prevent end users from installing DD-WRT.

Third-Party Access Control: What prevents third parties from loading non-US versions of the software/firmware on the device? Describe in detail how the device is protected from “flashing” and the installation of third-party firmware such as DD-WRT.

And so I ask the same to you that I asked to 'PhasmaFelis, who seems to have been the first poster of that link in this discussion: what is the actual legal status of that document (which unlike others linked here does mention DD-WRT by name), and to which devices does/would it apply?

In many cases, especially with lower-end routers, the firmware running on the end user device is the firmware running on the radio component - there is no seperate radio CPU.

In those cases, restricting the use of unapproved firmware seems to be exactly what these regulations are supposed to do, and as I've argued elsewhere, I think it is a reasonable step to take given the damage one ignorant user can cause with these systems.

> * Contrary to some of the FUD being posted in this thread, the regulations don't appear to be about custom firmware for the UIs or even signalling on a device*

This is incorrect. Various sites have done a really shitty job at linking directly to the relevant documents, but see this: https://apps.fcc.gov/kdb/GetAttachment.html?id=1UiSJRK869Rsy...

Note that vendors are required to "Describe in detail how the device is protected from 'flashing' and the installation of third-party firmware such as DD-WRT."

This is specifically about blocking all use of custom firmware.

What is the legal status of that document? I was looking at one that 'thejosh linked to:

https://www.federalregister.gov/articles/2015/08/06/2015-184...

That seems reasonably clear in what it's talking about, or at least what they're intending to cover. For example:

"The Commission proposed to require that an applicant for certification explicitly describe the RF device's capabilities for software configuration and upgradeability in the application for certification. This description would include all frequency bands, power levels, modulation types, or other modes of operation for which the device is designed to operate, including modes not enabled in the device as initially marketed. Also, an applicant for certification would have to specify which parties will be authorized to make software changes (e.g., the grantee, wireless service provider, other authorized parties) and the software controls that are provided to prevent unauthorized parties from enabling different modes of operation."

There's no reference to specific firmware platforms like DD-WRT or OpenWRT there.

Is this a case of some documents describing preliminary consultations and others a final situation that differed from what was contemplated originally?

For that matter, is it actually the case that some routers don't fully isolate the radio control software as intended by the regulations, in which case installing custom firmware like DD-WRT or OpenWRT really would be a risk, in which case the intent of the regulations really would be to prohibit the installation of that custom firmware in that particular context because it does present exactly the threat they are trying to control?

Yep, that's exactly the problem. We even have a few posters in this discussion talking about how they hate these new requirements because it means they can't run their routers illegally anymore. A lot of the fuss seems to be that the draft documentation of things that hardware vendors should consider and put in their certification request is, "what happens if an end user installs a third-party firmware like dd-wrt." They're seeming to take that as prohibiting all third-party firmwares, instead of prohibiting third-party firmwares on devices that require binary blobs. A router maker who has a baseband that can't be modified wouldn't be affected by this at all.

I've already covered a lot of this in my response to Silhouette, so I won't repeat it here, but:

> We even have a few posters in this discussion talking about how they hate these new requirements because it means they can't run their routers illegally anymore.

Are you under the impression that the sole purpose of custom router firmware is to access restricted channels? Custom firmwares contain many very useful, legitimate tools, and also happen to enable region-shifting for users who want to keep their old router when they move, which unfortunately lets malicious or ignorant users access locally unauthorized channels.

As for the rest, the U-NII Device Security document I posted above is pretty clear about prohibiting all unsigned firmware, explicitly including DD-WRT.

> What is the legal status of that document?

It's on fcc.gov, so it's probably legit. It dates back to March, so it's conceivable that it's been modified or superseded since then; I haven't been able to find anything to indicate so, but I'd be happy to be wrong.

Basically, there seems to be a lot of confusion stemming from blogs that heard about the document that I posted, but only linked to the document that thejosh posted. I found this doc via the CNXSoft post that appeared on HN yesterday [1], which mentioned the "protected from flashing" line and linked to Etherpad notes from a presentation, which linked to the U-NII Device Security v01r02 document that I posted above. It's kind of a ridiculous chain, so I can see why people were confused. You're right that none of the documents directly linked in this thread talk about blocking custom firmware; when people are talking about that, they're thinking of the U-NII Device Security document, not the "Equipment Authorization and Electronic Labeling for Wireless Devices" proposed rule.

> Is this a case of some documents describing preliminary consultations and others a final situation that differed from what was contemplated originally?

Possibly. They may also be describing different aspects of the FCC's evolving wireless policy. I haven't followed the rabbit hole deeply enough to determine for sure whether the FCC is still currently on track to block custom firmware, but it appears that they were as recently as March.

