I was with you until your final comments. Outside the US, MMOs/online games are moving to a Virtual Asset Purchase model. They are selling you the songs while sitting by the campfire is free (though you don’t get many songs until you pay, to strain your metaphor). Second Life should be the same. Where they have run into trouble is that they do not manage “song creation” / virtual asset creation by their players well so their DRM system is not effective – and has begun to undermine the player-driven economy. Whether it is possible to do so in a system like Second Life is a subject of a different discussion, but “song sales” seem to be growing rapidly as the business model (according to Gamasutra, microtransactions now drive about 50% of the Korean online game industry).
Well, yes. But no, at the same time.
Let’s take a step back and see what types of microtransactions exist, and what they are for.
All the issues of RMT and microtransactions typically get caught up in the question of what a virtual item is. So let’s settle that issue once and for all. A digital item is made up of database entries. It’s bits and bytes living on a server. It may have been made the same folks who operate the server, or it may not have. It may have been uploaded by someone who pays for the privilege of manipulating the data on the server, or it may have been uploaded by someone who gets access for free.
In some cases, these bits and bytes might be a unique arrangement of data, in which case it is probably copyrighted to someone. In some cases, it may actually be a record of activity instead, in which case and under some laws, it might be subject to privacy laws instead.
In no sense are any of these database entries “objects.”
When you play Pangya Golf, you in fact do pay using a microtransaction. You do so in order to increment a database entry labelled “points.” This is the same as when you purchase Microsoft Live points or when you purchase cell phone minutes or when you purchase acorns in CyWorld. From this point forward, there is no longer a real-world economic transaction. In the dominant model of microtransactions, you are actually using the “wallet” model.
Purchases of new outfits for your golfer, new Live Arcade games, or new furniture for your house are no longer truly “microtransactions.” They are the reduction of a number in one field, and the addition of some data to another. It may be data that says you have access to a given other field in a database, or it might be a change in the records that state that you have access to download something — at which point the download system can go ahead and verify that, then initiate the download.
This may seem stupidly obvious, but it’s worth going over because so many people misinterpret the issue of real money trades for digital items. When one person sells another an EverQuest sword, what is being traded is not actually an item (as in a retail supermarket) or even a stand-in for an item (as in markets for precious metals or pork bellies) or even bytes representing a bet on future price trajectories (as in futures markets), but rather moving some bytes from one column to another that have no analogue whatsoever.
In fact, in the case of EverQuest, you are not even moving a copy of the sword’s information. You are only moving a pointer to the sword’s information, because in practice, there is only one sword defined in a template database. Each local copy is an illusion; your inventory says “you have Sword X.” In many of these worlds, you can think of a piece of equipment and a skill as being exactly the same thing. The same is true for that clothing you purchased in Pangya or that chair you bought in Habbo.
Don’t worry, we’ll get to things like Second Life in a little bit. OK, actually, in a long while, because first we’re going to briefly mangle the history of copyright.
the only way such legislation is going to work in the rest of the world is if we begin to understand how intellectual property differs from material goods–and that ain’t gonna be easy.
The reason intellectual property seems like such a foreign concept to many in the Western world is that we come from a culture of ownership. We buy a house, we buy a car, we buy a pair of pants, and these things become ours to do with as we please. But intellectual properties like music, movies, and games don’t work this way. The $15 you spend for a CD or a DVD buys you the packaging and a hunk of plastic, but it doesn’t buy you ownership of the music or movie itself. All you’re paying for is a license to use the content within the guidelines established by the owner of the data. You have no legal right to redistribute that content to anyone who has not purchased a license from the content’s owner.
Where IP can come into play is that the company can assert that the arrangement of bytes in a given database field are a unique creation (for example, they can represent a texture, or a piece of text) that has been fixed into a tangible form. At that moment, it is covered under copyright law and possibly trademark law if it’s a trademarked chunk of data. People trying to make a buck by leveraging the content (since they are not actually selling it, but instead actually selling moving data from one field to another) can run afoul of IP law that way. For example, they might reproduce something copyrighted, like a texture, in order to persuade people to pay for the transfer of bytes from one field to another. But someone engaging in RMT services for, let’s say, powerlevelling someone, isn’t even doing that directly. Keep in mind that the actual transfer of bytes from one field to another is not in itself illegal — it in fact it is the reason why you pay for access to the database in the first place, and RMT would not happen were said transfer not assumed to be a basic part of the functionality of the virtual world.
Copyright laws are in fact a Western sort of invention. They exist because with the advent of the printing press, it was easy for someone to take something someone else had written, copy it and print it up and sell it as their own, and profit in a way that gave no revenues to the original publisher (or the author). Originally, the Stationers’ Company had a monopoly on copyright; whoever entered the title of a publication in the guild’s ledger first was held to have the exclusive right to make printed copies. Later, the Statute of Anne said that in fact it was the author who controlled the rights to make copies, and ever since, authors have negotiated away said permission in order to allow publishers to publish their work.
