Slashdot Mirror
Ask Slashdot: How To Get Paid For Open-Sourcing Your Work?
kc600 writes "Say you're a freelancer, using mainly open source solutions. You notice that customers, although they don't object to the whole open source idea, don't see the point in paying you for the time it costs you to properly open source your code. As a result, code is not released, because it would take too much time to factor out the customer-specific stuff, to debate architecture with the other developers, look at bug reports, et cetera. You feel there's something to contribute that many might benefit from. The code would also be better maintained if more people would use it, so the customer's project would also benefit. But you're not going to do it in your free time; you have enough on your mind and the bill is paid, right? What useful tricks can you think of to encourage yourself — and your customers — to properly share code, to the benefit of all, and get paid for it?"
The hard question really is "how do I convince people to give me money for writing software?". Open or closed are just details if it's being sold as part of a provided service.
http://en.wikipedia.org/wiki/Johnny_Appleseed
When I open sourced the programs that had made me some money, but I had no time nor the stamina to keep working on them, I didn't expect to get paid for that.
Instead, I thought of Johnny Appleseed.
The programs that I open sourced, to me, are old stuffs. I could have kept them under closed source, store them in CD-Rs or external hd or old computers, or ....
I could have done that, but if I did that, it wouldn't benefit me, nor anybody else.
When I open sourced those programs, I didn't even know if anybody else wanted them in the first place. I just placed them online, did some advertisement on related sites, and then, let go.
If the "appleseed" blooms, good.
If they don't, well, it'd be the same as I locked them up in CD-Rs.
The most important thing is that I've set them free. Their "lives" after being set free depends on their "fates", or in spiritual kinda speak, "karma".
Once they are open-sourced, they do not belong to me anymore. Now, they belonged to the world.
Muchas Gracias, Señor Edward Snowden !
So instead of a programmer grabbing a Linux kernel, and spending a few days to add a driver for customer specific hardware, you would advocate that he wrote his own operating system instead ?
I work for a 5-person tech cooperative. We were writing code, documentation, etc. that we wanted to contribute back to the community. So out of our "profit", we made sure that we set aside some funds for our members to spend some of their time abstracting code, packaging it up for release, etc.
The basic principle is the same for a freelancer - you have to raise your rates. Are you charging $100/hr? Charge $110/hr. Use the extra money to pay yourself to package up the code.
In terms of "useful tricks" - well, as a freelancer, you don't have the privilege of someone keeping you honest to your goals. You can change your personal rules whenever you want. But frankly, I would say my co-op has made a net PROFIT on open sourcing our material. When we open source it, we post about it on our blog, Twitter, etc. This increases our referrals from other developers, it means more folks are finding us on the search engines, we gain credibility with other developers when we need them to fix a bug in their module. Maybe the "trick" is to remind yourself of those advantages.
I have been paid for writing open source software but only in the following context:
Open source project X almost meets our needs, however it is missing the following 3 features. I could spend two weeks implementing those features (but we will need to contribute it back to the project) or two months implementing the library from scratch, which do you prefer?
Basically, I would say that you need to present a very concrete value proposition in front of the customer and let them pick... starting an open source project as a contractor and on the customers dime is pretty much always going to be a non-starter.
But if you were to make it into a one-off Kickstarter project, it wouldn't be pulling a Lunduke. Personally, I'm getting sick of all the Kickstarter campaigns that are "I want to make a profit, but I'm not willing to risk my own time and money -- you guys take the risk, I'll make the profit, m'kay?" and would relish more campaigns that say "I want to make an honest buck -- pay me fair and square for my time, and I'll forego future royalties," because that's really the whole point of risk-reward. People working on royalties take a high risk, gambling on the reward. Eliminating the risk without eliminating the chance of a high payout, it's, well... unfair.
Got them moderator blues I blieve I walk out the do', With these mod-points I been gettin', I 'most never post no mo'
Mod this up to the skies, please! In my opinion this applies to -everything- forgotten books [I've been thinking about trying to find the rights owners for some of them], forgotten music [same problem, I'm from the 60s there's some wonderful stuff buried] etc. Unhappily, for me, this non-publishing is collateral damage that [in the spirit of Bentham: http://en.wikipedia.org/wiki/Jeremy_Bentham%5D should be increasing the sum total of human happiness and instead lies locked or buried.
The second point, for software etc., is the 'standing on the shoulders of giants' idea. That is, something fairly simple or a few bits can be used to build something spectacular or inspire it. Same point of inspiration idea for music too.
Great post with a great many 'extra' implications, thank you.
On y va, qui mal y pense!
Another poster already called you an idiot, so I can skip that part and get onto exactly why you are wrong. People pay me to write software because they need that software written. That is the best motivation for writing software and the reason why about 90% gets written. The remaining 10% of commercial software is written because someone thinks it's a good idea and that they'll be able to sell finished versions.
Pretty much all of the software that I've been paid to write has been released under a permissive license (MIT, FreeBSD, or UIUC license). This is because there is a non-zero cost associated with maintaining a proprietary fork, which is basically what happens when you make any nontrivial changes to open source code and don't push them upstream. New features and bug fixes upstream may change some interfaces that you depend on and this means that you end up either having a version with known bugs (including security holes), or you spend money backporting the changes to your branch. If you upstream the code then someone else pays for this and it is cheaper for you. In FreeBSD land, we're currently working with Netflix and Juniper to upstream a load of their changes for exactly this reason: they want to spend developer time (and therefore money) on new features, not on keeping old ones working.
More importantly, if you don't upstream your code and it's useful to others then eventually someone else will implement the same feature, but often in a different way. You then either throw away the code that you've paid to have written and use theirs, or maintain a fork that is now radically different to upstream and therefore much more expensive.
As the developer, you also benefit from being able to use the code elsewhere. Everyone wins: your current customer gets a lower long-term maintenance cost, your next customer gets a smaller up-front cost, and you don't get bored implementing the same thing lots of times.
I am TheRaven on Soylent News
I work for a company that does a lot of Open Source stuff. Here is how we manage it: We have core toolkits that are open source, and custom applications that are closed source, made for specific customers. When ever a customer needs new functionality, we try to generalize it and put it into the toolkits, which we then release. We tell the customer that we have this open source toolkit which we use for the project, and which we keep improving. But we don't specify how much of the work goes into the toolkit, and how much on the custom side.
Those toolkits have been our main marketing effort, and have certainly paid off. Within our very narrow field we are world famous, and our toolkits almost dominate the market. Nobody can afford to build a competing one, when ours is free. Although anyone may use our tools, we happen to know them best and have most experience with them, so we can often do any given job faster than others. The company has survived over a decade, and has expanded internationally, and is now all of 15 people.
In Murphy We Turst
There is nothing in the GPL that requires you to contribute back to the community. The only requirement is that you give the same rights to anyone that you give derived works to that you received. For example, you can take GPL'd code, extend it, and then sell the result to a company, with a contract prohibiting you from selling or giving it to anyone else. They are then free to redistribute it under the terms of the GPL, but they are not required to. The GPL does not require you or them to return any code upstream, only to pass rights downstream.
I am TheRaven on Soylent News