Open Source Software and Piracy

In Piracy, Adrian Johns describes the history of the idea of intellectual property starting with the invention of the printing press. Far from static concepts, our ideas about intellectual property have changed many times through history. In its conclusion, he speculates that we may be due for another change. Many of our largest industries are primarily based on the sale of intellectual property – making them vulnerable to ip theft, or what is known as piracy. Piracy cannot be eliminated, but new models of sale can help to limit its effect. The software industry in particular has developed several methods of sale that I will discuss in this paper.

The problem that must be solved in selling software is the inherent copy-ability of digital content. In my second paper, I described a class of products I called “social products.” These products are valued because they act as tokens to gain access to a group, not because they have intrinsic value or usefulness. I argued that those products are especially prone to piracy. The marginal cost of each item is small, so a pirate can cheaply make a knock-off or counterfeit, but they can be sold at a high price, because the counterfeit has the same social benefits as the original. Not all pieces of software are social products, but they have many similarities. Both derive most of their value from the intellectual property they contain. In software, this is through copyright or patent. Both have relatively small marginal cost. For software, this is the cost of copying the digital data – basically nil. And, for both, a copy is just as useful as the original. While for social products, this was the social currency it gives, for software, a counterfeit is often identical to the original.

However, the software industry is not blind to this dilemma – how to make a profit selling something with little marginal value. It is clear that software cannot be sold the same way that an intrinsically valuable product can. Still, software like Microsoft Office runs for hundreds of dollars, and the Adobe Creative Suite for thousands. It is one of the fastest growing industries on the United State. There are many interesting ways that the software market has adapted to compensate. They range from very traditional – following current ideas of intellectual property, to very subversive – breaking long-standing ideas about authorship and ownership. The most traditional is the DRM model. Some companies choose to continue to sell copies of the software as if they were physical products, and use DRM to increase the difficulty of copying. This model is not unique to software – some form of copy-control is used by many industries from musicians through Monsanto, but I will focus on software. The second strategy, also somewhat traditional, is to not distribute the software at all, and sell access to servers running the software. This is the software-as-a-service model, or SaaS. More subversively, some companies have decided to give the software away for free – freeware – or pay-what-you-want, but keep the source code for themselves. This breaks some ideas we have of price, but keeps normal views of ownership and authorship. Finally, the most subversive is the open source movement, which provides the software and source-code for free – allowing others to build upon or modify their work.

DRM stands for digital rights management. It is a technical measure that aims to limit copying of a digital work. Most expensive software contains some form of DRM. For example, Windows will only operate in a limited capacity unless it is given a valid license key upon installation. Others require a physical item to operate. For many pieces of software, the original CD is required to be in the drive to operate the software. This prevents casual copying. For example, one customer cannot buy a video game, then let all of their friends install it as well. Only one can play at a time, because there is only one physical disk. For even more expensive software, a dongle is often required. This is a device with a unique id that connects to the computer and provides access to the software. Anteres Vocal Processing – the software used to auto-tune musician’s voices – is one such product. More recently, some programs require an active internet connection to the company’s servers, even for single player modes. A recent, high profile example is SimCity 5. It was claimed that much of the expensive processing of the game was calculated on the company’s servers, necessitating an active connection – basically, they claimed that the game was a software-as-a-service, described below. This connection, of course, would also check that the copy of the game was legitimate. Those claims turned out to be false. The active connection only existed to check that the game was legitimate, and no substantial processing was done by the server. In addition, when the game was released, there was so much traffic on those servers that users had to wait for hours to gain access to play a single player game on their own computers.

DRM software has a critical flaw. Because the code is ultimately running unencrypted on the user’s computer, someone sufficiently skilled can bypass the restrictions, creating an easily copyable version. It’s not clear that unbreakable DRM is even possible. Someone with some skill can create a clone of the original disk that the game will accept. With some more skill, they can simulate a dongle as well. Others will modify the program to skip the checks, removing any need for the physical products. These versions are known as “cracks” or “warez.” Once a cracked version is created, it is basically game-over for the DRM. Casual pirates may be detered, but filesharing sites make it increasingly easy to distribute the freely copyable versions of a piece of software. Cracked software has become another market. Rather than just removing DRM, many cracking groups added their own content as well. “Crack intros” are like the title sequences before a movie. Many cracked games contain these advertisements for the group that created the crack. In many countries, cracked versions of popular software, such as Microsoft Windows, are available in stores in professional-looking packaging. Alongside the Microsoft logo is the logo of the pirate group.