> For that matter, is it actually the case that some routers don't fully isolate the radio control software as intended by the regulations, in which case installing custom firmware like DD-WRT or OpenWRT really would be a risk

Oh, yes, certainly. Most of these firmwares are designed to be useful for users all over the world; IIRC, the one I use (Tomato) lets you select your region from a pull-down, and then enables the channels appropriate to your claimed region. (You can achieve the same thing by downloading the manufacturer's firmware intended for a different region; custom firmware is not required.)

I think this was well-meant on their part, but I agree that it is a problem, and it makes sense for new routers to block access to block access to unauthorized channels on a level that custom firmware can't touch. I just think that blocking custom firmware entirely is throwing the baby out with the bathwater.

[1] https://news.ycombinator.com/item?id=10137470

It's on fcc.gov, so it's probably legit.

But legit what? For example, I don't see any indication of whether it's an existing policy, a draft/proposal for future policy (which is what most of this discussion seems to be about) or something else. Who is required or recommended to read it? What happens if they don't do what it says?

Most significantly, it also doesn't seem to specify what constitutes a U-NII Device for the purposes of these rules. It would be bad if the rules had the effect of blocking custom firmware on something like a wireless router when that device included a self-contained radio component that was a black box and itself constrained to follow the normal rules for transmission equipment. On the other hand, if we're talking about running custom firmware on that radio component and/or on the main device but with the ability to reconfigure the radio component to perform outside the authorised spec, it seems to me that prohibiting the installation of that firmware is exactly what is intended, and for good reasons.

Receiving is not illegal. Sending on a restricted band, it is.

And how do you expect this is going to happen when the device modified it is usually a combined transceiver?

What if what they want to do with the device they paid for is modify it to send on licensed frequencies?

Do any android devices use open source firmware for their cellular radios? I was under the impression that all cellular radios worked only with closed source firmware + closed source radio driver, in which case _you already_ can't modify some of the software on any android device?

It does NOT restrict you from modifying the software. Read section 24.

"The Commission proposed that certain changes in layout, included components, operating software, or variations in overall electrical or mechanical constructions that do not substantially change the overall function of the device do not require a new FCC ID."

"The Commission proposed to continue to permit Class I permissive changes for those changes that do not degrade the device parameters normally reported in an equipment authorization application (including a decrease in the fundamental emissions that does not increase spurious emissions; an improved spurious emission performance; minor variations in the enclosure or components; and software changes that do not affect RF parameters)."

All of that says that a manufacturer can make certain changes to software and hardware without having to get a new authorization. It says nothing about what the user is permitted to do.

Yes, and the cheapest way to be compliant with that is to put the wifi chips in its own locked box, separate from the rest of the phone. There are several chips available right now, like the ath9k series, that already work like this. This issue really is a tempest in a teapot.

I'm repeating this in a few places because no one else in the thread seems to have found it yet: https://apps.fcc.gov/kdb/GetAttachment.html?id=1UiSJRK869Rsy...

The relevant passage is "Describe in detail how the device is protected from 'flashing' and the installation of third-party firmware such as DD-WRT." The FCC is specifically mandating that all custom firmware must be blocked.

A) The relevant portions of the rules only apply to the 5 ghz band. 2.4 ghz doesn't have that set of rules.

B) The document is not hard and fast rules, it's a set of guidelines to ensure that the device cannot operate out of the rules. So, putting the radio controls in a separate unflashable baseband chip would be ensuring that the radio is only operated by authorized software, while the rest of the router is operated by other software.

That's not right. Look at this: https://apps.fcc.gov/kdb/GetAttachment.html?id=1UiSJRK869Rsy...

One of the requirements for hardware vendors is "Describe in detail how the device is protected from 'flashing' and the installation of third-party firmware such as DD-WRT."

They are explicitly trying to illegalize custom router firmware, even though all the things that need to be prevented could be handled by separate hardware lockouts that wouldn't affect the legitimate uses of DD-WRT.

That page states that it's a set of general guidelines, not hard, definite rules. All of the examples, like flashing dd-wrt, are essentially asking, have you thought about x, y, and z? Tell how so. It's almost certainly perfectly acceptable to state, "we use a discrete chip for our wifi system without any way to upgrade the firmware, so if a user does flash their system, the wifi chip cannot be made to drive outside of legal parameters." A hardware lockout is definitely one way to comply with the guidelines.

People aren't paying attention to what this is for. It's to prevent people from using encryption to obfuscate and share their illegal data, and other criminal things, and from interacting with other criminals outside the bounds of the manufacturers intention. It's to protect devices from crime that was prevalent in the days when the internet was first invented, and for decades afterwards.

This is a good thing.

Reply on news.ycombinator.com