The case

Banshee’s Amazon MP3 store plugin was developed by Banshee star-developer Aaron Bockover, who announced on his blog last August that all revenue of the plugin would go to the GNOME Foundation. The plugin consists of two separate extensions, one for integrating music importing from Amazon’s MP3 store into Banshee, the other for embedding the store’s website. Both are open source, and available from Banshee’s GIT branch.

After discussions between Canonical and the Banshee developers, Jono Bacon announced on his blog that the final settlement was that Canonical would receive 75% of the revenue of both music stores, and direct 25% to the GNOME Foundation. Some people were outraged by Canonical taking such a large share of the revenue, arguing that the company was simply profiting from the work of others.

When are you allowed to sell?

I want to investigate this issue by going from the bottom up. Let us first establish why we pay money. We can’t do everything ourselves, because we don’t have infinite time and skills. Therefore we use the services of others, and pay them in exchange for what they produce. That money allows them to buy the products of others, so they can focus fully on their job. Money is thus awarded for a service.

In open source, most of the time you will not have to pay for the software. However, the GPL license does not prohibit selling your software. The Free Software Foundation defines free software not as ‘gratis’ software, but says software is free when a user is free to run the program, change the program, and redistribute the program with or without changes. (Read its piece on selling (free) software if you want to know more.)

When are you entitled to sell?

You pay money in exchange for a service. In the case of the Banshee Amazon MP3 plugin, Amazon gives a share of the revenue to Banshee, as a reward for bringing users to its store. Banshee subsequently chooses to give the revenue to the GNOME Foundation. Note that it is not the end-user who is the customer here, but Amazon!

Under the current plans, the Banshee Amazon MP3 plugin on Ubuntu will give Canonical 75% of the money paid by Amazon and the revenue of the Ubuntu One Music Store. The GNOME Foundation, via Banshee, will get 25% of both. I shall focus on the Amazon MP3 plugin. There are two ways to look at this. The first way is to consider Banshee an involuntary customer of Canonical, buying the service ‘broader access to customers’. The win for them is more income. The second way is to consider Amazon a customer of both Banshee and Canonical, who jointly provide the service Amazon pays for.

How does this happen? The Banshee Amazon MP3 plugin, developed by the Banshee project, is the direct means used to make the Amazon MP3 Store available. Other important factors are the attractiveness of Banshee—courtesy of its developers—and distribution via Ubuntu, the most popular Linux distribution on the desktop.

We have seen that both Canonical and the Banshee project deliver a part of the service that Amazon pays for. Canonical is the final distributor, bringing the product to the customers’ doorsteps, Banshee can be compared to a more specialised producer, providing a specific product to the distributor. If we look at the real world, we can see that it is often the distributor at the end of the chain that determines the prices. Farmers, for example, earn often very little for their crops. Most of the revenue on produce goes to the supermarkets that distribute the goods to the customers. Supermarkets may not be the sole method of reaching customers, but they are by far the most important channel; the farmer depends on the supermarkets. This simple fact allows the stores to dictate the prices. It is an economic law that says that when a good—in this case access to the customer—is scarce, the costs will go up. Here it means the costs for the farmer will go up in the way of lower revenues.

Canonical can be compared to the ‘Superunie’, the joint procurement organisation of the major Dutch supermarkets. Like supermarkets, it doesn’t actually make everything it offers itself. Instead, it is responsible for the selection, integration and fine-tuning of the components, and maybe for baking the fresh baguettes. Its large market share in the Linux desktop world gives it a lot of power. Some people are principally opposed to it and say it abuses its power.

Access to many potential customers makes Canonical’s contribution to the ‘service’ provided to Amazon much, much more important. It is very likely that 25% of the Banshee Amazon MP3 plugin’s revenue when enabled by default will be higher, than 100% of the same plugin disabled by default. The service of enabling the plugin by default is therefore a valuable ‘product’, which is sold to the Banshee project at a not unsubstantial price.

This high price can be justified by the fact that Canonical is selling a scarce good to the Banshee project. However, Banshee has little choice than to accept whatever benevolent offer Canonical deigns to make. Because they’ve chosen for a free license, there is no real transaction to be made. If Canonical doesn’t like what Banshee demands, then it can just replace the affiliate code and keep everything for itself. Banshee is powerless. That is the difference with the farmer-supermarket analogy, in which the farmer can decide to reject and offer and not give his or her produce.

So, what  amount can you ask for this substantial additional value? It is impossible to determine the true economic price of it when only one side can make demands. The 75:25 ratio is therefore not a representation of the true values of what both sides have to offer, but instead the representation of what the only party with any power over the matter considers the values to be. It is a subjective determination.

Whether or not you agree with the chosen ratio depends what value you attribute to the services provided by Canonical and by the Banshee project to Amazon. It is not possible to do this fully objective, and in any case you need extra data to say something definitive.

To me the demands from Canonical don’t seem very unreasonable at all. The value of the huge user share Ubuntu has to offer seems to be worth the 75% slice at first glance. However, we’ll first have to see the statistics from the Amazon MP3 plugin in action on Ubuntu to verify this assumption. If it turns out that Ubuntu brings in a lot of revenue, then the 75% fee is justified. If it turns out that the revenue is relatively low, or average, then Canonical’s share should be lowered to compensate for the proven lower value of the ‘service’ offered by the company. I would propose to do this check not too long after the launch of Ubuntu 11.04, make the results public and swiftly announce change when change is justified.

