Two things though:
* Developers, in general, are not business people. This may be different in Silicon Valley, but in a lot of the world, programmers have spent their life concentrating on one thing, learning to program. That doesn't mean they don't just "not know how" to start a business and find clients. It means they don't even know how to do the basic, first steps to registering one, knowing what to charge, etc. Let alone the complications of supporting an open source model on top of the usual business things like tax. It is just easier to throw up a paypal link and be done with it. Yes, something should be done about this, but there you go.
* When people pay for things, suddenly they expect it to .. work. As in, be stable. And be supported. And be kept up to date. All of a sudden you go from being able to work on your favorite project to supporting clients and responding to their needs. Many programmers don't want this. Admittedly, if you're floating such a project as your main income, perhaps you should be considering these kinds of things, but on the other hand one thing that keeps an FOSS project going, and gives it advantages, is not having to worry about exactly these things. It also might have a huge disincentivising effect towards contributions; why would contributor contribute if they don't get a cut? There are specific challenges in business-oriented FOSS funding models that would need to be addressed.
I couldn't find anyone who could help us do something like run a small business or just help us figure out how to sell those useless stooge licenses I keep hearing companies want to buy instead of sending donations.
I had no idea what I was doing. I found a local lawyer who incorporated me for $1000. He wrote the commercial license for my software for another $750. I had a friend who was a CPA and taught me how to keep books. I was terrified but literally every business is started this way.
Need help? Walk outside your house and ask the five nearest businesses who they use for legal and accounting. Talk to a professional.
I'm also happy to chat about starting an OSS business during my weekly happy hour, all welcome:
One possibility: Find a company that buys a ton of Matlab licences in order to get access to a specific library. Implement that library well, GPLed, for a price that is below their annual license cost.
It's a tiny niche (big company, small library), but if it can be found, it is one the Octave team could do well. Everyone, except Mathworks, would win.
Proprietary libraries do offer sponsoring companies greater competitive advantage, but the first mover to a cost savings is already a winner.
Getting a long list of big Matlab customers is a place to start. Sometimes they boast about the biggest fish in advertisements.
If running out of money is a critical problem for Octave development, make finding GPL revenue the core goal. Octave is now complete and capable enough that it can take care of itself for a year with bugfixes alone.
The first possibility has been attempted. In effect, that's partly why the GUI happened: a company really wanted it and paid for its development. jwe has been able to coast on this kind of gig for a while, about 8 years now. It's just that it's never been a very financially stable thing.
The rest of your suggestion seems to be about running a business. That's the thing we've just never really known how to do. There are a lot of business details that we've never really figured out because we'd rather be writing code.
the nice thing is that octave is established enough that you won't need to do a lot of marketing to develop awareness of the product, which is usually the hard part. you also have some validation for your product (a large and successful competitor and a bunch of support contracts for octave). it sounds like, instead, the hard part for you is figuring out what you want to sell and getting going. people seem willing to help, so take them up on the offer! i'm also happy to lend an ear or a hand if need be. good luck!
Every project has to determine their own path to sustainability. Knowing only what I've read here, I'd sell support in seat packs, e.g. $1000/yr for every 250 users. A good sized university might have 1000 engineering students and $4000/yr for a critical tool used by all their classes seems reasonable.
The great thing is that you define support however you want. 24/7 phone? Email during business hours only? Just make it clear to the customer so expectations are properly set.
A business might only have 10 people that use a tool like Octave but $1000/yr isn't a big deal if that tool is necessary for their daily work.
Still, the money from what you call "useless" licenses is minimal compared to the one that the businesses are ready to pay if they don't have any other alternative but to pay. Business are the systems that are designed to make profit. If they don't have to pay, they typically won't.
The key thing after that is to separate your income and expenses into personal and company. Never mix funds. Get a separate company credit card. You'll probably have to start out using the LLC's bank debit card - that's fine.
I can only speak for me: Because I did not pay anything for the foundation I built on. If someone comes and asks for a cut (and I had this more than once), I give it to them: A fair share for their 3-line patch to a million LOC codebase, minus my fee for reviewing, for tutoring and for project management. Unfortunately, so far, nobody paid after I sent my invoice :(
Just doing one thing all your life is not an excuse for not taking care of yourself. Developers know how to brush their teeth and feed themselves, and they know how to navigate really complex abstractions, they can certainly learn the basics of incorporation and contracts.
Using your analogy: to some (and I include myself), telling them "just start a business", is like telling your grandmother to post a question on stackoverflow when she can't figure out which little picture to click to open the word processor.