In the case of the book published back in 1710 when the Statute of Anne went into effect, all the way up through the last book you bought, what copyright controls is who gets to make a buck. The book is a container. Under the Doctrine of First Sale, which was first enshrined in US law back in 1908, you can resell the container (or in fact, sell the container for the first time), and the original author of the copyrighted work has no say whatsoever over what happens to it. There’s little question that the copyrighted work within the container essentially boosts the value of the container; try selling a blank book for the same price as a first edition of Dante’s Inferno. However, you cannot profit by taking the copyrighted work and doing something else with it, say by issuing an audio recording, or creating a new edition of the copyrighted work and printing and selling that instead. The doctrine of first sale is about reselling the original containers, not making new ones.
There’s two big exceptions: software, and phonorecords. The exceptions are about leasing, lending, and renting the container objects, not sale. And these exceptions, frankly, have zero to do with logic and everything to do with who had better lobbyists in Washington DC at the time. (Ever wonder why you can rent movies at Blockbuster, but not audio recordings?) In the muddled case of software, publishers rely on EULAs and shrinkwrap agreements, not copyright.
What does all of this have to do with microtransactions?
Well, Moorgard said of Second Life,
Second Life, in so many ways, is an aberration that cannot continue to exist in its current form. As with Mark Cuban’s point about YouTube, Second Life will eventually have to face the music (literally) for allowing the unlicensed distribution of copyrighted content.
But in many ways, it’s the copyright regime that is an aberration. For one, it’s only a few hundred years old; prior to that, people copied and reproduced with abandon. Today, you can get sued for parodying (and even though you shouldn’t lose this case, sometimes you do); for briefly excerpting; for repurposing; for outright use. What’s more, the length of time for which a given copyright lasts has been hugely extended. The Statute of Anne only granted 14 years for new works.
For another, we’re increasingly entering a world where the notions of “container” are being stretched quite a lot. The stretching started to happen when things like cassette tapes, photocopiers, and cameras came along. All of a sudden, reproductions of copyrighted works could be created by instantly creating a new container for them.
Then there’s the issue of streaming content of various sorts. The courts ruled a long time ago that it was OK for you to record audio off the radio or TV signals onto videotape. The reason is that there are plenty of non-infringing reasons to do so, most specifically time-shifting. In fact, you the consumer pay a premium for blank media of all sorts precisely because of this. It is presumed that the added hidden costs on the blank media cover the losses that the copyright holders incur from people pirating stuff off of “streaming” media. (The copyright holders actually wanted to instead make illegal the sale of methods to record streams, but the courts found against them).
On the Web, everything is streaming. In fact, when the content reaches the computer, it is in fact the first time it has been put into a container for sale. This meant that once you bought something via download, in theory you could have then turned around and sold the disk you downloaded it onto, under the doctrine of first sale. Except that putting it onto a disk in fact is making another copy. In the world of digital data, everything is infinitely, perfectly reproducible. Basically, MPEG encoding or any other form of encoding is just sort of an alphabet, and the plastic disc you have that has MPEGs on it is very much like the paper-and-glue object that carries stuff encoded in a different alphabet. The difference lies in how easy it is to read — some alphabets require special reading tools (like DVD players and screens) and others do not, because our brains are adequate to the task.
This fundamental disjoint with the notions of how copyright and first sale should work is what ultimately bequeathed us the DMCA. In a nutshell, the DMCA says that in this brave new world of easy copying, that it’s illegal to try to get around the technological barriers that publishers may have put into their “containers.” It went further than any previous copyright legislation, and made it part of copyright law that there be anti-copying measures in the decoding tools.
It also, fortunately, provided something called “safe harbor” for all those poor intermediaries who found that because of how digital media work, that they were making copies of everything that passed through their routers and servers, willy-nilly, without even knowing it. You see, when you send something to someone else via the Net, you’re not sending “it,” you are always sending a copy. What’s more, tons of copies get created at intermediate hops all along the way. Once it gets reassembled at the other end, there’s another copy there, and if someone then scrapes that content or archives it, there’s another copy. The DMCA says “well, they didn’t know where it was coming from, so they can’t be blamed.”
Now, finally, we get to Second Life. That “safe harbor” provision is exactly what both YouTube and Second Life rely on, you see. That’s why they don’t check in advance what gets posted; if they were in an editorial position, they’d be liable. By not putting themselves in such a position, they can instead just accept takedown notices as they occur. As long as they follow the letter of the law, they should in theory be OK. Lawsuits against the likes of YouTube arise because copyright holders feel that YouTube is following the letter of the law, but not its intent: that they are knowingly profiting off of a flood of copyrighted content that is illegally uploaded, and their business model depends on it, even though they don’t actually have any one specific piece of copyrighted content in mind.
