Coder's Rights Project – Reverse Engineering FAQ(eff.org)
eff.org
Coder's Rights Project – Reverse Engineering FAQ
https://www.eff.org/issues/coders/reverse-engineering-faq
8 comments
In Norway, we have an un-waverable right by law to reverse engineer computer systems for learning purposes.
They are described in Åndsverkloven §§ 39h and 39i.
Roughly translated excerpts:
§ 39h Whoever has the right to use a computer program, can copy, change and work on the program to the extent necessary in order to utilize the program in accordance with its purpose, including in order to correct errors in the program.
Whoever has the right to use a copy of a computer program can, during such reading, screen display, running, transfer or storage of the program that the user has the right to do, monitor, examine or probe the working of the program in order to determine the ideas and principles on which the individual parts of the program is based. [This right is un-waverable]
§ 39i You are allowed to produce a copy of a computer program's code and translate the form of the code, when this is a prerequisite in order to produce the information necessary to achieve functional interaction between a self-developed program and other programs, if
a) the actions are performed by a person who has the right to use a copy of a computer program, or acting on behalf of someone who do,
b) the information is not already easily accessible, and
c) the actions are limited to the parts of the original program necessary to achieve functional interaction.
[A few limitations on utilization of the information]
These rights are un-waverable.
They are described in Åndsverkloven §§ 39h and 39i.
Roughly translated excerpts:
§ 39h Whoever has the right to use a computer program, can copy, change and work on the program to the extent necessary in order to utilize the program in accordance with its purpose, including in order to correct errors in the program.
Whoever has the right to use a copy of a computer program can, during such reading, screen display, running, transfer or storage of the program that the user has the right to do, monitor, examine or probe the working of the program in order to determine the ideas and principles on which the individual parts of the program is based. [This right is un-waverable]
§ 39i You are allowed to produce a copy of a computer program's code and translate the form of the code, when this is a prerequisite in order to produce the information necessary to achieve functional interaction between a self-developed program and other programs, if
a) the actions are performed by a person who has the right to use a copy of a computer program, or acting on behalf of someone who do,
b) the information is not already easily accessible, and
c) the actions are limited to the parts of the original program necessary to achieve functional interaction.
[A few limitations on utilization of the information]
These rights are un-waverable.
> for learning purposes
> A few limitations on utilization of the information
I don't know how it is in Norway, but I see those clauses as problematic.
There is not much point in reverse engineering a piece of software if you can't use the results freely, including implementation of directly "competing" software.
That is, most of RE is just figuring out the data formats and protocols. And I fear that in too many jurisdictions one who had analyzed those can't also be the person who can legally write an alternate implementation, and a "clean-room engineering" is required.
> A few limitations on utilization of the information
I don't know how it is in Norway, but I see those clauses as problematic.
There is not much point in reverse engineering a piece of software if you can't use the results freely, including implementation of directly "competing" software.
That is, most of RE is just figuring out the data formats and protocols. And I fear that in too many jurisdictions one who had analyzed those can't also be the person who can legally write an alternate implementation, and a "clean-room engineering" is required.
As far as I know there is no legal precedence (and I'm not a lawyer), but I think that figuring out data formats and protocols counts as "ideas and principles" -- which is okay. It's hard to say, really, courts aren't famous for showing any understanding of computer science.
There is the DeCSS-case, which led to acquittal in Norwegian courts: http://en.wikipedia.org/wiki/DeCSS
There is the DeCSS-case, which led to acquittal in Norwegian courts: http://en.wikipedia.org/wiki/DeCSS
It seems countries in Northern Europe have some of the most liberal laws on RE - I wonder if the fact that the demoscene (which originated from software cracking culture) being quite established in that area is related in any way, or the fact that a lot of security researchers, who by necessity have to RE, are also from around that area.
This reads like disassembly of programs is legal. What about publishing that information?
I'm curious if non-US entities were ever prosecuted by US companies for reverse-engineering?
Dmitry Sklyarov?
http://en.wikipedia.org/wiki/United_States_v._ElcomSoft_and_...
http://en.wikipedia.org/wiki/United_States_v._ElcomSoft_and_...
Were any successfully prosecuted?
I'm curious if reverse engineering software which is classified or otherwise protected from disclosure by the force of government is legally risky. I presume doing so for those who have access to such software because of a security clearance might face a legal penalty under national security laws.
Good question, particularly if interpreted in a very broad sense. The
FCC is the portion of the US Fed Gov that handles things like validating
the acceptable RF interference and frequency transmit/receive parameters
of products. If you reverse engineer something, and the resulting
reimplementation violates the FCC rules, then you'll have legal trouble.
