Opening The Filthy Lucre
It is well known that if you can pay developers to work on software, we all get more software, quicker. Many Open Source projects have had a crack at trying to generate revenue and many have failed. Jono Bacon and Stuart ‘Aq’ Langridge explore the different options available to Open Source projects, determine which ones are viable and what challenges and opportunities some of these options have buried inside them.
Of course, we are the very start of the discussion! What do you think? Have you been involved in raising funds for a project? What were the challenges that you faced? Did it work out well for the project? Do you donate to projects? What methods would you be happy to give money to a project? How do you decide which projects to donate to?
42 Comments to “Opening The Filthy Lucre”
Leave a Reply
sorin7486 (83)
Derek (72)
marxjohnson (61)
Flamekebab (50)
b1ackcr0w (49)
Hessiess (42)
James Duncan (42)
mg (42)
tola (40)
Shane Fagan (39)
Right now Lukáš Tvrdý is getting paid for working on Krita by means of community founding, which I think is a really good way to get paid for improving free software.
What do you mean by “community founding”?
Lukas turned out to be available for three months between the end of classes at his University and his exams, so we appealed to the community to donate money. We needed about 3000 euros and thought we’d need three months to get there. Within one week we were over target. You can read the appeal here: http://krita.org/index.php&option=com_content&id=20.
One thing we did beforehand was making a detailed plan for the things Lukas would do: mostly optimizing, but also usability work.
That’s a really useful idea; people can much better judge whether they think the donation is worth it to them if they can see exactly what it’ll be used for.
Another thing that people appreciate is Lukas’ weekly updates on his blog: http://lukast.mediablog.sk/log/. Plus, the Krita project is pretty good at chronicling progress, updates like http://ardour.org/node/3413 are a weekly matter with us.
That last idea Aq came up with about keeping the Linux version free while selling it on other platforms is about as good an idea as trying to reduce CO2 emissions with tax incentives.
Basically, you’re incentivising people running your software on Linux. If everyone does run your software on Linux, you get no money. If everyone runs your software on something other than Linux, no-one’s using Linux. It’s pretty much lose-lose. Plus the users of other platforms have one less reason to choose your program over commercial/proprietary competitors.
Jono: remind us how long it took the Open Rights Group, after the pledge finished, to get you to set up your £5 a month Direct Debit?
Gerv
Original, Gerv…original.
Yep, it took me a while, for which I think I have been scorned enough by the Lugradio community.
More seriously, the trouble is that the money-making model differs wildly depending on what the software does. Media players can integrate with music stores; browsers can make deals with search engines; database engine hackers can usually get embedding contracts; podcasters can sell merch or do shameless product placement, and so on. I’m not sure if it’s possible to give general guidance…
Gerv
Wow, snarky…
I think the point was not to be sarky so much as to say that every case is different.
Theoretically there is a solution – make all software open source – programmers would get paid for programming with benefit to everybody. Law that make people pay hundreds of dollars for cd copying has to end someday – we all have cd/dvd recorders now, oh and we have Internet too. Now everybody ‘invents wheel’ all the time.
But we live in real world. I don’t see solution different than just people like you doing what you do – showing benefits of open source software to broader audience including companies.
What about Flattr a model that could work with modifications maybe?
http://flattr.com/beta/
This looks really interesting and sounds like a great way of contributing to a range of projects.
I am curious what the costs are involved in using flattr: e.g. if I click 100 projects and end up giving each 11p, does the cost to transfer the money outweigh the contribution?
I hope some folks from flattr come on and share their views on this.
I can’t find my digg comment on this, but as a person who blocks all online ads, I have wanted to be able to use a system like this for Ars Technica, Phoronix, and Jamendo for quite a long time.
I had sent correspondence to my ISP, Cox, to provide an optional monthly charge to be distributed to the sites I visit, but they didn’t really go with it.