This gets even more tangled when you realize that much of the Web works by linking as opposed to just uploading. Generally speaking, you can point to anything you like, and because of the way the web works, it’ll show up as part of your content, indistinguishable to the end user. Worse, an end user could link to something and make it look like part of your site… you get the idea.
I mentioned before that when you buy a piece of furniture in Habbo Hotel, you were actually not buying anything. Let’s look at the specific process that we go through:
- An artist creates the particular arrangement of pixels that we see as a chair when parsed as a particular image format. (Keep in mind that it could also be parsed as audio, as braille, as ASCII, or any other of numerous “alphabets”). It is now “fixed” in digital form, in copy #1 on their machine.
- Copyright immediately vests in this artist.
- Except that as a condition of employment, they actually assigned the copyright for their work-related creations to the company.
- The work is probably copied up to several servers: a version control database, a build server, a run-time server, and so on. Copies #2, #3, and #4 are made.
- From there, more copies are made as the item is reflected out to the local databases of all the other developers.
- A user maybe pays the company for the right to connect to the server which has, among other things, a database entry somewhere with the bytes that look like a chair when parsed in a particular way. Or maybe the user doesn’t pay, and they can connect for free. Either way, when they connect, they agree with the company operating the servers that there are some activities they won’t engage in with the database, and that the operator can kick them off if the operator so chooses.
- The user pays money in order to increment a field in the database associated with their account. Said money goes to the company.
- The user voluntarily reduces the value of that field in their account, and in exchange gets a pointer in their account to the database entry that holds the bytes that when assembled correctly, look like a chair.
- At some point, the bytes are fetched from the database, and sent over the network to the user’s client. Many copies are made along the way in this process, but most of them are erased instantly.
- A copy now exists in memory on the user’s computer. If it’s cached, it might also get written to the hard drive (so the user has two copies, at least until they exit the program).
- The “alphabet” is parsed and the bytes are arranged into a graphic that we recognize as a chair when it is displayed as colored dots on a screen.
- The excited user takes a screenshot, and…
Oops. Even though there have been probably a dozen copies made throughout this process, it’s this last one that is suddenly a possible violation of copyright. Even though the user has two or three copies on their machine at that moment (one in the cache and one in main memory and possibly one in video memory) nobody is worried, because “that’s just how it works.” In fact, if you made a copy of your cache as a backup, in case you had to format your hard drive, nobody would really say Boo.
But start uploading that screenshot, and someone might get a takedcown notice under the DMCA. Clip out the pixels that represent that chair, and paste them into another texture and upload it to Second Life, and someone might get a takedown notice under the DMCA.
You see, even though the data stream from the server may be encrypted (and it’s illegal under the DMCA to circumvent that), there’s the equivalent of the “analog hole” — you can always just capture what appears on the screen or plays through the speakers. Once the decoder is our senses, there can be no DRM.
The Habbo scenario and the Second Life scenario (I told you I’d get there eventually) are in fact identical in every way except the issue of who holds the copyright on it. You can trace the entire chain and get the Second Life scenario with the sole change of dropping bullet #3, the one that reads “Except that as a condition of employment, they actually assigned the copyright for their work-related creations to the company.” And that is why Second Life made headlines back when they irst announced this policy back at the first State of Play conference.
Many companies have asserted that the act of uploading grants them either copyright in the work, or a variety of publication rights. In fact, Second Life also has to assert that it grants them some rights, because they are in the business of replicating and distributing this content. They don’t stand in an editorial position, though, because to do so would require that they vet the content for copyright and trademark violations, and that’s expensive. So they take the position that they do not look at the stuff, they just blindly copy it and redistribute it, just like a forum does when you post on it.
Where does this leave microtransactions?
Well, everything I have said till now is based on following the letter of the law. But in practice, here’s something we do know. The law doesn’t work. Copyright, since the days of the Stationers’ Guild, is premised on social consensus — it is completely unenforceable against anyone outside the law or who does not share the same social framework. Some countries have something called a “moral right” to creative content, but like most morals, it’s visible only to those who share the same morality.
Ever since there has been copyright, there have been people copying, because copying is the default mode of learning. We learn songs by listening to them, writing them down, watching a performance, and sharing tab on illegal guitar tab websites. We learn to paint by copying the Old Masters, and we learn to write by imitating other authors. In many of these cases, the copying is subtle; in others it is blatant. In the cases of fan fiction, we leverage so much of the original creation that there’s little doubt we are co-opting someone else’s creation, without which ours would not exist. The line between a legitimate use and an illegitimate one is often questionable — see the recent case of the sequel to Peter Pan, for example.
