Open Source Doesn’t Make Money Because It Isn’t Designed to Make Money

Unfortunately most of the wealth created by OSS is being reaped by the large cloud vendors because relatively no-one pays for OSS but they do to host their production systems on it, atm this revenue goes directly to the major tech cloud companies which are using it to build out their multi-billion dollar infrastructure moats, ensuring a barrier to entry no other will be able to partake in.

This is why cloud vendors have become so "OSS friendly" - they love the status quo as they've become the primary beneficiaries from everyone developing and distributing OSS for free as long as they're able to collect the rent when it gets hosted in the end.

You're starting to see the conflicts as more companies like Redis Labs, Elastic, Confluent and MongoDB who realize this and start distributing their future investments under "OSS free that's free for everyone else but can't be used commercially by cloud vendors" as a way to force licensing so Cloud vendors are unable to use their own investments to compete against them - and force some revenue share back from when Customers pay for OSS, when they host it.

So when you hear how much cloud companies "love OSS" and how their initiatives like "Open Distro for Elasticsearch" is to ensure it stays open, know that they're not doing it out of altruism, it's to repackage Elastic's OSS investments and make it available for free so they can be hosted on AWS without having to host with Elastic, it effectively lets them avoid licensing and revenue sharing back to the companies who've invested their time and resources into developing it.

We don’t build the code ourselves. We do own, review and manage the code. We handle the project management and we facilitate access to our business processes and employees, but the actual development is something we buy.

Once things get build we often end up buying the various projects as SAAS solutions because we aren’t really geared to operate OS techs. We do have excellent federated authentication, network security and API services, but we genuinely don’t want to operate a horde of different tech stacks. This makes us a perfect match for anyone who knows how to do SAAS in the cloud though, and most of the things we build end up being run in Azure or AWS by a supplier.

So far it’s been a really beneficial relationship. We get ownership, better quality and faster more agile development. Our local development houses still make money, both from development but also from SAAS.

It has been especially good for our smaller dev houses, because a system isn’t locked down to a single supplier for years.

This last bit is entirely my personal opinion, but I prefer buying software from a local shop and helping them build and expand their businesses. On the plus side of this, they co-operate a lot better too. Municipalities are trying partners who can’t make up our minds, and a small agile shop who doesn’t point to the 4 year old contract every time you want a tiny change, has just been so refreshing.

First, this isn't true. Plenty of people make a living doing each of the above.

(Art is somewhat of a special case, because some will say that what you're doing isn't art by definition if it's attached to some functional purpose like a business model. But you don't have to listen to those people.)

Second, these things aren't cheap because they're not valuable. Water is cheap. Water is also necessary to life. It's immensely valuable. What makes it cheap is that it's easy to provide, so competition drives prices down. People aren't going to pay a lot of money to get something valuable if they can pay less to get the same value elsewhere. Value isn't the only thing that matters when determining price.

> There are dozens (hundreds?) of successful open source projects that have tried to become even just modest commercial enterprises, some very seriously. Results aren’t great.

Are results much better for normal people trying to start businesses that aren't based on open-source code? I'm not convinced they are. Businesses are hard.

Intellectual property can only make money by preventing its use, otherwise it just spreads everywhere to everyone for free. Given this property of the Internet, and the decision of open source/Free Software to embrace rather than fight this property, what side-channels can be constructed to return some value to the developers?

> Anything paired with a physical device. People will judge the value based on the hardware and software experience together.

Not just embedded, but also SaaS. The fact that the physical device is off in a datacenter somewhere isn't so relevant. (Another implication of this is that open hardware will never behave quite like open software!)

One of the main problem that we technologist fail at is that a lot of the times we assume technology == product. When we do this, we completely overlook all the necessary infrastructure and processes that we need to have in place to capture the value that the technology provides. We need customer success to guide the customer through a continuous successful implementation, marketing to very clearly articulate value propositions that change in different verticals, a sales team to capture the value and turn it into dollars consistently, support to listen to the customer when things go wrong and hold their hand through the fixes or workaround, lawyers to prepare support contracts and cover procurement requirements, hr, etc etc. I could go on forever.