You have to understand that some things that from your point of view are easy, are hard for other people. Starting a business is NOT "brushing your teeth." There's basic "taking care of yourself," and then there's how to create a self-sufficient semi-profitable business to work on projects you enjoy. I don't see how you can make this false equivalence.
The point isn't to start the most successful company ever its to just be able to function and do what you want to do (programming). By your analogy your grandma is willing to navigate a confusing and dissorienting world to get what she wants (probably facetime with her family) but you are just going to sit around waiting for someone else to hand you what you want without any effort on your part to learn something new?
You dont need partners, you just need a business checking account and an LLC so you can invoice companies. I recommend waveapps for creating and managing invoices btw.
Sorry if my tone is a little tough but I read so many defeatist comments in this thread and it depresses me that so many smart individuals willing to do so much good for the world arent willing to invest some of their talents (like relentless google searching and asking questions on forums) in taking care of themselves.
I suppose the simplest way would be to offer dual licensing. GPL for non-profit projects, proprietary for enterprises. The problem is, if the source code you are working on depends on third party GPL software it's very hard to do that.
I think this is what should be fixed. There should be some way to have a chain of dual licensing, so that if someone uses a dual-licensed library, they can forward revenues simply to the original license owner from their proprietary license sales.
Source remains open, and developers can eat (and pay for other amenities civilization has to offer).
Blockchains?
There should be funding bodies for projects like this who can fund projects and also provide a semblance of orchestration. The Linux Foundation (with the Core Infrastructure Initiative) and the Cloud Native Compute Foundation are already working in the right direction, imo.
There are other organizations that work through contracting work to develop free and open source software. Red Hat and Canonical are obvious ones. But there's also smaller outfits like Continuum Analytics, Igalia, The SQlite Consortium.
Further, there are loads of technology incubators and VCs who would do well to fund projects like Octave to make sure their Data Science startups can rely on these tools.
> Blockchains?
I'm scared to ask what you mean by this.
A much better way than 'government funding' would be for users to just pay for it.
It's called 'capitalism' it works very well.
Oftentimes, it's difficult for the government and centralized bodies to know 'what to fund' and what not to - it's a hard thing, it can get political. Money goes entirely to the wrong places, even with good intentions.
It's not always possible to 'sell something' but often that works well because the amount of money allocated to the project is commensurate with it's value. Roughly.
So - 'bad projects' that are either crap, or nobody really wants - don't get funded.
And the ones that are awesome - and everybody wants to use - get support and expand.
Obviously it's not always ideal, but it's rational.
Maybe - instead of someone offering this guy a job, they could teach him how to charge a nominal amount for the software, in a manner that still respects it's 'open sourcedness'. Like a $5/month price for access to a super set of documentation, that everyone using Octave would love and have no problem paying if the used Octave.
And honestly, this idea of perfect meritocracy, the idea that if you don't get funded, then your project was crap, has been disproven a long time ago. There was a kickstarter that raised insane amounts of money for someone to make potato salad. Not as a product, but making a bowl of potato salad for themselves.
I pay Netflix a nominal fee each month and whenever I feel like it I can choose to watch between hundreds of shows (currently working through Penny Dreadful), without having to hunt down a torrent for it. Especially a torrent of good enough quality.
Same with Steam. Sure, I could try to find a game to download via torrent, but I know I can find it on Steam, often at a steep discount (or put it on my wishlist and wait until the inevitable one arrives and they notify me of it).
Additionally, I know it won't have any malware or viruses hidden in it, and will be the most up to date version (seriously, try finding torrents that are up to date for most things, it's almost impossible).
If those people won't pay a small sum to use this - or can find some easy substitute, then that could be an indication that this open-source project simply isn't that valuable.
If there is no easy substitute, and he is creating consistent value, he should be able to get some kind of revenue for something, possibly enough to pay himself and a few others to make it better.
It's not always easy, but it's not remotely irrational for him to try a few things. There's no reason for him to 'work for free' to help other people in their endeavours. They can pay him, and everyone still ends up a winner. It's called 'comparative value' - it's how most of the economy works. It's good. Apparently, he wants to 'get a job somewhere' writing code. How are his future employers going to pay him? By asking people to pay for the software that he writes. Or something along those lines.
In fact - given that he does have a 'high degree of expertise' in this open-source he made, that apparently people are using - it would make good economic sense overall if he (and possibly some others) were working on that, as opposed to him 'getting a job' elsewhere. It's a matter of plugging into the revenue stream. Again - not trivial, but something that could surely be tried.