This needs to be built into the Ubuntu Software Center so that application developers have strong incentive to properly package their applications and get them into Universe. I’d love to see Canonical buy Flattr and integrate it into Ubuntu One.
10% commission is the figure I have heard touted.
Well from what I understood from the video it seems they are creating a micro-payments system on top of the existing banking one. The problem with banks is that they have huge inefficient over-engineered barely working systems that are to costly to use for payments smaller than a certain amount. I should know as I write such software for a living. Flatter seems to handle what goes where on their own and only make the transfer to the content provider once it gets up to a reasonable amount. It’s a really good idea and I really like the way it’s implemented.
It just occurred to me: it works exactly like sponsored adverts… I mean just like you pay google a certain amount to put your link in the advert and then that amount is divided to a great number of clicks and a small portion of it goes to each of the pages where a click originated from. I mean it’s a model already wildly used but they just twisted it a bit to serve another purpose. Quite ingenious.
Add this to APT, and I am sold
. So, somewhere I indicate that I want to donate, say $25 dollars per month. A popcon-like system will make note of the packages I have installed monthly, and distribute the money to upstream for the package. Of course, the systems needs to be that it is opt-in for upstreams (I do not want to donate to projects that are, say, dormant).
Well, thanks to Jono & Stuart for the mention in this interesting discussion. Its very humbling (and, to be honest, exciting) to be viewed as a successful model of at least one funding model. A few comments.
To clarify the “Ardour model”, these days the money flows via 3 related pathways. There are standalone donations, initiated for whatever reason (although Ardour does have a post-export nag screen that can take the user’s browser to the appropriate page if they choose to do so). There are annual subscriptions, which are conceptually similar to donations, but take the form of 12 monthly payments, thus providing more stability to the cash flow. Finally, downloading Ardour requires going through a “pay tunnel” where you get to declare, Radiohead-style, how much you’re willing to pay for the software (which is available as source code or pre-built OS X bundles).
Of course, Linux users who get Ardour from their distributions don’t see the pay tunnel – they may never even see the website. This is a serious issue, but not a showstopper for now.
Its possibly also worth noting that I received support from 2 companies (SSL and the other one prefers to remain unnamed) that basically took form of an completely industry-standard salary with the expectation that I just worked on Ardour. These two instances of corporate support were vital in getting Ardour 2.X into release.
I think that its more likely that other companies would want to form partnerships (not necessarily of the legal kind) when the software involved is big and complex (like Ardour). This happened with SAE (albeit not very successfully) and is now happening with Harrison and the Mixbus product that is a blend of Ardour and Harrison signal processing know-how and technology. With Mixbus, I get a (relatively small) cut of each sale of Mixbus, which means that even if its existence makes Ardour itself less compelling to other users, I still get funded at about the same rate.
There are a couple of companies who plan to build ready-to-run boxes (“computers”, but … well, more so) with Ardour installed. This is another potential revenue stream, but at this time I have not discussed any revenue sharing arrangements with them.
There was some mention of a “bounty” model. We have such a feature on the ardour.org website, and to be honest, its been mostly a failure, for a couple of reasons. The first is that most users don’t really understand how much work is involved in developing even relatively minor features. We regularly end up with bounty amounts on the order of US$100 for something that represents a week’s worth of serious development effort. The second problem is that when it comes to bug fixes, it is often the case that it takes more work to get to the point where a good estimate of what is required to fix the bug than it does to actually implement the fix. This doesn’t really line up well with user expectation or with a system where you’d like a developer to say something like “I could fix that with about 2 hrs work” – if it takes 4 hrs to be able to make estimate, then a renumeration system based on paying for the fix itself generates discincentives to fix “deeper” bugs. That said, for some medium-sized, widely desired features/fixes, I think that a bounty model can play some role.
