What's the best way to do free software bounties?

First, I'd like to thank Joey, Noah and Jeremy for providing much needed feedback on a related blog post. Thanks guys. It really got me thinking. What if instead of a contest we figured out how to do community funded bounties? Wouldn't an open, continual system of free software bounties be much a better idea than doing another contest?

In contrast to a contest, bounties would open up both the funding and the funded goals to the community. Everyone could vote with their wallets for specific goals they'd like to see accomplished. Rather than dictating from above Alon and I could participate in the process on the ground level (e.g., starting new bounties or supporting community created bounties).

There wouldn't have to be be winners and losers. If multiple people or teams collaborate on a solution to a bounty we could divide it amongst them.

To be frank the more I think about it the more my enthusiasm for TurnKey sponsored contests diminishes. The main problems with contests is that they're top down (I.e., funding and direction is dictated from the top) and reward competition instead of rewarding collaboration. You get what you reward.

By not rewarding the right thing, a contest would discourage rather than encourage the spirit of win/win collaboration that is critical for the long-term success of TurnKey as a true free software project.

Sure, figuring out how to do The Right Thing may be more difficult than going for the low hanging fruit but it would pay off in the long term.

Something about this idea really excites me.

What do you think? What would be the best way to implement a system of free software bounties?


For those of you who haven't been following developments, last year we squeezed our build infrastructure into a single self-contained system image. We want to see more people in the Debian / free software community using this because we believe it could do a lot of good, not just for TurnKey based solutions but for any Debian based project.

In fact, we'd like everybody who could find this useful to know about it and use it, even if they don't contribute anything back to TurnKey: what if all Debian/Ubuntu based dists used TKLDev?

Note that this is what Alon and I actually use to maintain all 100+ apps in TurnKey. Without all of the automation in our build system that wouldn't be remotely possible. Nobody could pull it off. There's just way too much work. But the system we've developed makes maintaining a ready-to-use Debian system template as easy (sometimes easier) as maintaining a single Debian package.


Hi Liraz,

There is an old joke "How do porcupines mate...carefully". I am not a big fan of money and opensource. I believe that money associated with opensource is best sustained if orthogonal (support, consulting, brand, trust, etc). Mating them together should be done...carefully.

As soon as money is targeted specifically to building opensource it becomes a commercial venture and the software becomes commercial software or worse...open-core (software that pretends to be opensource but isn't...the free car with the $30K specialised steering wheel).

Potential risks include:

- People giving money will want to know their interests are prioritised so a few individuals end up driving the agenda;

- People who contribute code and time for free get disenfranchised;

- The amount of money that code contributors do end up with is likely only enough to buy lunch.

Money is rarely higher than 3 on peoples' compensation priority list (http://tedleonhardt.com/blog/index.php/news/its-not-about-money/).

If you are determined to put this in place you might consider a fixed contribution per enhancement model to keep a few contributors from driving the agenda.

I applaud your intention with this and accompanying posts regarding the future of TKLX. I hope you find the right formula that works for you, Alon and the TKLX community. I'm just not sure what problem you are trying to solve though! Is it commuity participation? If it is, then I would recommend concentrating on your vision statement. Communities follow leaders with vision and passion. You have the passion, I'm just not sure what the vision is. You can be reasonably certain  though that it is unlikely that TKLX community members would join or leave based on money-related matters.

Just one man's opinion...


Tim (Managing Director - OnePressTech)

Jeremy Davis's picture

And I agree with him to a point. I also see your point Liraz re the competition. And whilst a competition works for the idea of new appliances, it doesn't really lend itself well to other things such as documentation or maintenance, or even enhanced features really...

So I see the sense of bounties, but as per Tim's response I also agree that how it would work in practice might be interesting (and/or problematic)...

All that aside though, as synchronicity would have it, I stumbled across Bountysource the other day and I wondered along very similar lines to Liraz... Perhaps that might be a good platform to give it a go (if that's the way TKL heads)? It would mean no real effort on TKL's part to trial the idea...

FWIW Bountysource have a paragraph regarding introduction of money into an open source project in their FAQ:

I'm worried about introducing money into my community

This is a reasonable hesitation. However, money has already been a part of open-source for decades. Many open-source contributors are paid by their employers to work on open-source. Many projects already have donation buttons. Many open-source developers have consulting businesses around their projects. Ultimately, the motivations behind open-source contributions already vary widely. The easiest way to make sense of it all is to focus on the code itself. At the end of the day, if high-quality code is being contributed to a project, the incentives behind the code should be irrelevant.

Obviously it is in their interests to get people to use their platform, but I think it still holds some validity.

Hi J. Great link to Bountysource. I'm sure I will find a use for that.

I agree with the quote you included. That was what I meant by money needing to be "orthogonal". As soon as you pay anyone directly for contributing to opensource software...why would anyone contribute software for free! And if there is only a little money being spread around...isn't it just a poor paying I.T. job!

In general a modular sotware architecture seems to be supportive of an open community with most of associated commerical partner employee contributions going to the core and most of the community participation going to the plugins. Another way to ensure clear separation between paid and unpaid contributions. The plugin community essentially becomes the truly sustainable open community within the initiative...as long as the commercial organisation doesn't pull the funding for core work.

There is a risk to having a community largely founded on side-contributions from a commercial organisation of funded employees as well...you end up with open-core or the company pulls the plug and the community folds. Case in point...take a look at http://maqetta.org/ . This is a solid piece of opensource that was funded by IBM. When IBM pulled the plug https://groups.google.com/forum/#!topic/maqetta-users/dLGzdERrkG8  there was no community outside of IBM to continue the work http://www.ohloh.net/p/maqetta.

I think that TKLX community, as it matures, will be a stable ecosystem that will contribute more over time and things will level out. I doubt money will change the current TKLX community dynamics.

Perhaps a documented vision that links TKLX into the emerging DevOPs field and consistent marketing where DevOPS is discussed could provide all that Liraz and Alon are looking for.



Tim (Managing Director - OnePressTech)

John Carver's picture

Kickstarter and other similar sites have had surprising success in funding many projects.  Why can't the model for collecting pledges for new appliances or enhancements to existing appliances be applied here?  The bigger problem in my mind is how to fund ongoing support for the many existing appliances.  Despite the automation provided by TKLdev, there is bound to be development effort for each new TurnKey release.  There is also effort needed to keep on top of security issues like Heartbleed.  If we learn nothing else from the OpenSSL fiasco, we should understand that we can't continue to expect to receive support for free.

Information is free, knowledge is acquired, but wisdom is earned.

Jeremy Davis's picture

I think that a Kickstater campaign could be a good idea... I think we should consider that further...

Marina Yana's picture

Informative blog with good explanation. Thank you


Add new comment