If people are not willing to pay for something, then it's a very strong indication that it has no economic value.
Almost by definition.
When people pay $4.00 for a Cappuccino, and are not willing to pay $1 for 'an app' - then yes - 'the app' is worth less (to them) than a cup of coffee.
Sometimes there can transactional friction, IP issues, regulatory issues which distort the market etc., but generally speaking 'price' is a very good proxy for value.
If nobody is willing to pay money for this guys product, then it is worthless, and he should probably stop wasting his time by volunteering his time to make something that nobody derives any value out of.
Those people using it would in that case obviously have something else they can use that is 'just as good', or, they just don't need it at all.
After all - if they will pay $1 for a Bic pen, and not this software, well, that doesn't bode well for how useful this software is for them.
But mostly likely, would rather seem that this software is quite valuable to many researchers. Which is good, because they would likely indeed pay for it, just as they would a pen, a book, their computer, their lab-coat, their eyeglasses, their calculator or their lunch.
Governments and other organizations rarely know 'what to fund'. But there is a good model that is already working in other fields - the grant process.
Why couldn't there be an umbrella organization (not gov't) that accepts donations on behalf of the community, and then has funds available to support open-source projects? There could be grants written, peer review, etc...
Why would you do that?
Why not just have people pay for it?
No overhead needed.
"But there is a good model that is already working in other fields - the grant process"
'Grants' are speculative redistributions of capital into areas that some decision making body deems worthy - but which may have no bearing at all on what companies and people actually want to use - ergo - not usually very efficient.
We have no choice in some cases (i.e. pure research) but for this stuff we do.
This is just software, it's not rocket science, there's likely a way to just have people pay for it without too much overhead.
A researcher can buy his breakfast, his home, his equipment, his computer, he can buy software as well.
Sales, marketing, payment systems, and invoicing are plenty of overhead.
>This is just software, it's not rocket science, there's likely a way to just have people pay for it without too much overhead.
Sure. People are doing it. Red Hat, Canonical, Continuum, Igalia, and Mozilla are all organizations that pay for developers to work on open source. I would guess OP could look at NumFOCUS or get a job at Continuum Analytics if they want to help move Matlab developers into the open source world.
He could make something very easily himself, 1-click to $X a month for some bit of software or service.
In fact, if he did it properly, he could be wealthy, or chose to hire a bunch of people to make his software even better.
Mozilla is a great org, but I'm not sure if they are the best example as I believe they depend on the benevolence (or strategic foresight) of entities like Google etc.
So should I go tell Pepsi and Coke that they're wasting their money?
Plus, while Octave is an established project, there are many projects being started each day. No project comes into being fully established.
"He could make something very easily himself, 1-click to $X a month for some bit of software or service."
Do that, instead of working on his project. Plus, the source is already out there. It's not going away.
"No project comes into being fully established"
Octave is well established, so he doesn't have that problem.
For those that are 'new' - it's extremely difficult for governments to decide how to allocate funding. There are 1000's of 'open source' ideas. It's very hard to tell which one's would work, and have merit. It seems this guy was able to make something on his own. Great.
Now he can generate some revenue and expand the project - OR - work full time somewhere else, and plug away at this in his spare time.
"Do that, instead of working on his project."
He's not going to be 'working on his project'. He's going to be 'working full time somewhere else' because his project has no income.
The time and energy he spends on some kind of 'revenue' is an 'investment' not a cost. Once it's up and running, there should hopefully be net surpluses. If it goes well, he can even hire a small team to do 'sales and marketing', and possibly hire some developers to make his project considerably better, so that it's even more useful to others, and possibly even more people want to use it.
There's no reason that he couldn't try to find a way to generate income from his project, it's creating a lot of value for others.
Paying money to each other as a fair exchange of value is how most of the world works, and it works well.
Really? They're "established and entrenched products." I think they're a pretty good analogy here. Even if you are established and entrenched, you still have to advertise.
"Paying money to each other as a fair exchange of value is how most of the world works, and it works well."
Billions of people in poverty, and rising income inequality would say otherwise.
They are completely irrelevant to this discussion.
"Even if you are established and entrenched, you still have to advertise."
He has a widely adopted product without having advertised before. If he didn't advertise before, why would he have to now?
"Paying money to each other as a fair exchange of value is how most of the world works, and it works well." Billions of people in poverty, and rising income inequality would say otherwise."
Total rubbish. In the last 40 years, the number of people in the world living in abject poverty has been more than cut in half.