In one sense, DRM software is very traditional. It is sold at a fixed price, and has intrinsic value because of the function it provides and the difficulty of copying it. DRM transforms the software into a commodity that can be sold like most products with intrinsic value. However, it is more complicated than that. It also subverts ownership for the buyer. Even after a company sells a piece of DRM software, they still retain some ownership of it. They are able to dictate to the new owner what they can and cannot do with the software. Honest users are the only ones affected. The customers must satisfy the requirements of the DRM system. They may only be able to install the software on a certain number of machines, or may only be able to operate it while connected to the internet. This dynamic – a product that exerts control over the buyer, is interesting and as far as I can tell, unprecedented.

Crack groups are interesting in their own right. They are certainly pirates, but they have more in common with the book reprinters of the Enlightenment than the buccaneers of the Golden Age of Piracy. They are not simply reproducing works, they are reinterpretting them. Crack intros have become an artform separate from their role in cracked software. The demoscene and keygen music evolved out crack intros. The demoscene is a community built around demos – displays of technical prowess and utilization of scant resources. For example, a famous demo creates an entire 3d video game in less code than an original Nintendo NES game. When a programmer created an intro, they often had only the space used by the original copy-protection code, so they became adept at using it efficiently. A “keygen” program is designed to generate a license key for a program so that the user can unlock it without paying. This is an expensive computational process, so to pass the time, these programs often play music. By incorporating these artforms, a piece of cracked software breaks traditional views of intellectual property ownership – it is both the creative work of the original producer and the crack group – a creative work that is not allowed to exist in our intellectual property world. Their copies are not mere copies. They are creative works in their own right, with their own subculture built around them, just as unauthorized works from the Enlightenment were extremely local: offering translations, rewordings, and new content.

Software-as-a-Service is another option for software companies, especially startups. Without the resources of a large company, they are unable to police the copying of their software. Microsoft alone spends hundreds of millions on enforcement and development of anti-piracy measures. In a SaaS model, the company does not release software at all. Instead, they sell access to the software on their servers. As the name implies, they are no longer just makers of products, but producers of the service economy. Nearly all “cloud” software fits this model. Very little code is run on the user’s computer. The bulk of it is run by the company on their own hardware. Increasingly, even large companies are moving to this model. The latest version of Microsoft Office, Office 365, requires an annual subscription, and provides conveniences such as remote storage and editing from any device. Adobe has also released their new Creative Suite, the Creative Cloud, which also requires a monthly or yearly subscription, and provides similar services. They have side-stepped the problem of selling a resource that can easily be copied. However, it comes with a set of trade-offs. Unlike DRM software, this explicitly acknowlodges that the user does not own the software. They merely have purchased access for a time. But, the company has more responsibility. Rather than just providing DVDs or downloads of the software, they must also do substantial computation for each user. This requires physical hardware. (Companies such as Amazon, with its Amazon Web Service, and related products, have filled this niche by providing a Platform-as-a-Service for Software-as-a-Service companies). Because all of the important computation happens on their servers, it is only useful with an active internet connection. This limits its usefulness – many computers must operate in areas with poor or no internet access. Despite its disadvantages, this model is not new and will not go away soon. A service economy has existed for as long as there are records and is only growing today.

Freeware is the third model. In this, the company embraces the fact that software is usually effectively free. They give away the software. There are different strategies for monetization. RedHat, an operating system used on many computers run by large companies, sells access to support for the software. Recently, they have surpassed one billion dollars in revenue. They are selling a service, but this differs from the SaaS model, because the service they provide is not the software itself. An interesting emerging model in this category is pay-what-you-want. Many recent independent video games follow this model. They provide DRM-free downloads of their video games for any price that the user decides. Depending on the site, they may also show the average price people payed, and provide bonuses for those who exceed it. In this way, they eliminate the problem of buying power differences across the world. While I do not necessarily suggest that Windows move to this model, it would eliminate the market for cheap, counterfeit copies in the developing world, where the cost is too high for many to afford the real thing. The bonuses are interesting, because they are usually digital content as well. A pirate could get the software and bonuses for free if they wanted, but that is not as common. Because the bonus is a reward for supporting the developers, those who obtain it legitimately get the value in the bonus, and the value of knowing that they are helping the developers, an emotional value. This neutralizes some of the market for pirated copies. Freeware turns many ideas about price, value, and ownership on their heads. Even though the software is given away, it still has value. It is still useful. This demonstrates that value and price are not simply related. Experiments with pay-what-you-want also break economic theories of rational self interest. The Nash equilibrium of the pricing game is for everyone to get it for free. However, many choose to pay for it. They are able to cooperate, ensuring that the content producer receives enough money to produce more content in the future.