I don’t have the financing thing solved – not by a long shot. I have somehow miraculously (and perhaps temporarily) ended up being able to make a moderate income (compared to what I could get if I went for a regular programming job). Being entirely funded by users is exhilarating. Its also scary, and reminds me of the endless “tricks” the proprietary software vendors use to try to keep sales and support income up (or at least flat) – its very, very easy for projects to just fall off the radar. It is frustrating that we still do not have enough revenue from Ardour to pay any other developers – with 2 more reasonably committed devs, world domination would be ours, and we could probably even convert Jokosher users to Ardour too
I love what you are doing Paul, I really want you to find as many revenue streams as possible for Ardour.
Just an idea but, have you ever thought of giving your recurring donators a little extra content. Such as official PPAs or a members mailing list (forgive me if you already have one). Just it seems to me the only incentive ATM is one of charity.
Keep up the good work,
Sam Pattuzzi
Hi Paul,
great to have you comment here. One thing that was on my mind already when listening to the FLOSS-Weekly episode with you and Jono was: Wouldn’t it be the most natural thing to have a donation system built into places like the Ubuntu Software Center? I could imagine that this would be a great place to raise awareness of FOSS project financing amongst users. It could be made dead easy to donate to any project using one account associated with the software management application. When distributions and software conservancies would work together on a system like this this could have a huge impact on the advancement of free software.
Of course I am looking at Canonical here, them spearheading the whole software store idea in Linux currently. It would be for their own best. E.g. perhaps also the best argument Canonical could make to win over the “nay-sayers” in the whole commercialisation discussion.
Cheers and greetings,
OolongBrothers
I think the main bit of wisdom my 38 years has for this shot is “you cant please everyone ,so you gotta please yourself” Honestly this subject frightens me with its terrible flame potential!
hi, interesting commentary!
this ones certainly the difference between good dedicated support and a project falling off the rails.
i’m quite aware of the ardour project, and appreciate paul’s creative attempts to finance his development, he certainly sets an example to others running the same marathon.
ardour is however, effectively in a different league to many projects – it fills a hole required by many industries which at some point need to record and mix-down audio, and who often demand a high price for their services. professional DAW software is usually costly to begin with so the bar is already raised.
i think the task of raising revenue becomes more difficult as a software targets more niche areas of production – the same issues exist for proprietary software and successful products often demand high license fees.
drawing the interest to a concept beta, which might allude to a product, but only scratch the surface is very difficult, few people will see the potential to get it over the hump which might make the prototype actually useful. this kind of thing would usually be the product of someone tinkering, or the rare in-house development of a company prepared to let the r+d trickle out to the opensrc community (usually if their use for the product is fleeting or non-existent).
to get a project over the hump that comes from likely burnout of creative design and prototyping is something that really needs to be discussed. i feel that there are many great projects that have fallen prey to this slump, which don’t meet the expectations of the user or don’t find the developers whose imaginations might engage with the initial design.
i think this is why linux users haven’t seen a derivative to adobe packages like flash(authoring) and after effects – and to a greater extent, community development of a professional video/composting environment (cinelerra is by and large a one man operation and tends to suffer from this fact, blender is designed with 3D in mind and comes from proprietary origins).
one thing i think paul davis does exceptionally well is communicate directly to his users, and explain the difficulties of their demands, contrary to this i think the ‘bounty model’, might be failing for lack of transparent communication, users need to be made aware of the overhead costs in time and money of their feature requests before the notion of donation is pushed. i think this would allow the developers to be able to survey better what users are expecting and how they wish to work, and users would see what the development costs of a said feature would be and then perhaps reprioritise their demands. a good system that encourages users to help sculpt the tools they are helping to pay for would surely be twice the incentive to contribute money and ideas. (is this an open-src project in itself?/are there successful open-src marketing/distribution models?)
i really want to see ideas put back on the map, innovation is the key to new approaches to computing, and we in the opensrc community should be contributing far more to the perceived status-quo, there are many liberating factors to not having a company agenda, and new radical ideas should be an outcome. as a long time user of linux for mulitimedia its been refreshing to see so much autonomy in some excellent fringe products, i really hope that we can find a revenue model that matches this ideology and competes against (or challenges) greater powers.
