A few of the questions asked about "open source software" in such a way that, responding to them directly, I'd be classifying programs as "open" or "closed". That I will not do, because those terms presuppose a different philosophy based on different values.
Rather than give no answer to those questions, I modified them to say "free software" instead, and answered them that way. (Square brackets show these changes.) I hope the answers to these modified questions are of interest to readers. They are rather different from what an open source supporter would say.
What are your views on the recent NSA activities and how do you think it will change free software & the internet?
RMS: Nonfree software is likely to spy on its users, or mistreat them in other ways. It is software for suckers.
Awareness of this is spreading, which helps us make the case for free software to people who are not computing experts.
As for the internet, it has been turned into a spy network. A considerable fraction of the massive surveillance (but not all!) applies to the internet. Most use of the internet involves web sites that snoop on users, which is poisonous. That's in addition to the snooping by ISPs themselves.
Massive surveillance of people in general endangers human rights and democracy; but we should remember that US snooping agencies do this mostly by piggy-backing on businesses that massively collect data about people.
Therefore, it is not enough to legally limit the government's access to the digital dossiers about us. We must prevent those dossiers from being made, either by business or by government. We must legally require digital systems to be redesigned so that they do not accumulate data about people in general.
Here is my full position on massive general surveillance.
by Anonymous Coward
What is your opinion on cryptocurrencies?
RMS: In general, I am in favor of ways to pay each other cash on the Internet without going through a payment company that keeps track of all payments. I would like to be able to pay an on-line service with cash the way I pay cash for all the things I buy today.
However, Bitcoin payments are not anonymous. To serve this need requires anonymity at least for the payer. People are working on trying to improve Bitcoin in that way.
I am not an expert on encryption, and I can't judge the security of any particular cryptocurrency. What I do know, and what is illustrated by the recent collapse of several exchanges (banks, in effect) due to robbery, including MtGox, demonstrates that, here as in any field, the security of a practical activity that uses encryption is a very different question from the mathematical validity of the encryption system or the correctness of the software. It may take years to develop cryptomoney exchanges we can have confidence in.
These currencies raise economic issues, too; but not necessarily the way many people think. The number of bitcoins is capped, but new cryptocurrencies can always be created, so that the total number of Xcoins for all values of X has no particular limit. Does this mean that the value of all cryptocurrencies will inevitably tend towards zero? Not necessarily; that depends on how much people accept various other cryptocurrencies -- a sociological question, not an economic one.
I don't enjoy risk, so I will not do speculation in cryptocurrencies any more than I do in other commodities. I may use them for payment if and when it becomes possible to use them anonymously to buy something that I can't get with cash. To resist surveillance, I do buy goods with cash in a store, so that no data base knows what I bought -- therefore, I don't pay over the internet. But I would use an anonymous cryptocurrency to pay for services and downloads.
by Anonymous Coward
I read a little on your website about your take on technology that uses non-free software. Do you still not own a cell phone?
RMS: I certainly do not! A cell phone is Stalin's dream: its movements are tracked, and it can be converted (through the universal back door) into a listening device.
AC: If not, I'd love to hear your perspective on life without one these days, where its just assumed that people own one.
RMS: Please help teach everyone that this assumption is false!
There is a way to make a cell phone acceptable _for occasional communication only_: put a one-way pager in the phone, so people can page you if they are trying to reach you. That way, you can keep its radio connection off most of the time. When you get the page, you can decide when and where to reveal your location by connecting the phone to the network.
Of course, the software in the phone's main computer should also be free, but that is a separate issue. In other words, nonfree software in that computer is one assault on your freedom, and the phone system's location tracking is another.
The software in the baseband (phone radio modem) processor can't be free, at least not as things stand now. So the phone needs to be designed so the baseband processor can't talk to anything (peripherals, antenna, etc) unless the main processor permits it, and so that the baseband processor can't change the software in the main processor. Ideally the software in the baseband processor should be immutable, so we can treat it as a circuit.
AC: As a follow-up, where exactly do you draw the line concerning [freeness of software] and whether or not you use software. For example, do you toast bread in a toaster that runs proprietary code? Obviously we're talking about different things here, but I'm curious to know at what point you say "no thanks!" when it comes to locked down technology.
