Hacker News Reader

People who previously had no problem working on the project without direct compensation will now feel undervalued if they don't get shares, but other people do.

So it could work for new projects?

I think you'd need to define the value of contributions before you ever begin. Otherwise, time will be spent bickering of the value of work, and not actually doing the work.

True. Also, having quantified the value of each contribution might actually create a disincentive for joining the project: that 5-line patch that was going to be my very first contribution is now officially worth nothing, so why submit it? By laying a focus on the top contributors instead of the new/potential ones, this undermines the crucial ability of F/OSS projects to motivate people to _gradually_ extend their involvement.

Yeh sold. :) The embryo for this grew out of wanting to try and make those 5 patch contributions worth something tangible and to encourage them... sadly it seems the idea just got lost along the way and does the opposite.

Back to the drawing board :D

Maybe you could weight contributions differently, depending on what kind of behavior you want to encourage? For example you could make first contributions worth double, or have first contributions gain a bonus of x shares, which they can cash in/withdraw after their 5th/nth contribution, or something?

That encourages new-account spam. Maybe you could make them wait, like a "signing bonus", for a month? or until their third contribution?

Yes, that's what I suggested, actually.

You might not have to go all the way back...

I've thought of this concept before, and I think the key is to measure based on contributions to functional goals of the project, not amount of code added.

For example, participating in a discussion on requirements for a specific function is worth X points. Submitting the code is worth X points, too. Documentation is worth X points, as is communicating with the end users who requested the function or found the bug.

In short, reward people based on participating in a collaborative development process. Then that 5-line patch is more highly rewarded by someone who takes ownership and truly participates, not just a random coder with time to spare.

Might this discourage some people? Yes... but only people who wouldn't have time or inclination to be significant participants in the project anyway.

In the long run, the team and project should be stronger.

At least, I thi/hope so...

The implication that F/OSS donations should go to charity might not sit too well with a lot of people. They're developing software in their own time and giving it away for free - why do they now have to channel money to charity to improve their 'image'? Having some donations every now and then might be a nice thing, especially if you're the Mozilla foundation.

The same thing might even happen to the user base: the software is still free, but now you get a guilt thing for not having donated to Oxfam.

Weird psychology at work here, I'd guess.

The charity inspiration comes from 3 main things. Firstly that it makes the money distribution easier and would hopefully cut down a bit on arguments over who gets what. Secondly a lot of FOSS projects do support a charity with donations - we might not be talking Oxfam here, what about software related ones, like a Linux PC for every child? etc. Thirdly it means people know where the donations go - rather than to all sorts of contributors it is going to a recognisable charity :)

That said perhaps there are other avenues: the main point is to not give cash out to contributors because that defeats the object somewhat (in my eyes :D). For example it could go into some sort of software related fund (perhaps like a google summer of code type fund) etc.

Here is an idea for you, I won't go into criticizing your idea. If I knew what will work out, or for that matter ANY of the people here, we would be doing it. We can spend hours and hours talking about the idea, telling you the problems why it wouldn't work, but still be way off course of the reality. Recently I spoke with my ex boss who has been in IT for so long, now retired and he said, quoted "If someone asked me back then if Facebook would have worked out, I'd say no chance."

I think, we entrepreneurs shape the world around us to tune into new ideas. Persistence and continuous mutation/evolution showed to work for life, why not business too? Whenever someone asks a question here on HN, you get 90% of critiques and ways of discouraging. I say screw them. The spirit of real hackers and entrepreneurs is to just do it, be perseverant and have no fear of failing, but if it is to happen, fail fast and learn from it for the next iteration. No other can influence the success or failure than the entrepreneur himself, so I think you should hear the tone of only those that give you suggestions which way to go. I am sure you know what the problems are.

My suggestion. Build a prototype, use something that you can build it very quickly e.g. Rails and then hear out what the people are saying about it. Then modify it day by day from what you hear, and I guarantee you that you will eventually end up with something great. Your idea might be brilliant or stupid, but it is not on us to judge. Also, when you get discouraged, remember it is not on you either to judge.

I personally believe that is indeed needed and a great thing and has been going through my mind by reading a bunch of stuff. If I can somehow help, I'm in. I recommend you read this, that you would help solve:

http://www.strategy-business.com/press/enewsarticle/enews053...

Cheers!

Dejan/Aleveo.com

I'm one of the founders of FairSoftware and I have been thinking about the problem you describe for 3 years now...

You are on the right track, but I would suggest some tweaks.

1. Why does FOSS have to be given away for free? The freedom that RMS talks about is in accessing the source code and improving it. In other words, if you made donations to FOSS mandatory, you'd still have "libre" software, but not "free" (beer) anymore. Is that too shocking a concept?

This could make the amount of money flowing in significant enough that developers could start abandoning their day job and actually work full-time on FOSS. Today, if you look at who significantly contributes to FOSS, apart from a few exceptions, they are all somehow on the payroll of IBM, Sun, Intel or Google. That's not freedom.

2. In our implementation of FairSoftware, we used the paradigm of a corporation, where everyone is a shareholder and decides on future share allocation. To match better a FOSS implementation, I would recommend some sort of karma, where everyone can upvote contributions they like. In large, unstructured groups, it might work better. If I like what you do, I give you a +1. Think of it like HackerNews voting, if you will. This is somewhat subject to gaming, but nothing that can't be fixed.

It would provide a model for allocating the shares that is more in line, I think, with the FOSS community.

Finally, I disagree with the statements that if you start compensating people for contributions to FOSS in any way, you will pollute the well. If implemented improperly, it's true that you can alienate the volunteers. That's why I think the flexible karma approach would fit much better. You can receive karma without having to ask. A good leader will make sure that newcomers are welcome and give them some thank-you points as soon as they help.