What do you think? Do you agree with my conclusion? Did you spot any mistake? Please leave a comment!

As we’re welcoming these new people to the Ubuntu community and introducing them to the various tasks there are, we can see a set of problems arise. There is the problem of increasing scale—community manager Jono Bacon has written about handling this on his blog several times—and the issues introduced by the increasing diversity of the community. I’ve written before about the effects of language differences, but there is something else we should realise, which shows itself more and more now the composition community diversifies.

Not everyone is equally intelligent. When considering other people, it is very hard not to assume they think like us, and judge from that viewpoint. Especially when you are intelligent it may be very hard to realise there are also a lot of people who have more trouble thinking. Life is not like Sims, where everyone has an equal amount of character points, only distributed across different character traits. With the knowledge of the human we have nowadays we can only conclude that ‘life’ is not fair, because not everyone gets the same amount of character points. Some people are smart and beautiful and nice and happy, and other people are stupid, ugly, unpleasant, depressed. Both situations are rare, but both can occur.

This is an important thing to consider in a community that seems to assume intelligence. Every contributor is welcome, which is a good thing, because we don’t want to shut someone out. Every contribution helps us. We can see this in the mentor programmes too. If you create a page at the Ubuntu Wiki and sign the Code of Conduct, you’ll be assigned a mentor. You do not have to sit an exam. At the Ubuntu Developer Summit we are discussing the future of Ubuntu almost as equals, and there are a lot of people that have something interesting to say.

However, what we are doing is not easy work. Building an operating system, writing applications to run on it, running a community, these are all tasks that require skills. Skills that not everyone has. We do, apparently. That makes us more intelligent than the average person, because that is how hard it is to do what we do.

Not all prospective bug triagers, MOTU applicants or passionate artists have got what it takes to build something like Ubuntu. A bad bug triager only causes more work, an unskilled MOTU ruins our credibility, ugly artwork scares users away. Even when the triager, packager, artist is a very kind person, we should say that.

I do not doubt that the community members are honest enough to tell people when they are not delivering good work. But that isn’t the only side to this. Say you are telling someone (s)he’s triaging bugs badly, you’ll first have to have seen badly triaged bugs. That means someone will have to correct those bugs, if the reporter hasn’t already stopped responding. It could also mean that someone has spent some time trying to mentor the bad triager. Valuable time of a volunteer has gone to waste. That is also bad for the motivation of the volunteer.

We should not make intelligence a requirement for joining the community, we should not look down on people because they happen to be less intelligent, we should not become an elitist club. We should, though, make sure we watch out that we do get the right people at the right place. We should realise that not everyone has something to say that is worth listening to, though you cannot make that judgement until you’ve heard what the person said. We should prevent frustration for volunteers and prospective volunteers by being clear about what it takes to join the community.

We should not assume intelligence.

Do I want to say something is wrong with the Ubuntu community? No. I do not think that there are things we need to change to fix the issue I just described. Yet. We should be watchful, considerate and aware that not everyone is equal. The sentence “Every man was created equal.” is wrong, because men where not created, just like women, and they are not equal as well. Everyone is unique, in a positive or negative sense.

The Windows version was released months ago

It happens more often that the promised Linux support isn’t as good as expected, has fewer functions or gets less attention from the developers. A good example is Skype, its Linux client is nowhere as good as the Windows version.

I can already predict the responses of people saying it’s just logical, because Linux just has a much smaller marketshare. However, this doesn’t mean that you should pay less attention to an OS if you say that you do support it. I don’t say that Chrome didn’t do this right, but if you decide to spend less effort on the Linux version, please state it very clearly at your website so we know we should look for an alternative program that does value all its users.

Apart from the program installed on Linux, the OS itself also has got support problems. Some netbooks that are sold with Linux have less functionality than the Windows versions, or have bad drivers. This is a bad case and probably partially the reason for of MSI Wind netbooks with Linux being more often returned than the Windows version, which I wrote a blog about earlier.

It seems like the companies and users have a higher expectation of Linux and start with optimism. However, after a while the users get disappointed with Linux, because the hardware they’re using isn’t fully supported by the manufacturer that installed Linux on it itself or because it has less features than the Windows equivalent.
The companies find out that Linux versions aren’t as profitable or popular as Windows versions, simply because the lack of marketshare. They decide to spend less resources on the Linux version resulting in a program that has differences between its different versions. Or they started with a program that already had less functions and thus didn’t became as popular as they expected, which they could have caused themselves by not paying as much attention to it.

What could be done about this? Yes, I do agree that the marketshare should increase in order to make it more interesting for companies to maintain a good Linux version. However, I don’t think that a small marketshare is a good excuse for this. If you decide to support a certain platform, you should do it properly. When you don’t do so, you contribute to a negative image of Linux because people notice that their favourite program isn’t as good at the opensource OS. On top of that, why would you want to serve the whole world? You don’t have to have a huge group of users to make profit or be succesful.

With supporting Linux badly, the companies doing that are at least partially responsable for the reasons that made them support it bad. If they would support it on a good way, there would be another reason to use Linux, increasing the potential userbase for that version of the program.

And supporting Linux can really give you popularity. A good example of this is the MMORPG Regnum Online, which supports both Windows and Linux. Because of the lack of other good and free-to-play MMORPGs, it has become almost the default choice for Linux users that don’t like the overwhelming amount of roleplaying in PlaneShift. Although the group of users still isn’t that exceptional, it’s still attracting a lot of users and binding them to it because of it unique selling point: Linux support.