on closing, i just want to mention amazons quite horrendous global contributor workforce, the mechanical turk, which is almost inside out to the opensrc software model – getting users to participate in menial tasks for a pittance (mostly to alleviate boredom). what it does do well is distribute resources, perhaps there is something to be learned?
dmotd (dmotd at gee em ex dot net)
i didn’t really want to make any examples in my previous post, but a list to those i am somehow indebted to seems necessary, look them up! pure-data, supercolider, gimp, inkscape, scribus, fontforge, freetype, blender, cinelerra, ffmpeg, mplayer/mencoder, dvdauthor, jack, ardour, audacity, rezound, zynsubfx, ladspa/dssi.. many great tools that have managed to provide somehow, and many others i have used only once so specifically that i will never even remember using. THANKS!
In my opinion the key to the future development of Ardour and the financial security of its developers is not to find and rely upon new revenue streams that mirror those of commercial applications (such as tie-ins with hardware that ships with Ardour pre-installed, or similar sources that Paul mentions above).
The current subscription model provides the ideal interaction with users and development; with user needs and what is implemented. Paul has freedom to fix and add what best reflects the requirements of Ardour users, and everyone benefits (including companies who sell products reliant upon ardour) from this fact, and the nature of the code remaining GPLd.
As awareness of ardour increases, as it will exponentially (especially after v3 is released) the number of subscribers will also increase. The current annual sponsership method via paypal, which I have used for nearly a year, seems to be providing a good income despite the currently tiny number of Ardour users.
As more individuals, studios, schools and all sorts of other organisations use Ardour, I very much doubt that insufficient funds for Paul’s salary will be a problem in the near future.
The unreliable nature of the current donations system no doubt has problems for Paul, and cannot currently provide him with an ideal level of financial security. However, the situation is improving regarding awareness and uptake, and it really doesn’t seem worth risking subordinating project goals to the needs of a few commercial sponsers to get a small cut of their revenue streams to shore up project income at this critical stage in the project’s history.
Ardour is heading for ‘world domination’ of its market, and this will be a phenomenal benefit to all kinds of people in all sorts of ways. A subscriptions model of funding will provide Ardour with unparalleled freedom and focus to fulfill this role. Its current lack of commercial agreements is one of the projects’ greatest strengths, and I hope that this continues to be the case.
Why is a direct charging model so infra dig? Software does have an intrinsic value. Direct charging for software has a bad name because it has been abused in the past. Maybe it time we rethought whether it’s the model that’s wrong or the people setting the price tags are at fault. We’ve seen Paul Allen buy the worlds biggest toy boat becuase his company gets away with charging everybody $100 for a $20 product. And we’re right to be annoyed, because most of us have no problem paying for what we use, as long as we’re not being conned or squeezed for our cash. I also think the term “Free” (as in beer), also has more negative connotations than we realise. Readers Digest Free Prize Draws are “Free”, you get mail spammed, a subscription you didn’t want unless you tick the right boxes or cancel it. In the US, as we can see from the healthcare debate getting something for free is seen the mark of Socialism and anti-American. Is there a section of potential Open Source paying customers who are excluded because we don’t offer a simple transaction?
“Why is a direct charging model so infra dig?”
Because:
“…people have a strong tendency to assume that software has both value characteristics of a typical manufactured good. But both these assumptions are demonstrably false.”
“…maintenance…makes up the vast majority (more than 75%) of what programmers get paid to do. Accordingly, most programmer hours are spent…writing or maintaining…code that has no sale value at all.”
“…the theory that the sale value of software is coupled with its development or replacement costs is even more easily demolished…the price a consumer will pay is effectively capped by the expected future value of vendor service“.