What happens to shares when people stop contributing? Are they recycled? If not then how do you deal with the possible eventuality that there will be no more shares to award? Printing more on the other hand means this would be a currency, not shares.

On a related note the Capitol Hill Baby-sitting Co-op paper is well worth reading. This has been posted before, but in case you missed it: http://ideas.repec.org/p/kud/kuiedp/0335.html

Sounds complicated (transaction costs!) and people don't donate much to open source projects in any case.

I'll add: how to fund open source projects is a Difficult Problem. A while ago, there was a great mailing list called FSB (free software business), with a lot of illustrious folks. They never really came to any definite conclusions other than "it's difficult".

I was thinking about that.

Transaction costs (I assume you mean payout to the charity) weould be easy because you just store up the charity donations from everywhere till it reaches a reasonable value ($15 perhaps, or more) then payout costs are minimised.

As to donations - I had a big discussion about that last night. In the end we figured 3 things. Firstly collecting donations is a pain to handle so a lot of people dont bother (especially as there is little worth to them). Secondly there is no culture to donate even though many people probably would if it was easy and they could do it super cheap: a known central service to handle donations wold be recognisable and trusted. Thirdly as a portion of the beneficiary would go to charities and you could actually see that happening it might encourage more people to donate.

In terms of complexity: I think it just sounds more complex than it is. Really it IS just Twollars with a bit more on top.

Does that seem like a reasonable assessment?

If you're at all familiar with how shares are traditionally divided in a business, then you'll realize this is no trivial matter. If you want to think of an individual operating agreement or charter as an "implementation," then the complex negotiations, armies of lawyers, rounds of iterations, and continuing amendments to these documents are a lot like a huge enterprise app.

What happens when we decide, like businesses, that we need new classes of shares? Who gets paid first? Who has "voting rights"? What's the process to to pursue remedies if you feel you've been cheated, or that the "board" (project owner and his lackeys) has operated in bad faith to defraud share holders?

This is a complex issue even before you hit the psychological problems that others have already raised.

What I'm saying is that shares can't really divided in a trivial, algorithmic way. I like the idea in theory, but I think it's impossible to correctly implement.

Actually I wasnt thinking anything like as complicated :)

The divison would be based on awarded "shares" - so we're not talking stock exchange type shares (dont worry, I play the markets :) Im genned on realistic shares) but "share of the donation"

After I posted the original message I realised it was misworded. The project doesnt get X shares to divvy up. Rather share count is "infinite" and the project owner hands them out (either by automated means or manually) to others. So if 10 people contribute and each gets 1 share they recieve 1/10th of the donation pot (after costs etc.).

If the owner wants a "share" they must give themselves some ;)

What you were thinking and what the system would eventually become are two different things. I'm saying the inevitable consequence of a system like that would be the scenario I described.

Do you think company stocks started out that complex? No, they started out as an intuitive way to divide the profits of a business operated by multiple people, just like your intuitive way of dividing profits from a project built by multiple people.

Edge cases presented themselves, conflicts arose, opportunities for individuals to maximize profit came to light, and thus was our current system grown.

ok fair points; I assumed you were pointing out flawqs based on the model I mis-represented.

Im not 100% sure I agree the problem would be as dramatic/bad as your presenting though.

If the money isn't the motivator (gets donated anyway) you could use a virtual currency. "karma" or something similar. You might monitor svn/git activity and create a community where people could rate each others commits and give each other some "karma" that way (hacker news with code submits instead of news, sort of).

This proposal is not too dissimilar to the ideas behind P-BOS (http://lmframework.com/page.php?id=pbos)

P-BOS works best when the modules are small, and hence breaking down individual contributions within a given module does not get too messy.

As to the 'morality' of charging for FOSS, the issue can easily be sidestepped by giving developers complete control of the licensing and payment terms for their modules.

For example, a given module (component) could be free for everyone, or just non-commercial users. Developers should also be similarly free to route a certain percentage (0-100) to charity. The key point is to leave all the key decisions to individual developers.

If anybody wants to know more about P-BOS, feel free to drop me a line: david-lmframework-com

I just wrote some code that the community recognizes as useful and that thousands of businesses incorporate into their product or process. Thanks for the twollar.

Joel Spolsky talks about this over and over on the Stack Overflow podcast. People who are willing to do something for free !== People who are willing to do something for small bits of money or quantified reward.

It's possible that there is incentive to be had in the 'common goal'; "The Foo Project's Charity is the FSF; If we squash all the bugs by Labor Day, 80% of donations will go to the FSF. Otherwise, 50%"

I happen to disagree with Joel on that one - in particular circumstances :) also the aim might be to encourage those who tend not to do "something for nothing" to contribute :)

(The common goal thing is quite good....)

I think - like most ideas based around rewards the value will be seen in it's worth.

As a silly example; 1000 Twitter followers would not have been valuable in 2006, but in 2009 it holds a value. It's the same with HN ratings.

An issue there is that new users will feel "at the bottom of the pile", and possibly less likely to "rise up the ranks". At least once a project gets going.

Once you put a dollar amount on work you have done, and then hand it over to someone else you run into a can of worms regarding taxes. Uncle Sam will want a share, and will eventually get it.

hence the donations model :) the contributor gets no cash-in-hand, so taxes aren't a problem (just our taxes for donating to charities etc.)

may be you can use ohloh.net to see individual contributions on a project and derive some calculations based on that. or may be you can even link the "donor concept" into it. ?

Seems like donations could be put to better use than just being tunneled into another organization under a different name.

Ask HN: Critique this idea