In particular, since anything that can be seen by our senses can be reproduced, for better or worse, all digital forms of enforcing copyright are doomed to fail. Every form of encryption is moot, because everything must be decrypted in order for us to see it. At some point, the data is in the clear, and then it can be copied.
What’s more, the dominant mode of discourse on the Internet is no longer the central-server model that serves the microtransaction systems so well. Today, end users of remote Internet services upload as much data as they download. This is true even in the case of the closed service MMORPGs that make their living by effectively broadcasting the state of their database to you. You chat, and they log it, storing gigabytes worth of data. You move, and you do things, and they track a history of frequently highly private information. And you write reviews to upload to Amazon, you upload graphics in Second Life, and you post stuff on this blog — more data, in fact, in comments than there is in the posts.
In the end, microtransactions are an ancillary business model because there are copies of everything everywhere. Microtransactions from the server side depend on the notion of monopoly. The artist at Habbo or EverQuest is ultimately able to sell you access to that image solely because you cannot get it any other way. But other ways are increasingly available. In order to prevent you from getting it, they have to impose forms of DRM. They have to tell you “no, you cannot upload images.” They have to tell you “no, you cannot transfer access to these bytes to another user.” But all of these depend on the user not simply choosing to go somewhere else. After all, were there a completely open uploading platform like say, Second Life, they could simply take all the screenshots of Habbo furniture and upload them as textures there instead, and thumb their noses at you.
We see this in operation on the Web, where easily copied content has led to the result that pretty much everything on the Web is freely available all over the place. The only thing stopping you are scruples, savvy, and the vigilance of the copyright holder going after every place which hosts the content, sending them DMCA notices to make them take the stuff down. Scruples have a way of bending. Savvy has a way of always increasing. And the latter mechanism, IP-holder enforcement, only works as long as the host site shares the same social context (try enforcing copyright into North Korea, for example).
The reason iTunes can make money is not the content, it’s the service: it makes music easier to find, with a slick interface and useful databases. In terms of actual end-user utility, it’s far worse than illegally downloading stuff off of a P2P service. After all, it imposes all kinds of license-based restrictions on where and how you can listen to that music, and its database is far smaller than the extraordinary array of stuff available on the Web.
In the end, these are the truisms:
- RMT is not about IP. IP is just the method that companies use to maintain monopoly access to stuff in their DB.
- Microtransactions work as long as providers have monopolistic or cartel-level control over the data.
- In a digital world, it is impossible to maintain said control. Everything can be copied.
- Copyright only functions as long as everyone agrees to it.
- “Broadcast” unidirectional services are going, going, gone.
- Providers can still make quite a lot of money even if they give up on the notion of control, if they instead embrace the idea that what people really want access to isn’t the bytes that look like a chair, it’s the service.
In the end, what people want isn’t the picture of the chair, or the picture of the sword. It’s the value of having a chair in your apartment in Habbo Hotel, the value of having a badass sword in EverQuest because you can kill more stuff. The actual IP bit is mostly useless outside of that context.
To say that Second Life (or indeed any world which permits uploads of user content) is an “aberration” or doing things “wrong” because they aren’t wedded to the broadcast model and the notion of DRM is silly. In fact, saying that SL didn’t do enough in terms of implementing DRM is silly — it actually has MORE of it than Habbo does to protect its own assets!
If anything, it’s the market in Korea not quite catching up to the full implications of the bi-directional Web. It’s the misapprehension that this is at core about copyright. What is really at stake is the question of how much permission users have to reach into the database and twiddle the bits, and whether the results of that twiddling can be considered a service provided to other users in exchange for payment. It’s whether it’s fair game to make use of, resell, or otherwise appropriate content once it’s tied to a container that you own. In the end, the debate is actually about whether the screen belongs to you, or to the person sending the bits down the wire.
The value is in the service, not the content. In the service, not the microtransactions. A digital item is worth nothing. What is of value is the context. People are increasingly not willing to pay for the experience of hearing a song by itself in the abstract. They pay for the concert as a whole (the iTunes experience as a whole, the CD experience as a whole, the movie-going experience as a whole, the EverQuest experience as a whole), and it will be smart venue operators who survive and make the money in the long run.
Edit: two things I didn’t cover (because this was damn long already)… One, trademark is a different beastie than copyright, and there are several legit ways in which trademark protection could be used as an argument against RMT. And secondly, not all data is of course displayed on the client end and therefore will be available for copying. Much server-side value can be gotten from stuff that is simply never displayed. The mistake is to rely on copyable stuff for revenue.