“…software is largely a service based industry operating under the persistent but unfounded delusion that it is a manufacturing industry.”
Basically charging per sale for software is one of the worst problems with existing commercial software development. A subscription model is a far more accurate representation of the costs involved with producing software, and furthermore it has the critical bonus (particularly relevant to the Free Software world) of keeping user needs aligned with development priorities – something that it increasingly rare in traditional development systems.
Therefore the direct charging model is rightly unpopular, and especially so amongst FOSS users and developers.
As I understand it, a subscription model implies regular delivery or access to a product. So, what about an initial one-time fee to download the software, then a subscription for regular updates?
I could see an open source project charging for access to a repository with pre-compiled packages. You pay to get in, and then you pay a monthly fee to stay in, with the understanding that updates will be published at least once a month (or whatever).
@B1ackcr0w – the biggest issue that I’ve noticed with a direct charge model is that its somewhere between irritatingly hard and almost impossible for software developers/vendors to release their software as a single downloadable “product” for even a modest subset of the X-hundred current Linux distributions. When I “sell” Ardour for OS X, I just build two bundles, one for PPC and one for Intel (if I was willing to do a bit more work, I could make a single universal bundle) and that’s it (more or less). For Linux, I get a headache just trying to imagine the work involved in making a binary of the program that will actually run on an arbitrary Linux distribution.
This forces distribution to occur mostly via repositories, which in turn pretty much rule out any kind of direct charge model.
Hi Paul,
repositories the way they are currently perceived do have that problem. But I don’t see why we could and should not try to change the repository front-ends to allow a financial feed-back channel to upstream. See my above post. I do really see a chance for repository-centric donation systems connected with upcoming software store approaches; as a big part of the infrastructure would be in place already then. Distributions could get a cut of the payment/donation for packaging and hosting.
It’s more a matter of will and sorting out the organisation. I think the successful implementation of such models is an absolutely central point in preserving the notion of ownership of free software amongst the growing group of non-hacker users that cannot so easily accomplish that via contributing to projects.
So let’s make that happen, folks. Or am I seeing this completely wrong?
Enthusiasticly, OolongBrothers
In fact, the Ubuntu Software Store seems like a pretty good place to implement something like this. I remember it being mentioned when it was first branded as a “store” that there was a possibility (likelihood?) that Canonical would make third-party commercial software available through the store. Surely then it’s not a huge leap of imagination to allow donations to free software projects through the same interface?
Paul,
Ever had a look at the Buildservice? It builds packages automatically for most major linux distributions. In the works (well, planning…) are mac OS X and Windows packages
Something which doesn’t seem to have been raised so far, and admittedly it’s not a solution for all projects, is that of encouraging our government agencies to spend their money on sponsoring development of software, rather than on licenses.
It’s my view that it’s better to spend money on improving, for example, the open source equivalents of VISIO and MS Office to bring the (missing from the point of view of the government department) features they need, rather than paying repeatedly for the software with no benefit for tax payers.
If my taxes payed to improve software which helped run my government, and in the process benefits me by providing that same software to me, as I’ve already paid for it’s development, then great. (What software have I received as a result of that billion pound deal the NHS did with Microsoft for example?)
Interesting blog post from sh on this subject at http://shermann.name/?p=1449 too!
I’ve donated to projects in the past, as well as that I help to pay Leos’ wage on Twit with their “Tip Leo” option every month.
There are only a handful of projects I’ve donated too and it would be nice to have a model that allows you to be more specific with where your funds go, but you don’t have this model with charities, you send them money, and it could go directly to the cause or it could go into fuelling the planes to get there, or other office costs. Who knows?
I’ve been thinking about this issue for many years, and for the past year have been formulating a fairly specific idea about how to solve it.
What’s missing is a marketplace for open source development. Sure, the whole Internet is basically that–but that’s an awfully big place to look, and it’s a bit disorganized.
We need an eBay for open source development, with a twist, a cross between a bounty system and a crowd-funding mechanism.