Similar would be true for Europe with the CE regulatory/testing body.
For example, most wifi (802.11-) chipsets have "modes" for various countries due to the fact that the "freely available" RF spectrum varies from country to country. Though illegal (according to the FCC/CE/etc), it is often possible to make wifi chipsets use chunks of frequency that are not legally accessible (within some jurisdiction). This can be done via software (drivers) or firmware (blobs), as well as through the use of additional or modified hardware (amps, antennas, ...).
When you get into "licensed spectrum" that is supposedly "owned" by a company like a mobile carrier, things get even more "legally risky" since messing around in those bands can wreak a lot of havoc. I don't know if the operation of the (mobile) baseband processors qualify as "classified" but such info is definitely protected from disclosure by both technical (hardware/software protections) and legal (contracts, NDA, etc.) means. Of course, using a self-modified baseband processor is definitely illegal since the device has not been tested and approved by the required regulatory body (FCC or CE), but the laws do have some limited exclusions/registrations for test/development purposes.
If you cause problems for a carrier that spent many billions to own/license a chunk of spectrum, then you also open yourself up to civil liability for damages, in addition to the legal problems you'll have with the Fed Gov regulatory body. Needless to say, if you go mucking around in the frequency bands dedicated by law for use by the military or law enforcement, then you are going to have a whole lot of legal problems.
For example, most wifi (802.11-) chipsets have "modes" for various countries due to the fact that the "freely available" RF spectrum varies from country to country. Though illegal (according to the FCC/CE/etc), it is often possible to make wifi chipsets use chunks of frequency that are not legally accessible (within some jurisdiction). This can be done via software (drivers) or firmware (blobs), as well as through the use of additional or modified hardware (amps, antennas, ...).
When you get into "licensed spectrum" that is supposedly "owned" by a company like a mobile carrier, things get even more "legally risky" since messing around in those bands can wreak a lot of havoc. I don't know if the operation of the (mobile) baseband processors qualify as "classified" but such info is definitely protected from disclosure by both technical (hardware/software protections) and legal (contracts, NDA, etc.) means. Of course, using a self-modified baseband processor is definitely illegal since the device has not been tested and approved by the required regulatory body (FCC or CE), but the laws do have some limited exclusions/registrations for test/development purposes.
If you cause problems for a carrier that spent many billions to own/license a chunk of spectrum, then you also open yourself up to civil liability for damages, in addition to the legal problems you'll have with the Fed Gov regulatory body. Needless to say, if you go mucking around in the frequency bands dedicated by law for use by the military or law enforcement, then you are going to have a whole lot of legal problems.
There is nothing "secret" about the operation of basebands, or what they put out on the air. This is just the companies protecting trade secrets, trying to avoid patent litigation, or just general open source averseness.
Linux includes an open source regulatory information database that several open source wifi drivers use, so FCC licensing is not an issue for open source software either. It is not a form of DRM and does not require any "black boxing". Obviously you might be able to modify the software so you violate FCC rules, but you can also build your own radio transmitter and violate the FCC rules just as well. There's no special case for it being software.
If you want to experiment on the radio bands, you can trivially get a ham radio license and modify the radio as you see fit, as long as you stay within the ham radio limits.
Linux includes an open source regulatory information database that several open source wifi drivers use, so FCC licensing is not an issue for open source software either. It is not a form of DRM and does not require any "black boxing". Obviously you might be able to modify the software so you violate FCC rules, but you can also build your own radio transmitter and violate the FCC rules just as well. There's no special case for it being software.
If you want to experiment on the radio bands, you can trivially get a ham radio license and modify the radio as you see fit, as long as you stay within the ham radio limits.
Interestingly, four out of five court cases mentioned are from games industry.
The gaming industry is well known for exercising as much control as possible over their hardware and software. Console manufacturers want to control what runs on the console and software authors want to prevent piracy.
Has EFF ever addressed if we have the right to root our phones?
Because smartphones are getting more and more locked down, they have reached the point in late 2014 where many models simply cannot be rooted by 3rd parties.
Note I am not talking about carrier unlocking, I mean root.
I want full control over any device I use.
It is like buying a car but not being allowed into the trunk.
Because smartphones are getting more and more locked down, they have reached the point in late 2014 where many models simply cannot be rooted by 3rd parties.
Note I am not talking about carrier unlocking, I mean root.
I want full control over any device I use.
It is like buying a car but not being allowed into the trunk.
It would depend on what exactly prevents you from getting root rights on your device.