Once China ended their Communist insanity, and embraced free markets in the 1980's, they've lifted 100's of millions out of poverty and they are flourishing.
I'll gather you've never worked in a business role?
I think the dual licensing method would extract the most value and give the most of said value back to the developers.
That may be, but the point is to keep coders coding. The overhead cost of delegating fund raising to another person who isn't committing code is worth it.
>, and also gives those middleman organizations that do the funding some measure of power that i feel is also undeserved.
I think that it's often very deserved if the organization is well managed. If programmers can keep their hands on keyboards instead of rattling a tin cup then I think it's a better solution than the dual license.
On the other hand, it ensures the developer makes enough to keep working on it quite steadily.
Can I develop my proprietary application using the GPL version of PyQt?
Yes, so long as you have purchased the commercial version of PyQt before you distribute your proprietary application for the first time. (Note that this is different to the terms of the Qt commercial license.)
To copy from another reply: "I meant it would be convenient if there was some platform to maintain the licence contract chain so that a dual licenced project could use a third party dual licensed library, and the license fees would be forwarded to all license owning parties in a transparent manner."
I'm not claiming it would be a good idea. Just a thought.
Maybe there could be a startup that helps people monetise their free software output?
this is a YC-fundable startup. Here are the projects I can think of:
1. Celery - jeez. It runs half the world's high availability python software, yet has severe funding crunch.
2. Octave - obviously
3. Jupyter+Pandas - http://jupyter.org/
4. Vue.js - http://vuejs.org/support-vuejs/
People already pay big bucks to learn how to code, use that to fund the senior devs, the junior devs gain the experience they need to change their career trajectory.
This guy should not have been the director and doing the majority of the legwork for 25 years! He needs to be put in a position to get an awesome job at either classic big tech corporate, then some high prestige research unit. And there should be an understudy in the queue set to take over the work, and eventually earn his reward in the afterlife.
I mean, it's only OpenSSL which was for a long time the foundation of much of the online economy.
Or the various BSD's.
Perhaps the model needs support / services included.
Edit: or perhaps humble bundle is just more oriented to consumer goods.
Presumably we're past the point where anything with a .com in the name gets showered in VC money.
In open-source there is a huge difference between value generated and value collected. The people behind open-source projects are often clueless (sorry, but it's true) how to extract value from their work while keeping community happy. How about a service which helps authors monetize open-source software for some percentage?
Lots of open source projects don't need to bring on a business partner, but they might benefit from something like this. Just like not everyone on Kickstarter would go out and raise money by themselves.
And just because he is asking for donations doesn't mean he's against monetising.
And then, when they can't afford to work on their project anymore, they want our pity. They don't have mine.
More projects just need to use it.
Gratitude-ware is not constrained by what you are selling.. only by 1) what can pass through corporate accounting 2) how convenient it is.
Though it would be nice if the majority of people who intend to make money from opensource use Apache or MIT, however it is perfectly viable to make money from GPL.
This is the crux of the problem - creators generally think "can I sell X with Y license?" and scared and give up. Just look at Pfsense - it is making money using a single click support license contract.
Hell, people will pay for documentation. Make it corporate friendly and for please please generate invoices.
>> Gratitude-ware is not constrained by what you are selling.. only by 1) what can pass through corporate accounting 2) how convenient it is.
I've bought support contracts and promptly shredded the support contact information effectively to just make a donation for software we use extensively but is free.
Make it easy for us to pay you. Finance departments will not ask if the license or support contract covers all aspects etc so it's easy to buy those (PO number please!). But the last time I tried to do a straight up donation, I spend months before giving up.
I understand the intention, but you should have "used" support.
You don't need to be picky and call it for every minor thing, but calling attention to the author about a problem or issue that can be improved is valuable
This really isn't an option for a lot of things, and gives you some really perverse incentives to make things complex enough that they need support. Who's going to buy support licenses for /bin/ls, or most of coreutils, or most of libc?
You'll get people buying support licenses for MySQL and other difficult to configure & operate codebases, but not for the vast majority of code they use on their systems.
I work for a big company that runs a shit-ton of open source, we sponsor the two/three biggest projects we use, but by volume it's 0.1% of the total number of open source projects we use, at best.
This is a general problem for open source, everyone's using a huge long tail of infrastructure code that needs to be maintained, but any one company has no strong reason to support it.
There are details, especially with large projects that have many contributors, but it is done.