In the first essay, I described a definition of a pirate – one who gains the authorized form of currency through unauthorized means. Long John Silver, then, is a pirate of money. However, by giving up his life of piracy – giving up the chance for more money – he becomes a different sort of pirate. He becomes a cultural pirate. He found the authorized currency of contentment without acquiring more money. This is piracy in itself. The software world has produced another type of cultural pirate. They don’t break any laws, but these pirates reject ideas of intellectual property that have existed for over a century, and even ideas of authorship that stretch back at least as far as the Enlightenment. This last model is the most interesting, and the most subversive. Open source software not only gives copies of the software, they also give away many rights of authorship. An open source project is released under a license. The license is a declaration from the copyright owner – the original author – giving up some of their rights. For example, the popular 3-clause BSD license allows anyone to make any modification and use of the code, as long as the original authors are credited in the resulting product but their names are not used to endorse or promote the new product. Other licenses are even more permissive. The Creative Commons licenses are applicable to many more creative works. They allow mixing and matching – “some rights reserved” copyrights. The most permissive of these is the CC0 license, a carefully crafted legal document allowing the author to disclaim all rights to the work to the maximum extent allowed by law. (Some jurisdictions make it difficult for authors to give up certain rights). This license effectively places the work in the public domain.

“Copyleft” licenses take this even further. Like the BSD license, they require attribution, but they also require any subsequent versions be placed under the same license. The GNU General Public License is the most common of these. Any software that interacts with GPL code must also be put under the GPL, so that others can use it. This is a formalization of the concept of public good in copyright law. By placing the software under the GPL, a programmer gives up many of their rights, instead assigning their rights to the public as a whole. The public has exclusive rights to author the code. The public has exclusive rights to see the code. The public is the copyright owner, and they will protect their rights. The Linux Kernel is the most commonly used piece of software in the world, and is licensed under the GPL. I own at least 4 devices, all from different manufacturers, that run variants of Linux. While Linus Torvalds, the original author, still has the final say in what is included in the mainline code, an open source project has a much more complicated provenance. The mainline has many well developed “forks.” Anyone can take the original code and make their own version, with the caveat that they must allow others to do the same with theirs. This has led to a huge proliferation of versions. There are over 300 actively maintained Linux distributions. They form a huge web, with lines of influence tracing down branches, across disparate forks, and back “upstream” to older versions. The GPL and other copyleft licenses completely invert common ideas of copyright. Using the mechanisms normally in place to protect the interests of a copyright owner in a creative work, they protect the interests of the public in a creative work. This is a powerful system that is still developing.

The public interest angle does not preclude the developer from making a profit, however. RedHat’s main product is RedHat Enterprise Linux. As I described in the freeware section, they sell support for their software, and have become very successful doing so. They are also the largest corporate contributers to Linux. They did not create Linux originally, but because of the GPL, all of their changes are available for the public to use, and many have been incorporated into other versions of Linux, even moving all the way upstream to the mainline branch. Many smaller projects are dominated by a small team of developers. While the current version of the software is free, they will develop new features through sponsorships. If a company is using the software and would like a new feature, it is often cheapest and easiest for them to pay the developers for it. As a bonus for everyone else, that new feature will become free for anyone to use in the future. So, even giving up all ownership of a work, it is still possible to make a profit.

The current state of the software industry shows that it is possible to make a profit selling digital goods, even when piracy is cheap and easy. None of the models I have described are the best model for all software, and none are going to disappear in he near future. However, there are a few developments that will soon change the nature of these models. The development of DRM has led to the idea of trusted computing. While Microsoft is looking out for their own interests when they release software to check if a copy of Windows is legitimate, they are also ensuring that the copy of Windows does what you expect. DRM demonstrates that the actions of a computer do not always align with the interests of its owner. Viruses and other malware are another consequence of that. In a somewhat big-brother fashion, trusted computing would build DRM into a computer so that neither viruses nor pirated software can run. Already, phones and tablets demonstrate some of these features. In the next few years, this will increase, and it will be interesting to see what the public gains and loses in the process – some safety gained, but some freedom lost. Software-as-a-Service allows the profusion of cloud-based software today. However, expectations of privacy have not yet caught up. For example, most people use a web-based email service. They store all of their past emails on a server owned by a third party. Because laws were written before this became commonplace, all emails older than 6 months are treated as abandoned, meaning that law enforcement can access them without a warrant. Soon, expect laws to change to fit these new norms. Independent game development has exploded recently. Many are using the pay-what-you-want model. It remians to be seen if that is viable in a larger market. Already in the first couple months of 2014, more indie games were created than all of 2013. This will be a critical test of freeware models. The open source and copyleft models have started to penetrate other fields as well. Many books, websites, and other creative works are adopting this model. Johns’ prediction of a radical shift in our intellectual property ideas could come from one of these changes.

Word Work