If it's DRM, to address that you'll need to fix DMCA-1201. Or more precisely, allow breaking DRM for any non infringing purpose. Right now it forbids it, and instead provides idiotic method of Librarian of Congress deciding what exceptions to grant.
This corrupted law was passed undemocratically using backdoor method of WTO trade agreements.
If it's DRM, to address that you'll need to fix DMCA-1201. Or more precisely, allow breaking DRM for any non infringing purpose. Right now it forbids it, and instead provides idiotic method of Librarian of Congress deciding what exceptions to grant.
This corrupted law was passed undemocratically using backdoor method of WTO trade agreements.
> allow breaking DRM for any non infringing purpose
Why not just make building DRM into a consumer product illegal?
Why not just make building DRM into a consumer product illegal?
That's perfectly fine with me, I view DRM as unethical. But trying to propose banning all DRM will more likely to fail than proposing to allow breaking it for non infringing purposes.
> This corrupted law was passed undemocratically using backdoor method of WTO trade agreements.
No, while the anticircumvention prevention is in a Title of the DMCA named for WIPO treaties that (among other treaties) were implemented in US law in other provisions of the title, the anticircumvention provision was not part of those treaties and, in any case, neither it nor any part of the DMCA was "passed" through trade agreements, the DMCA was passed through the normal legislative process.
No, while the anticircumvention prevention is in a Title of the DMCA named for WIPO treaties that (among other treaties) were implemented in US law in other provisions of the title, the anticircumvention provision was not part of those treaties and, in any case, neither it nor any part of the DMCA was "passed" through trade agreements, the DMCA was passed through the normal legislative process.
Here is a more elaborate source for you: https://www.eff.org/deeplinks/2013/03/ustr-secret-copyright-...
> At the conference, Lehman [known as the architect of DMCA] admitted the law was the product of a deliberate end-run around the democratic process. Lehman was an advocate for several hardline proposals to criminalize digitial rights management (DRM) circumvention. Unable to sell the proposals domestically, Lehman pressed the Office of the U.S. Trade Representative (USTR) to propose them at the UN World Intellectual Property Organization (WIPO) instead. Many have accused Lehman of using the treaty process to avoid Congress. What was Lehman’s response to those accusations at the event? “I would say that they're right.”
Nothing more to add there, it's straight from the creator of DMCA. Corrupted law in practice.
> At the conference, Lehman [known as the architect of DMCA] admitted the law was the product of a deliberate end-run around the democratic process. Lehman was an advocate for several hardline proposals to criminalize digitial rights management (DRM) circumvention. Unable to sell the proposals domestically, Lehman pressed the Office of the U.S. Trade Representative (USTR) to propose them at the UN World Intellectual Property Organization (WIPO) instead. Many have accused Lehman of using the treaty process to avoid Congress. What was Lehman’s response to those accusations at the event? “I would say that they're right.”
Nothing more to add there, it's straight from the creator of DMCA. Corrupted law in practice.
> the DMCA was passed through the normal legislative process.
Wrong. First of all those treaties took place before DMCA (including 1201) was passed. And there was nothing normal about it. Normal process actually rejected such idiotic restriction outright (since DRM proponents first attempted to pass it normally).
When they were kicked by normal democratic process, they turned to WIPO treaties which have no democratic oversight. Corruption there is much easier, so they squeezed anticircumvention provisions into such treaties affecting US. Then they turned back to the Congress and said: "you can't ignore your international obligations! Pass this law". And Congress passed it without much of opposition. That's not called a normal democratic process, it's called corruption and backdoor dealing.
See: https://en.wikipedia.org/wiki/WIPO_Copyright_Treaty
Wrong. First of all those treaties took place before DMCA (including 1201) was passed. And there was nothing normal about it. Normal process actually rejected such idiotic restriction outright (since DRM proponents first attempted to pass it normally).
When they were kicked by normal democratic process, they turned to WIPO treaties which have no democratic oversight. Corruption there is much easier, so they squeezed anticircumvention provisions into such treaties affecting US. Then they turned back to the Congress and said: "you can't ignore your international obligations! Pass this law". And Congress passed it without much of opposition. That's not called a normal democratic process, it's called corruption and backdoor dealing.
See: https://en.wikipedia.org/wiki/WIPO_Copyright_Treaty
> First of all those treaties took place before DMCA-1201 was passed.
But they aren't self-executing, and, in any case, DMCA 1201 isn't -- despite being in part of the DMCA titled for various treaties that were implemented in that Title -- part of them in any case. DMCA (the whole thing, including 1201) was passed as normal legislation by the House and Senate and with the signature of the President. You can argue that that is undemocratic.