Since you can see the donation-amount in the blockchain, I'm free to say that "blockchains" are not the solution to this problem. People do not donate, no matter if via paypal or bitcoin.
Just an idea, not sure if it would work non-parhologically in the end (i.e. some mega troll buying the licensing rights for a set of critical open source libraries, for example ).
"10.000"
I reeeeallly don't like it when I have to think about whether my bench multimeter is reading 115 volts or 115 millivolts.
That kind of behavior needs to be called out.
Wow
However the sad reality is that outside a few unicorns, working in open source means being sponsored by a large corporation or via support contracts, otherwise better give up.
Both don't suit that well to desktop software for example.
This is against the GNU GPL license terms. One can't conditionally license some code under GPL. Ie, a software released under GPL can't be restricted for some purpose (here, 'for non-profit').
The "no string should be attached" is for good. Say for example, a project says "For people to help learn writing code, I'm licensing my project under MIT License," it would be wrong to not release the source code (and only the binary) even if the code is in MIT license. Because that is the purpose why the developer choose MIT for his/her project.
IANAL, but some corporate may be able to turn such carefully crafted words into money with their pricy lawyers. Most of the people are unaware of this. And for this reason, GPL can't be done the same way.
You can release code under however many incompatible licenses as you'd like, as long as you own all of the code in question.
You could choose to only grant GPL licenses directly to non-profits. They would be free to pass along the same rights to commercial entities, if they chose. It's possible that in a practical sense this would present enough of a hurdle for larger companies to prefer to pay (but plenty possible that it would not).
It is also possible to grant other licenses for situations where the GPL prevents some particular for-profit activity - that kind of approach is presently making some people some money, although the incentives can be weird.
However, everyone who has received your software under the GPL can sublicense it to the people who you denied the GPL to. You can't forbid it, but they don't have to, and you can nicely ask them to never sublicense it.
This method also only works if the copyright if fully owned by a single person or entity, with all code contributors signing a contributor licensing agreement to give away their copyrights to a single entity. This is problematic because now there is the possibility that this single entity gets bought out by an evil corporation (say, Oracle), which can choose to continue developing the project without releasing the modifications under an open source license.
As they are with the GPL so long as they don't distribute the modified code outside of the organization.
Contributing to the community is all good, but you know, eating is slightly higher on the priority list.
You don't need the draconian measure of dividing the OSS community, turning the idea of Free-as-Speech software into what's essentially shareware with source code access.
Large corporations have (obviously) vast financial resources. And – more importantly – they are (usually) staffed by humans. Have you ever felt the wish to be generous to someone whose software you used extensively? That feeling doesn't go away when you step into an office that requires a name tag to pass security.
Even mid-level managers or programmers usually can spend 4-figures easily without much oversight. And even if: I have witnessed a few cases where such requests went up the chain and came back with an added zero, despite HN third-favourite cliché being "companies have a duty only to shareholders and anything that deviates from 'greed is good' is a criminal offence'".
Sometimes it helps to offer a some sort of service contract, sponsoring or a workshop. They're bought as tokens to cover all eventualities.
So what's missing? Probably something like the FSF is doing here, handling paperwork and making it tax-deductible. I think it should be tied into some sort of coherent marketing strategy, kinda like the Apache foundation but more accessible to smaller projects, and possibly even a brand identity (/website) that doesn't trigger nostalgia for teletext. Mozilla is doing the best job in OSS marketing I can think of, but doesn't have a model for such projects. Maybe a Github foundation could also find success with a Wikimedia-style campaign. Although I'll hate myself for suggesting it if it ever happens.
Plus, unfortunately, a tiny bit of effort by developers. If you're just coding away at home and loving it, I can understand that it seems unfair, useless, out-of-personal-scope etc. that you should be doing fundraising. But it's also a bit too easy to expect good things to happen just because there's good code. (I have no knowledge of the octave situation specifically, and they may have done an excellent job and I'm wrong – this is just my general impression of some of these lone-coder projects).
Let's also not forget that OSS as a whole is probably doing better today than ever before. For every case such as this, there are dozens of cases where people are hired by large companies and largely left to work on their projects, or OSS created and released from within corporations, or successful startups with an OSS strategy etc.
Yeah, and "blockchain" isn't actually a magic spell that solves random problems...
You're asking to have your cake and eat it to: Give software away (altruism) but also feel guilty when he complains that he doesn't have any money. You can't have both.
You are likely to have a hard time to get external contributions with that kind of licensing.