RMS: The case of the toaster is very clear: we can't tell, except by taking it apart, whether it has a processor and software or a special-purpose chip. Since that we can't tell the difference, it makes no difference: therefore, a program that will never be changed is equivalent to a circuit. I don't care whether a toaster or microwave oven contains software.
A very common design approach nowadays is an appliance or peripheral that contains software that could be changed, but normal use does not include changing it. I think we can still disregard that software, as regards the ethical issue of free vs nonfree software; it is just a short way into thr gray area. However, such devices can be a terrible security threat, because a corrupted computer can install malware in them that will propagate. Devices which have this problem include USB sticks, microSD cards, disk drives, and the cameras that go in computers.
Where is the other side of the line? If the device has an "update firmware" button, that firmware is software meant to be changed, so it is unacceptable.
by Anonymous Coward
Dear RMS, I for one am very interested in what your view is concerning the future of GNOME and specifically GTK. In the past there were concerns over licensing between GTK and Qt and there seems to be a rise in uptake of Qt. My question is whether you see there being a future in GTK and should developers consider moving their projects to Qt?
RMS: I can't see the future, because nobody can. I hope that GNOME and GTK will be very successful. Please help make it so.
Please share your vision for where you would like to see GNU/Hurd, and GNU software over the next 25 years, and what people would be doing with it.
RMS: I regret to say I have no response. I never try to think about what computing might be like 25 years from now; it would be a waste of time, since I know that I don't know.
I can tell you something about free software 5 years from now: most of it will be the same as today. Free software does not change rapidly. (I think that is a feature; our society teaches people to overvalue innovation so as to distract them from more important things such as freedom, democracy, and giving everyone a comfortable life). Most of the GNU/Linux system in 5 years will be the same as what we have now; some components will be new, but they will be a small change compared with the system as a whole.
The GNU Hurd kernel (and the GNU/Hurd system, which is GNU/Linux with the Hurd instead of Linux) is not a high priority for us any more, because it would be a replacement for the free parts of Linux, and we don't need to replace those. Volunteers continue to work on the Hurd, because it is an interesting technical project.
The parts of Linux we need to replace are the nonfree parts, the "binary blobs". But replacing those has nothing to do with the GNU Hurd. The main work necessary to replace the blobs is reverse engineering to determine the specs of the peripherals those blobs are used in.
That's a tremendously important job -- please join in if you can.
Free hardware? Why not?
In my experience; it is far easier to obtain; install and work with Free Software than with Free Hardware. I asked you about this in person 2 years back; but you brushed it aside saying hardware is not trivial to copy. Recent events have proved me right; I feel. We simply do not have access to Freedom Hardware at low cost - even the Raspberry Pi has proprietary components in its hardware.
RMS: When you say "free hardware" I think you mean hardware whose specs are known, so we can develop free software to run it. I call that "documented hardware". When I say "free hardware", it means to transpose the concept of free software to hardware. This means People are free to copy and change the hardware; if it is made from a design, that design must be free, with the same four freedoms that define free software. But that is mostly an issue for future technology. Documented hardware is what we need now.
The scarcity of documented hardware is indeed a tremendous problem. In general I don't see any way we can fix it except by reverse engineering to figure out the specs.
jkrise: Why can't the FSF pool resources; license technology from ARM Holdings; and build a truly Free Tablet, Free Cellphone and Free PC running Free GNU/Linux instead of the pseudo-free Android? I am sure the community will pay any money to buy truly free Hardware from the FHF.
RMS: This would cost millions of dollars, and we have no skills or experience in hardware manufacturing, so we couldn't do it.
We could try to raise funds to pay for reverse engineering of the VPU in the Novena laptop -- if we could find skilled reverse engineers ready to take the job. Can you introduce me to any?
I believe you're in favor of much-reduced copyright terms - a few years rather than the endless decades of today. If copyright were reduced to, say, five years, then the vast majority of GNU code would become public-domain - copyleft depending on copyright as it does, this would mean anyone could create a [proprietary] fork of, say, emacs. How do you feel about that?