> When they were kicked by normal democratic process, they turned to WIPO treaties which have no democratic oversight.
Since any treaty, to be law in the US, requires ratification by the Senate, it is simply false to state that they have "no democratic oversight". But its irrelevant in this case, since DMCA 1201 is a provision of statute law adopted through the regular legislative process, not something that is binding law as a result of at treaty without implementing legislation adopted through the regular legislative process.
> Then they turned back to the Congress and said: "you can't ignore your international obligations! Pass this law".
A tactic which is not actually effective when Congress wants to ignore the "international obligation" at issue. At most, one could argue -- if the anticircumvention provisions were actually contained in any of the treaties that were being implemented in the DMCA, which they weren't -- that the treaties provided political cover for something Congress wanted to do.
> That's not called a normal democratic process, it's called corruption and backdoor dealing.
To the extent there was corruption or backdoor dealing involved, it wasn't WTO or WIPO treaties "passing" the provision. You can argue that the regular legislative process was subject to corruption or backdoor dealing -- such is not at all unheard of in US legislative processes -- but any treaty process is a sideshow to that in this case, since the provision didn't become US law by being included in a treaty, it became US law by being included in a regular bill passed through Congress and signed by the President.
But they aren't self-executing, and, in any case, DMCA 1201 isn't -- despite being in part of the DMCA titled for various treaties that were implemented in that Title -- part of them in any case. DMCA (the whole thing, including 1201) was passed as normal legislation by the House and Senate and with the signature of the President. You can argue that that is undemocratic.
> When they were kicked by normal democratic process, they turned to WIPO treaties which have no democratic oversight.
Since any treaty, to be law in the US, requires ratification by the Senate, it is simply false to state that they have "no democratic oversight". But its irrelevant in this case, since DMCA 1201 is a provision of statute law adopted through the regular legislative process, not something that is binding law as a result of at treaty without implementing legislation adopted through the regular legislative process.
> Then they turned back to the Congress and said: "you can't ignore your international obligations! Pass this law".
A tactic which is not actually effective when Congress wants to ignore the "international obligation" at issue. At most, one could argue -- if the anticircumvention provisions were actually contained in any of the treaties that were being implemented in the DMCA, which they weren't -- that the treaties provided political cover for something Congress wanted to do.
> That's not called a normal democratic process, it's called corruption and backdoor dealing.
To the extent there was corruption or backdoor dealing involved, it wasn't WTO or WIPO treaties "passing" the provision. You can argue that the regular legislative process was subject to corruption or backdoor dealing -- such is not at all unheard of in US legislative processes -- but any treaty process is a sideshow to that in this case, since the provision didn't become US law by being included in a treaty, it became US law by being included in a regular bill passed through Congress and signed by the President.
See my response here (https://news.ycombinator.com/item?id=8840676), it quotes the creator of DMCA himself who clearly states that he intended to bypass the democratic process. That's unquestionably corruption and a perfect example of undemocratically passed law.
It also highlights how broken the current system is, when corrupted international treaties are used to shape the law bypassing normal democratic ways.
> A tactic which is not actually effective when Congress wants to ignore the "international obligation" at issue.
It should ignore it in theory, but in practice it doesn't have the guts and gives the power away to USTR and the like. That's what all this fast track idiocy is exactly about.
It also highlights how broken the current system is, when corrupted international treaties are used to shape the law bypassing normal democratic ways.
> A tactic which is not actually effective when Congress wants to ignore the "international obligation" at issue.
It should ignore it in theory, but in practice it doesn't have the guts and gives the power away to USTR and the like. That's what all this fast track idiocy is exactly about.
> See my response here (https://news.ycombinator.com/item?id=8840676), it quotes the creator of DMCA himself who clearly states that he intended to bypass the democratic process.
How he describes his intent is irrelevant. The fact is that the law was passed through the House, Senate, and Presidents signature in exactly the manner laid out in the Constitution, not "by WTO treaty". And, while some of the law implemented things previously agreed to in WIPO and other treaties, the particular provision at issue -- Section 1201 -- did not.
It might be undemocratic and/or corrupt, but if so the undemocratic/corrupt mechanism by which Section 1201 was passed is the normal US legislative process.
> It should ignore it in theory
It does in practice, when it is something that Congress doesn't want to do anyway -- this happens quite a lot on non-trade issues.
> but in practice it doesn't have the guts and gives the power away to USTR and the like.