There's this story from Freakonomics: A Kindergarten in Israel started to charge a fee for parents arriving late to pick up their children. Result: More parents would arrive late, because paying a fee was less of a burden than the bad conscience they had previous;y had when they made the employees stay late.
This would work in reverse. That good feeling I get when I contribute a patch for some random project I improved at the margins would be devalued by whatever $0.29 I'd get wired at the end of the quarter.
You hit the nail on the head what troubled me about my own suggestion of easy monetization but could not quite put into words.
Perhaps the few poverty striken maintainers are necessary outliers in this equation. I still feel irritated there can't be some simple opt-in value capture program for recognized bdfl's.
Sorry, my heart just can't ache somehow!
Only for Bitcoin.
- Octave won't die. It won't die even if you ditch it altogether. A lot of companies depend on it because licenses are free and that makes it easy to do massive parallel computations by launching thousands of instances. Matlab doesn't let you do that. As long as it stays GPL, the employees within those companies and academic institutions will continue to contribute to it. (Python is competition, but there are still plenty of Octave supporters.)
- You might consider working for a company that is built around Octave. They would likely gladly welcome you as well as support you in spending a percentage of your time to maintain it, since their business depends on it.
- Build something that people will pay for. Like a decent GUI for Octave that does everything Matlab's GUI does. Or a few awesome closed-source toolkits. That alone might support your personal income needs.
How many of those are there? I give talks pretty frequently where I start by asking people in the audience what tools they regularly use for technical computing, and I'm legitimately surprised the few times anyone says they use Octave for things beyond classroom-level work.
No offense intended to its developers, but Octave as a project seems like misdirected effort to me. Matlab the language isn't the most useful thing to try to emulate, Matlab the environment and tool set for getting non-software engineering jobs done is why people use it. For example, an increasing portion of the people who still use Matlab do so largely because of Simulink, for which there isn't a comparably effective open source alternative yet (you can sometimes use Modelica but there's less overlap in users than you'd think).
As much as I detest the MATLAB licensing reaming (and sometimes the inelegance of the language itself), there is very little out there approaching the sheer depth of the associated toolboxes and Simulink models.
When I see people asking if they can do this, they get laughed out of the room by others who claim that "open source freedoms" does not restrict commercial use.
This is kind of a serious issue. Artists get such a license, a high quality one (CC-BY-NC). Programmers don't. Programmers are not licensing law experts, they won't make their own, and if they do they won't do it right.
If you make it as easy to ask commercial bodies for money and still be OSS as it is to just build stuff for free, this story you mention will happen far more often.
That's true and I'll argue that's a feature not a bug.
It's also by design. The OSI specifically does not allow "for educational use only" and other types of usage discrimination. "The license must not restrict anyone from making use of the program in a specific field of endeavor. For example, it may not restrict the program from being used in a business, or from being used for genetic research."
IMO, this was one of the things that open source software licensing got right. Non-commercial licensing of open source is intuitively appealing and is also a great way to not get software adopted.
You're correct that CC, by contrast, does have an NC clause as an option. Even leaving aside philosophical points, I hold that this is a not a good license variant. Just about anything that isn't a trivial use has some commercial element to it. And what I might rationalize as non-commercial may well be different from what the copyright holder considers non-commercial.
It's probably telling that CC has been kicking around the NC problem for something like a decade at this point and AFAIK is no closer to putting a stake in the ground now than they were then.
See also: https://www.cnet.com/news/does-the-noncommercial-creative-co...
I long for a world in which someone can make a living out of open source work, similarly to those making a living out of games, mobile apps, youtube videos etc, without having to either work at a company or creating extremely restrictively-licensed software.
And very few indie game/mobile developers or would-be YouTube stars make a decent living out of it either however they license or sell their work. Same for most writers, independent consultants, etc. Being a freelancer is hard and doubly so when you're selling things that aren't necessities and have a lot of competition.
Unfortunately the AGPL has a chilling effect on using the software at all (see RethinkDB). I suspect the same would be true of any almost-but-not-quite-OSS license.
Apparently NPOSL addresses an issue with some non-profits making warranties, not to restrict usage to non-profits:
http://marlacorinne.4parkers.com/2014/06/20/non-commercial-o...
Source : book keeper for a students union for years, dealt with this shit with our tax consultant all the time
News at 11
Hardest part of dealing with American companies seems to be that they usually don't understand (or don't consider that) things that are different from the US (UK companies seem more open to this, still)
The fact that nobody pays invoices with checks outside of the US for example, or that different countries have different payment methods.
