[Foundation-l] Bounties and expenses

[Erik Moeller]

Hi,

I know that I'm still technically on wikivacation, but I'm beginning to  
feel full of energy again and want to briefly share my thoughts on the  
matter of bounties and expenses, which has been hotly debated in the last  
few days.

First of all, I think that Angela should absolutely go to Paris on the  
foundation budget. This is not an entertainment trip. This is the first  
opportunity for Jimbo, Anthere and Angela to sit down together and discuss  
what will happen in the next few months.

Angela went to Berlin from her own money to observe the creation of the  
German chapter (with me trying to give her a real-time translation of what  
was going on), and in Paris she would have a similar opportunity to talk  
to French Wikimedians about the French chapter, both sharing her own  
experiences and getting new input. This is exactly what board members  
should do.

Face-to-face meetings are much more productive than IRC simply because  
real human interaction has a much higher bandwidth than letters in a  
window on a computer screen. That was obvious at the WOS in Berlin, as I  
know that some people are still reeling from the whole experience. ;-)  
Getting Angela to mingle with the French Wikimedia community will benefit  
the whole project.

Of course I understand the general objections, but calling foundation- 
related travel expenses "perks" is ridiculous and offensive. Anthere and  
Angela are giving a large part of their personal lives to this project at  
the expense of their career and family. Asking them to either fund their  
own travel expenses or stay at home undermines the whole purpose of the  
board, which is to keep an eye on the development of *all* of Wikimedia,  
and for that it is absolutely necessary to actually meet with real human  
beings.

It would be a shame if we returned to our usual anglocentric way of doing  
things while Jimmy is still on his missionary trip through Europe.  
Wikimedia is an international organization, and real world board meetings  
should take place in different locations to give Wikimedians from the  
whole world an opportunity to directly talk to the trustees they voted  
for. Now just because these experiences might actually be enjoyable to the  
trustees doesn't mean that they aren't valuable to Wikimedia as a whole as  
well.

I do believe that funds which were not explicitly designated for the  
purpose of funding foundation organizational activity should not be  
permanently used for said purpose. So what we should do is clarify on the  
donations page how much of the money is going to be used for which  
purpose.

But we will soon get a check over 10,000 euros from the Prix Ars  
Electronica award, and that money can be designated by the trustees for  
various purposes, and a certain amount of it (say 2000 euros) should  
certainly be designated for organization expenses. Temporarily withdrawing  
the necessary funds from a non-designated pool until we get the check is  
not a serious issue.

The key here is that everything is transparent and open. The fact that we  
are seriously debating whether we should give one of our trustees 400  
bucks or so to meet with the elected board and discuss the creation of a  
new chapter shows quite well that we are already much, much more open than  
virtually every other organization of the same type.

Now, designating money for org. expenses does not preclude us from doing  
the same for development expenses. As some of you know, the creation of a  
development bounty system was a core part of my election platform. From  
communicating with Jimbo and Angela I got the impression that they share  
the belief that selectively funding specific tasks would be a good idea. I  
don't know where Anthere stands on the issue.

Again, we could use a certain amount of money from the Prix Ars  
Electronica funds for a first test drive (I'd suggest $2000). If it turns  
out that such a bounty system does more harm than good, we can always stop  
doing it. It is unlikely that a single experiment will have devastating  
effects, but it is quite possible that it will lead the way toward a  
complementary development process.

The key question is how to define priority tasks. Because the developers  
are the benefactors of such a system it is somewhat dangerous to let them  
alone make the decision, even if that is done through voting. On the other  
hand, non-developers often do not have the understanding necessary to make  
these decisions.

I do not yet have a final answer to this question. For the experiment  
phase, I think appointing one developer and one technically-minded non- 
developer who have to reach consensus would be a simple solution. I would  
like to nominate Tim Starling and Daniel Mayer for these two roles. If Tim  
doesn't want to do it, I would suggest Jens Frank, who has already said  
that he wants to leave bounty tasks to others, so he would have a certain  
level of objectivity.

(I'm not nominating myself because I would consider participating in the  
bounty system, and because I think I'm a little too biased in favor of  
certain tasks to be an objective judge.)

Essentially, these two people would be in charge of evaluating "grant  
proposals", which could be made by anyone (developers or users). In the  
long term, I believe it would make sense to replace them with an appointed  
or elected committee, which would have to include at least one leading  
developer representative with the power to veto certain proposals (for  
being infeasible, impractical, incompatible etc.).

But again, we should experiment with different approaches.

The amount of money for each task should of course be related to its  
complexity, and be decided by the bounty managers. The bounty would be  
paid if the developers and the bounty managers agree that the task has  
been completed.

Regarding Erik Zachte's remark that $100 is not a lot of money, that is of  
course correct. However, the purpose of this system is not so much to give  
participating developers a salary, but to provide a little extra incentive  
for completing tasks which we all agree need to be done, but which have  
been largely ignored for months. We can always raise the bounties if it  
turns out that they are ineffective. $15/hour seems like a reasonable  
starting value. I know Java programmers who work for less than that.

To preempt the inevitable comment that we don't do the same thing for  
articles or wikibooks, that's true, but that doesn't mean that we never  
will. If this bounty system works, it is quite possible that we will try a  
similar approach to fill important gaps in the various Wikimedia projects.  
As always, it is of key importance that any such process is open and  
transparent, and that all funds which are used for this purpose have been  
designated for it from the start.

With all this talk about expenses, we have to keep in mind that Wikimedia  
will quite possibly be an organization with a multi-million-dollar budget  
in just a couple of years. I am very confident that we will be able to  
raise $100K or more through grassroots donations this year. Like the  
energy of our content contributors flows into many different areas, the  
money which is given to Wikimedia should flow wherever it can be usefully  
and productively spent.

A lot of our money will be going into hardware purchases for quite some  
time, but it would be irresponsible not to carefully consider and explore  
other ways in which money can help along our mission of educating  
humanity.

All best,

Erik