I proposed a “Funded Development” model, based on funding features you want developed. Basically, I’m planning to set up a site where people can request a feature for any open source project. Unlike crap sites like eLance or oDesk which basically farm it out to the cheapest regions for development, first crack at bidding on the project goes to the core development team. If they want to pursue it, they name their price. And then everybody who wants to see the feature can contribute any amount until it’s raised. The developers create it, and then get the pot of collected money.
So in Paul’s case, when he gets a request for a feature to be developed that’s going to take a week of hardcore development but only has a $200 bounty, he can say, “this is going to cost $2000.” And then the original requestor can put it out to get other people interested in the feature to chip in. Once there’s a pot of $2000 raised, Paul knows he can get paid for getting the work done, and as soon as it’s done and verified, he gets paid…
I’ve done a lot more thinking on how to make this work, how to build in some checks and balances (and also recruit moderators/project managers who get a cut of the transaction if the project is successful). Developers get paid to develop. Users have a single place to request features. The site acts as a neutral third party, to both ensure the developers get paid, and the users don’t pay if the feature never gets done.
That’s my vision of what’s necessary– I do need some people to help make it all work. If you’re interested, please drop me a line at john at freelock.com.
So, to confirm I understand this, this is essentially the Street Performer Protocol applied to coding?
ha ha. Like that.
But it’s not really a street performer model. While we’ll certainly take contributions for projects in general, I think people will pay more, and more often, if it’s going to some particular pain they have.
If a project has a list of features that people want developed, along with the amounts needed to get them done, that’s a pretty powerful sales pitch. The users no longer are being asked whether they could spare some dollars for the project, but instead asked which feature they want to see first? And they vote with their dollars. (Sorry, American here). There’s a bit of psychology going on here which I think will work a whole lot better than just a “donate” button.
There are a few open source products I would pay for.
Firefox (despite the fact I have all but abandoned it except when I have to use windows).
Google have a huge part in keeping Firefox going – altho if Google ever stopped Bing would be in there so fast.
Openoffice (I would pay for this). (Oracle better keep this going)
I would pay for Linux Distros on CD (Just as a donation and save me a bit of time and hassle for dl and burning).
I apologize for coming to this shot late as my MP3 player has been on the fritz.
When I heard this shot, I felt like a kid trying to get his teacher’s attention in elementary school. I couldn’t wait to get back to my computer and post a reply.
I think Open Source community could learn a lot from the Role Playing Game community in this regard. In recent years, a number of independent developers led by people like Wolfgang Baur (http://en.wikipedia.org/wiki/Wolfgang_Baur) who came up with a publishing process that can bootstrap a game publisher through the creation phase of a project. He calls his version the Open Design project (http://www.koboldquarterly.com/k/faq) which is, essentially, a leveraged patronage system.
Patrons pay for a proposed idea up front. After getting a critical mass, the project moves forward with input from the patrons as one of the perks. After the project is finished, the patrons get a book, or game, or supplement that meets their individual desires and Baur (and his Open Design project) can then sell the finished product (or start other patronage projects based on it).
Other people try ransoms. Role Playing Public Radio has a campaign setting that it has been slowly building up using public ransoms (http://slangdesign.com/rppr/the-new-world-free-4e-dd-campaign-setting/). They post the ideas for the next installment on Kick Starter (http://www.kickstarter.com) where they take pledges until they meet their cost threshold. Then they create the supplement and release it for free to the world.
It seems to me that an open source project could use tools like Kick Starter to put a patronage price tag on various features that they could likely implement. They could put real prices on the time it would take to implement those features and then use tools (such as the private forums that Baur uses) to work with the patrons as an invested beta test and focus group. Once a feature is completed, perhaps patrons could have a discount on support (if the project offers that as its revenue model).
It may not be a direct fit and I’m not sure of all the details, but it would be interesting to see someone try it.