Finally we slapped their logo on our site for the trip, added the tax we would have to pay on the money, then sent an invoice. As our tax consultant explained to me, they can reclaim the added tax through their own tax return, resulting in a net exchange of exactly 1000€- whatever the labour costs of everyone involved.
Now, this trip was not too bad, but possible complications are :
* What you get sponsoring for needs to be in proportion to the amount of money, leading us to need way more volunteers than originally planned just to deal with the overhead of the sponsor services in a career fair we organised * If you are tax exempt for charitable purposes, you need to very cleanly separate donation funded things from sponsorship or otherwise business funded stuff, because subsidising a lossy business is only rarely legal
German-cultured(DACH) companies get the idea of "give money as a donation because what they do benefits us" a lot more than Anglo-Saxon cultured ones, at least in my experience, which might be reflected in the law
It might be that the legal framework is there, but the process within companies is not. I'd be interested in someone correcting me if I misunderstood things
I think the critical different is that the sponsorship is for a specific activity, but I'm not sure.
There's certainly tax structure in place for donations in the US, but perhaps there is also a custom regarding sponsorship.
So, you want Octave to sell support contracts because accounting will never do donations. The two examples aren't selling just stooge licenses. In the case of pfsense, they're selling support. That means changing what jwe is doing from developing Octave to handling support calls. In the case of sidekiq, they're actually just selling plain ol' non-free software, which is what open-core is.
jwe will never want to sell non-free software. He's already selling some support and has a few customers, but it's not easy.
I often have heard engineers (who seem to sound angry or offended with the concept of a donation) say things like what you are saying, but I've never figured out how to satisfy them without either (1) just plain selling non-free software or (2) completely changing the focus of the work from developing Octave to doing some mind-numbingly boring task like getting Octave to run on RHEL 5.
In sidekiq pro case, the "support" is a right to send email. If i may say so, that scales with the money you receive.
How to set it up? take a look at sidekiq - they use https://plasso.com. There are tona of alternatives in this space.
More importantly, we are here to help. Please don't hesitate to ask. Also take a look at indiehackers.com to see how others did it.
lastly - we are not angry.. we are frustrated because we have the ability to help. But open source software creators must empathize with procedures and stupid stuff like accountants. Unfortunately thats unavoidable, no matter how much you want to.
Sidekiq pro is just selling non-free software. It's not just the right to send email.
My second idea was to move to an open core model: hold
back more complex or enterprise-specific features from
the OSS version, sell those features as an "expansion
pack" on top of Sidekiq. Thus, Sidekiq Pro was born.
This proved to be popular and forms my business today.
I can't tell about handsontable, but it looks like again it's just plain ol' non-free software, although they do seem to let you look at the source code under what sounds like a restrictive license.
pfsense also seems to sell a lot more than just an ebook. They sell backups, what sounds like another non-free addon "plus many more features". Maybe most people don't use any of that extra stuff?
None of this sounds like it's just a matter of selling stooge licenses, and the sidekiq guy even said that selling stooge licenses didn't bring in any noticeable revenue.
It's a donation in all but in name so that the accounting department will approve it.
In all the debates about MAGA or how to invest in the country I don't think i've ever heard someone raise the issue of funding importent open source software (only indirectly).
Actually for any country that can afford to invest, it provides so many benefits economically, competitively, academically, and strategically. Best of all it can boost macro productivity, which benefits everyone even if they have nothing to do with tech.
Even lowly web development which people love to bash as changing too much. you can look at all of it as creative experiments in engineering that eventually end up advancing productivity.
Who handles PR for open source anyway? Shouldn't a professional communications person work on building awareness, explaining the value, advocating for the cause?
ALso key is to locate Octave within the solution space, e.g. Startups who can't afford matlab. High schools. A formal online course. Makerspaces. Job seekers wanting to work with matlab.
That said, I would understand if he didn't want to try an licensing like that for fear of angering The Mathworks.
With pypy, funding goes towards particular projects (e.g. STM).
This may not be suitable for all projects or developers (libjpeg turbo being an example - he wants the freedom to be research lead and work on what he likes).
If anyone, then the FSF should be looking into this problem, study all the models used, which are successful and provide some sort of template model for people to use.
Licensing may be an issue, can you do the Sidekiq or Wine model when your code is GPL ?
The solution is to just pretend you're buying a minimal support contract and that'll get approved no worries.
Selling open source is like selling any other enterprise software - you need to prove value.
An alternative is to sell support (like pfsense).. which is pretty much the right to file bugs. We also buy a license for Handsontable Pro : we get right to file bugs and the fact that we get new features 3 months before the free version.