More likely, on trade issues Congress takes advantage of treaties to provide for political cover for things Congress would like to do but where there might be greater political costs without that cover.
How he describes his intent is irrelevant. The fact is that the law was passed through the House, Senate, and Presidents signature in exactly the manner laid out in the Constitution, not "by WTO treaty". And, while some of the law implemented things previously agreed to in WIPO and other treaties, the particular provision at issue -- Section 1201 -- did not.
It might be undemocratic and/or corrupt, but if so the undemocratic/corrupt mechanism by which Section 1201 was passed is the normal US legislative process.
> It should ignore it in theory
It does in practice, when it is something that Congress doesn't want to do anyway -- this happens quite a lot on non-trade issues.
> but in practice it doesn't have the guts and gives the power away to USTR and the like.
More likely, on trade issues Congress takes advantage of treaties to provide for political cover for things Congress would like to do but where there might be greater political costs without that cover.
> How he describes his intent is irrelevant. The fact is that the law was passed through the House, Senate, and Presidents signature
Using corrupt manipulation scheme and pressure by international trade agreements like described above. When democracy doesn't work, the fact that it's called democracy on paper is pretty meaningless.
> It might be undemocratic and/or corrupt, but if so the undemocratic/corrupt mechanism by which Section 1201 was passed is the normal US legislative process.
Yes, it is undemocratic and corrupt and it was passed through the legislative process, but as I said there was nothing normal about it. That process is not supposed to be manipulated by such treaties in such corrupt fashion, yet it is.
Unless you simply mean that such level of corruption is generally a norm and this participial incident should be a default expectation of how Congress would function. It doesn't fit with what Lehman said that initially his attempts to pass these laws were rejected. I.e. he didn't manage to manipulate the Congress directly, and had to pull the weight of international agreement to change that.
> It does in practice
No it doesn't. Otherwise such stuff as fast track proposals to relinquish Congress oversight won't be even considered.
Using corrupt manipulation scheme and pressure by international trade agreements like described above. When democracy doesn't work, the fact that it's called democracy on paper is pretty meaningless.
> It might be undemocratic and/or corrupt, but if so the undemocratic/corrupt mechanism by which Section 1201 was passed is the normal US legislative process.
Yes, it is undemocratic and corrupt and it was passed through the legislative process, but as I said there was nothing normal about it. That process is not supposed to be manipulated by such treaties in such corrupt fashion, yet it is.
Unless you simply mean that such level of corruption is generally a norm and this participial incident should be a default expectation of how Congress would function. It doesn't fit with what Lehman said that initially his attempts to pass these laws were rejected. I.e. he didn't manage to manipulate the Congress directly, and had to pull the weight of international agreement to change that.
> It does in practice
No it doesn't. Otherwise such stuff as fast track proposals to relinquish Congress oversight won't be even considered.
What are fair use regulations in EU in general? Are they worse than in US?
Also, how exactly does it work when contractual prohibitions try to take away fair use rights? What wins? In this FAQ it's only mentioned that it's risky.
Also, how exactly does it work when contractual prohibitions try to take away fair use rights? What wins? In this FAQ it's only mentioned that it's risky.
There is a major discussion now about GOG changing their TOS trying to prohibit reverse engineering and as well introducing DRM-like password restriction in their installer packages.
See here:
* https://www.gog.com/forum/general/please_fix_your_user_agree...
* https://www.gog.com/wishlist/site/to_retain_your_drm_free_st...
* https://www.gog.com/forum/general/on_gnulinux_has_anyone_be_...
* https://www.gog.com/wishlist/site/dont_slip_into_drm_swamp_s...
See here:
* https://www.gog.com/forum/general/please_fix_your_user_agree...
* https://www.gog.com/wishlist/site/to_retain_your_drm_free_st...
* https://www.gog.com/forum/general/on_gnulinux_has_anyone_be_...
* https://www.gog.com/wishlist/site/dont_slip_into_drm_swamp_s...
Just the other day I was pondering who to contact to determine whether my current method of figuring out how the undocumented API works (typing command guesses over the CDC-ACM serial interface in gtkterm) puts me at risk, but it doesn't look like it. I'm not dumping the firmware and don't have a copy of it, not decompiling the original software, there is no access control on the interface, no encryption, no handshake, no EULA on the hardware itself.
Seems that I'm in the clear from what I can tell.
I've gotten enough of the commands documented that I've been able to write a library based on what I've learned, and I'm hoping to publish my documentation and the code soon. It's been quite fun to figure out how it works so I and others can actually use these things in other software :)
[1] http://dl.ubnt.com/newsletters/0112.html