RMS: For this very reason, I oppose shortening copyright to 5 years without making some other change to prevent this harmful consequence. See http://www.gnu.org/philosophy/pirate-party.html.
With the 10-year copyright term I propose, this problem would not be significant.
People often identify proprietary software with copyright; there was a time when I did, too. However, that's a mistake. The two principal methods used to make programs proprietary are (1) EULAs (a legal method) and (2) keeping the source code secret (a technical method). Two secondary methods are (3) copyright (a legal method) and (4) putting the executable in a tyrant device (a technical method, see below). Patents are used too, but only to reinforce the others.
To defend our free software from being made nonfree, the only one of these four that we can use is copyright.
People like apps
There is an entire generation of people out there for whom mobile apps, mostly on iOS and Android, are the way in which they do their computing. The more successful apps are usually very well-designed with incredible user interfaces, an area where free software has not achieved much success, and sold at very low prices and,
RMS: These "advantages" can seem impressive to those who don't see what they cost in freedom. The most basic thing we must do is say, "I'd rather have nothing than have that," and then act accordingly.
thetagger: in many cases, also monetized through stolen personal data.
RMS: Please don't use "monetized" to mean "make money from". That word stinks of the attitude that "Profit justifies anything". See http://www.gnu.org/philosophy/words-to-avoid.html.
Besides which, the word's correct meaning is "to use something as a currency."
thetagger: It appears to me that the GNU project is mostly ignoring this important area - I am aware of Replicant and F-Droid but these are well behind their proprietary counterparts at the moment. What should we do? Ignore mobile and hope it goes away,
RMS: I personally will ignore it, because there is nothing about it that I want. Even if we assume it is has no phone radio connection, so it is not Stalin's dream, a computer with a small screen and no keyboard is so inconvenient as to be useless for me.
However, we need to try to bring freedom to mobile computer users. We must not ignore them.
thetagger: try to get onboard with Replicant and F-Droid,
RMS: If you want to use mobile computers, please contribute in this way.
But we will never have, in the free world, the sort of "social" snooping apps that so many internet users spend their time in. We can't compete in terms of the misguided values that our adversaries promote in order to ensnare people, and if we did, we would be doing wrong. We have to set an example of rejecting those values.
thetagger: try to bring in a new generation of free software developers that is native to the mobile environment,
RMS: If this is meant as an alternative to the previous two, I don't understand what it means. We welcome people of any and all generations in everything we do.
thetagger: or avoid the mobile "ecosystem" completely
RMS: In general, I avoid the word "ecosystem" in connection with computing because of its amoral premises.
In this case I'm at a loss for what it means. I don't understand how this option differs from the first option, "Ignore mobile and hope it goes away."
thetagger: and try to work on the hardware side and try to make free hardware that is not inherently trackable/centralized and then run free software on top of that instead?
RMS: When you say "free hardware", I think you mean documented hardware. (See above.)
We can in principle make our own documented hardware, but the only way that would directly help is by avoiding the need for reverse engineering to figure out how to run the peripherals. In practice, though, I think reverse engineering is probably easier.
However, preventing the tracking is another matter. The only way I can envision to prevent the tracking of geolocation of mobile phones is if you have them disconnected from the network nearly all the time. (Well, in theory it might work to carry a parabolic antenna so you can communicate with just one tower. Maybe that would prevent the use of triangulation to figure out where you are located. I don't know whether this could be made to really work. Does anyone want to try it?)
Fundamentally, privacy-preserving computing has to be done mainly in your own computer. We have to reject the dependence on servers that the proprietary world is pushing people into. Freedom requires local application programs, rather than "web apps" or server-backed "mobile apps".
Do you foresee a viable Free Car OS?
Automobile user interfaces have become increasingly complex and de-standardized as computerization reaches into the driver's seat. The major vendors don't seem to care about possible legal liabilities of designing inherently dangerous UIs. Google has enticed Honda, GM and Audi to join the Open Automotive Alliance, but that project seems more oriented towards selling android and nVidia products than providing an objectively better car OS. Do you see a future where a real Free (or at least Open Source) car operating system is a reality, or do you think the car makers will just continue to create unsafe and unstandardized vehicle UIs indefinitely?