Its just a way to package gratitude-ware to pass accounting inspection.
...which you get through the FSF donation link he provided.
It's amazing that this is the top comment.
I do corporate accounting for baroque software and I can write an invoice and pushing people to pay for it is still quite a sales challenge - you need to be able to talk to high enough people and prove the value of what you're offering on top of the software you are already giving for free. It's not different to a standard issue sales exercise, except a little harder. This is not a skill that most open source project maintainers would consider themselves experts at.
Giving advice like that is very dangerous - "only if you did that, you would make money". It's not a given that it would lead to something more than couple hundred dollars.
Free and open software has created the expectation that software has no cost. Trying to overcome that after the fact will take time and education.
There is a community site I frequent that keeps itself afloat with a freemium model (if you pay you get more features; but you still get 98% of features even as a free user; there are no ads on that site). Many people understand this and pay to keep the site running, but something like 10%-20% of users claim that it's just a money grab because "facebook offers essentially the same features and they can do it for free, so obviously this site is trying to make a profit on our back".
There is expectation that because some software is free, most software is free. Similarly, there's an expectation that if some websites are free, most similar websites should be free as well.
p.s.: the main cost to using Facebook in my opinion is privacy, and I value my privacy much more than what I get from Facebook. Obviously. other people value privacy and facebookness differently.
It might not be that hard if you're motivated and have that vision, but I can completely understand someone not wanting the added pressure after all this time.
Why?
If you want money from companies, you have offer the trinity that suits their payment rules and habits: An order, an invoice, a payment.
What you sell doesn't have to make sense to the beancounters. They're used to that. But it has to be something they're able to order and pay for.
Having a proper invoice (client address, your company address, VAT number) makes life significantly easier for accounting department and thus makes paying money significantly more likely.
(There is also a lesson here for other US-based companies and conferences that can't produce a simple paper saying "Invoice" and a pair of addresses. Yes I'm looking at you 1Password.)
For a realistic example in Italy, having a VAT number implies a minimal fixed yearly cost of ~300eur (combining all various expenses) even for zero income. If you already are an employee, any extra income is likely to be taxed higher and require extra bookkeeping, to the point that cost parity is only attained if you can invoice for at least 450-500 eur.
This is purely wasted money, without considering the extra work that you have to put into for the tax records.
It's simply not worth it.
For any reasonable company (even a single founder one), 300eur is nothing
With this limitation, many developers simply do not (and rightly so) put the effort in accepting anything in that ballpark because it's just extra work.
I'm pretty sure in most EU countries the profitability of a service supported by patreon (that is, a passive income in sub 10k range) would be almost nullified if correctly accounted for taxation.
I had long discussions with friends about legally accepting money for side projects. It's a very hard topic to discuss.
I'd be curious to know if anybody in an EU country is able to legally accept and declare a side income and retain anything more than 30%.
And yes, I know that VAT isn't required piece of information on all invoices, but it WILL make the payment go smoother through the accounting departments which is what you want ;)
AFAIK the big problem with VAT, from a seller perspective, is the fact that most governments, knowing full well that 90% of businesses fail in the first 2 years, ask for big upfront payments on estimate VAT remittances in that window, to make sure they don't lose out. After 3 years, as long as your accountancy is correct, it should be smooth sailing.
There are other way to start companies that are not directly attached to a person, but this involves a much larger upfront capital, which is designed to avoid exactly this scenario.
But you cannot give a paypal link and expect donations. As a company, I cant do that. I need an invoice. Hell, if you can get a business account, I daresay you will get subscriptions.
I like to call this "gratitude-ware".
Check out Sidekiq Pro and their experience making 80k USD per month. https://news.ycombinator.com/item?id=12925449
At first I built Sidekiq as an LGPL project and sold commercial licenses for $50. Revenue was laughably small, but the response I got was encouraging: people told me they were saving $thousands/mo over previous solutions and wanted to buy the license just to give me something as thanks.
Octave currently offers support packages for which you have to write in to the maintainer and have an email discussion. Compare that with Sidekiq : http://sidekiq.org/products/pro
Its one of the best designed gratitude-ware page...even works amazingly well on a mobile phone.
We personally also buy pfsense licenses https://www.pfsense.org/our-services/gold-membership.html
TL;Dr Donations wont work. Engineers cant give an excuse to corporate accounting. Make a pro subscription with ANY "pro level" feature. I can get my accounting to sign off. And no "contact us to find out about support contract".