This page is a set of ideas I formulated while I was trying to study the possibility of bounty for developers.

The discussion around payment led me to read many opinions on the topic, and among which two web links of important texts apparently well known from developers of the free software mouvement (and often cited by them, Jimbo included of course). I am not sure these links are listed anywhere on meta :-) Here they are for non developer scrutiny

These are probably relevant texts to help consider the issue of whether we should somehow pay the developers for their good work or not.


In his essay, Raymond points out that human beings are fundamentally competing for social status, and that this drive for status expresses itself in different ways, largely dependant on the degree of scarcity of survival goods. For example, higher status allow to get more food, or more women or better job.

He mentions three major cultural systems; the first is command hierarchy, where goods are owned, controlled and allocated by one central authority. In such systems Raymond says, status is essentially access to coercive power.

The second culture is the one based on exchange, where the social status is mostly determined by having control of things to use or trade. "Things" might be material, such as a server or money, or might not, such as information or love. Allocation implies to adopt a voluntary behavior of cooperation if only to get to know what the other have, negociate for some goods rather than others, and lead the other to produce the scarce goods you need. The whole system naturally requires scarcity.

Finally, Raymond mentions a third system, relying on what he calls the "gift" culture. This latter does not require scarcity, but abondance; social status is then determined not by what one controls, but by what one gives away.

He then mentions the fact that the open source developers culture is essentially a culture of "gift", where there is no serious goods scarcity (such as disk space, network bandwidth or computing power) and where status is mostly reputation among one's peer. What mostly drive people is the fulfilling satisfaction of personal achievement, and all consequences of reputation (such as persuading people to collaborate with one, or indirectly helping one to get a better job).

I have then been thinking of what was driving us on Wikipedia and other projects, and if it could really subscribe to that "gift" culture.

The satisfaction we all draw from participating to our project may be at several levels. First, it is for all of us the opportunity to share an experience with others, to feel we belong to a group, and though some try to claim the opposite, it is definitly a mean to build some friendly relationships. This is a good reason why many people join the community. How many of us would now choose a lonely evening in front of TV to a hot debate on irc or a good old edit war on Wikipedia ?
For many of us, the project quickly grows to be also a mean of personal achievement, attention and/or recognition from others. Few fail to appreciate getting a reputation of being a knowledgeable and unbaised editor, a thoughtful and caring mediator, a benevolent dictator, a fair and intelligent leader, a skilled and helpful developer…
And finally, the strongest and more fulfilling joy the project can give us, is the opportunity to “try” to reach our potential. We all have the opportunity to do some challenging and meaningful work, while using our own creativity, at our speed, when we have time and energy. I think that giving to individual the means to do this, is about the greatest thing in our project. Basically, as long as one follows some basic rules, one is free to make experiment, to try innovative solutions, to explore new paths, to take his time, to choose not to do anything when he feels like not doing it, to fail or to succeed, all this without a boss telling him what he should do, how he should do it, and when he should do it.


This opportunity to freely give is probably the main asset of the project. But when we “give”, we always receive something in exchange, self satisfaction and reputation at a minimum. Which we can use to get more things.


Our current system is based on the following principle :
Feel free to give, and if people recognise the work is valuable, receive in exchange appreciation and reputation. Keep preciously the appreciation for self-satisfaction, and use reputation on top : to gain some status (e.g. sysop), to get more cooperation (e.g. start a project and get people to join because they trust you), to get external reward (e.g. improve your resume and gain a new job, receive money by paypal, get a TV interview and have your mom proud of you).

Is there truely a "gift" culture ? Well, I do not think so. There is always an exchange done one way or another. The true gift, donated with absolutely no expected return in mind, does not exist imho. Even self-actualization it brings to help people, is a reward, hence an exchange.

Does that matter ? I think not.
What really matter is not whether it is a gift or an exchange, what matters is the certainty of what the nature of the goods exchanged will be. In a "defined exchange" culture, one will exchange money against a server. The nature of the exchange is defined before the exchange naturally occurs. The type of server is known, the amount of money to purchase it is defined as well. One may always try to lower the price, or to get a bit more memory, ultimately, once the exchange is agreed upon, the goods exchanged are known.

An attempt to a "defined exchange" Wikipedia culture is wikimoney. In such a system, the one "giving" usually does not create what he wants to give. He may propose, but he must always listen to requirements set by the other one with who he will make an exchange. The donor consequently lose freedom over what he is doing, and instead places himself in the position of trying to comply with what the other one is expecting from him. When the exchange occurs, he receives a good for having done something according to the other one desire. Some studies suggest that doing a task to get a reward decreases performance level, especially when the performance involves creativity. The performer will tend to quickly do the task as described in the requirements to get the reward rather than explore all options.


In an "undefined exchange" culture, there is a certainty that a gift given away will bring something in exchange, but one does not know what sort of a reward it will be. It may be just personal satisfaction to see other people use the goods. It may be happiness to receive a thank you note. It may be another good given in exchange. It may be a higher status due to peer esteem. It may be money, just coins or a big check. We just do not know in advance. That means we will do the task without being troubled by the potential consequences. The task will stay a mean, rather than an end.


The current system is essentially an "undefined exchange system" but for a few weak attempts at true capitalistic exchanges (wikimoney) and a tiny bit of power system (WikimediaFoundation).

The proposed bounty system will propose a major change in the current system, since it will clearly set up in a system where the one paying is describing what he wants, the developer working while knowing which reward he will get. The reward will be offered to have followed the instructions. Type of change is level 2 on Twelve leverage points, ie changing the mindset out of which the goals, rules, feedback structure arise.

In comparison, money rewards do not necessarily force us to change the system. Awards for the "best developer" is just a thank you new type of good, which only mean "we appreciate what you do". A flat salary, with very few instructions about what the employee tasks are, is another way to satisfy his basic needs, while letting him express all his creativity. In comparison, this type of change is more a level 5 on Twelve leverage points, which focus on stimulating productivity.

All this, naturally does not take into account the potential benefit of orienting development one way rather than another, by paying people for taking care of certain tasks.