RMS: I don't see the future, so I can't tell you what will happen. I can comment on the problems I know about now in the automotive field, but I can't tell you whether we will win, because that depends on you.
It will be a hard fight to free the software in our cars, but it is essential for drivers -- and not just those that might wish to soup up or customize their cars. The issue affects everyone.
Proprietary software is an injustice in itself, but it also leads to further secondary injustices, such as malicious functionalities. In the case of cars, those can include surveillance and back doors, as well as DRM in the entertainment system.
To exclude those malicious functionalities, the users need to have control over the software. In other words, if you want to have even a chance to make sure that the only back door in your car is the one that lets you reach into the trunk, the software must be free/libre. Anything less is inadequate.
The question asks whether open source software might be almost as good as free software. The main difference between open source and free is in the values they are based on: free software raises the issue as a matter of right or wrong, while open source studiously avoids saying that. However, what's relevant to this question is the practical extensions of the two criteria. Those are _almost_ equivalent; nearly all programs that are open source are free software.
Source code that is open source but not free is rare. On GNU/Linux you will probably never encounter any. In a car, however, you really may find programs that are open source but not free. The main case of nonfree open source programs today is when you can change the source but you can't change the executable.
How is that possible? In such cases, the source is released under a free license; it is free software, and it is open source. You can change this source, but that doesn't do you much good, because you can't run your changed version. The executable comes signed by the manufacturer, and the processor it runs in is designed to reject any executable not signed. (We call such processors "tyrants".)
In the cases I know of, this program is a version of Linux, and the reason they can make its executable nonfree is that Linux is distributed under GNU GPL version 2. If it were under GPL version 3, the seller would be required to give you the signature key to sign executables for your car.
Android uses Linux (but not GNU; the only thing in common between the Android system and the GNU/Linux system is the kernel, Linux). If Android is used in a car, its executable is very likely to be made nonfree in this way.
Of course, tyrant processors can contain software whose source code is nonfree, even secret, and this too occurs in cars. However, those programs are not open source either, so they are not a difference between free software and open source.
What this shows is that we must insist that car software be free/libre; open source is not good enough. It is not enough to be allowed to play ineffectively with source code.
See here for more explanation of the difference between free software and open source. See Evgeny Morozov's article on the same point.
Projects not being done
Ignoring preference of [free software] license for a minute, the [free software] landscape has lots of software to satisfy a wide range of users. What piece of software is still sorely missing from the [free software] landscape that isn't yet being seriously attempted by any project? Short version; what [free software] projects still need to be started?
RMS: The most important missing programs are firmware for various peripheral devices, to replace the "binary blobs" found in the vanilla versions of Linux. Linux-libre deletes the blobs, and all the free GNU/Linux distros use deblobbed versions of Linux; that gets us a totally free system but it can't operate those peripherals.
It is also important to develop Gnash enough to handle the current version of Flash. People like to imagine that Flash is dead, but reports of its death are premature.
Look here for other things we would really like people to do.
free software into law?
You argue that it is unethical for someone to distribute software in a way that limits any one of the 4 freedoms to users. If you had the option, would you make it illegal to do so? In other words, if you had the option would you make it so that software developers were forced by law to use a free software license? or would you leave the option to the developers and try to convince them (without coercion) that it is the right thing to do?
RMS: In an ideal world, there would be no nonfree software. I think it is possible to get pretty close to that. But I don't propose to make nonfree software illegal under today's circumstances, because it is a leap too far; the public is not ready for it. Most users do not think that nonfree software is an injustice. A law that does not have public support is going to meet resistance.
What I advocate, for today, is to ban some egregious practices found in many proprietary programs, including digital restrictions management (see DefectiveByDesign.org), censorship of applications (jails) or works that can be viewed, or requiring code be signed with a key the user does not have (as in Restricted Boot; see fsf.org/campaigns).
Of course, there are other measures governments should adopt to recover computational sovereignty and lead society towards freedom.
We should also ban the practice of asking users of digital works to agree to contracts (EULAs) that give them less rights than copyright law allows to users.