I disagree that there are not examples of companies that show how to make money from open source. Beyond the obvious examples like "RedHat is a public company" and "MySQL sold to Sun", I can think of an absurd amount of businesses whose secret sauce is ElasticSearch. I would argue, that even though Elastic is a public company, they have not been capturing all the value they have brought. Although there are consulting companies like EnterpriseDB and CitusDB around projects like PostgreSQL, thinking that "consulting" is the one and only business model around monetizing an open source database is a gross over simplification that glosses over models that are evolving in various other areas, for example the one Jetbrains is doing with IntelliJ (basically a subscription service now).

I can look at the current top 100 posts of HNs and point out quite a few of successful businesses that would not have been where they are right now without open source existing in its current form.

Isn't it odd that a piece of software, could be widely deployed, forked, starred on Github etc. etc. and yet with all this infrastructure there is no mechanism to ensure that any money finds its way back to the creators unless they ask for it? Some kinds of business people praise the selfless creativity of the open source community while simultaneously conserving their own profits and often denouncing taxation as theft. When challenged, such people attribute all human progress to the profit motive, and thereby in part to themselves, in manifest disregard of the facts. Creative people don't like this but they need to eat and keep a roof over their head, which requires money.

On the one hand, creative people with a mindset of mutual aid and consequent abundance, whose genius expands and structures much of our economy. On the other, well-resourced people who preach a gospel of scarcity and systematically give less than they take. These same people denounce any interference in their profit-making, and choke the unfortunate with externalities because they refuse to bear the cost of cleaning up their own waste, often going so far as to deceive the public about its existence.

How are open source creators any different from blacksmiths, forging the very chains and weapons which bind and oppress them? How much life must be given away, how much filth heaped on heads, before creators will unstop their ears to the laughter of those who exploit and rule over them, mocking their generous spirits and diligent efforts?

I think Open Source really works for infrastructure projects and it's possible to do really well either working on a system like Kubernetes or Elasticsearch.

Not so well for projects that are more focused on consumers.

My app, Polar, is focused more on consumers and is an app for people to managed their reading and notes. Similar to Evernote but also has spaced repetition support like Anki and we're working on collaboration features too:

https://getpolarized.io/

I wanted Polar to be Open Source but we've had almost zero source code contributions in the last six months.

Most of our contributions from users have been from people suggesting features, filing bugs, suggesting architecture changes or product/market fit issues.

These have been very helpful of course.

However, to date, we've only raised about $500.

To be clear I'm working on Polar full time so I've probably made about $0.15 per hour...

I think there's a lot of reasons for this actually but if I were to narrow it down.

1. Consumers hate paying for software. The plan with Polar is to go with a freemium model so we need 100k or so users.

2. There's a big of a 'tragedy of the commons' here with the donations aren't coming in because everyone assumes someone else will donate.

This is really a bad situation in a lot of ways for the 'self hosting' crowd. They want quality apps but they're not willing to pay for them.

We're polling our users about why they haven't donated and here are the current results:

https://i.imgur.com/cOUy10f.png

I think my next main goal is to do some sort of Kickstarter and try to raise $150k or so.

Either that or focus on raising grants or VC.

Those donations could then be used to hire full-time developers, and work on the kinds of software that is currently not well served by open-source (TurboTax replacements, games, office software).

Just as with public radio or podcasts, the price signal of donations would encourage certain kinds of products to be produced according to demand instead of enjoyment or itch scratching, but the end product would still have the benefits of open source, no license keys or DRM to worry about.

I expect 99% of people would freeload as they do with public radio, but that's ok, even a small participation rate could result in new kinds of open source software being produced, and a sustainable career path for creators.

It's just that selling like this is more difficult now that there's a high-speed internet for everyone, but I think more people should be trying to sell this way. App stores are one example, if they didn't have those ridiculous GPL-incompatible restrictions, but Red Hat does outright just sell the software in a GPL-compatible way: you want access to their repos and nothing else, no support (I believe they call this self-service), they sell that too.

It might not make you rich, it might not hockey stick, but I think if more people just tried to sell free software, we'd see more people being able to comfortably make a living.

Of course orgs have lost site of this. The sad reality is 99.999% of orgs are OSS freeloaders or way too conservative and hence want to in house everything. Or happy with modestly priced cloud hosted everything that ends up taking away some of OSSs maintenance savings upside.

Only through legally enforced IP protections do people even pay for classic proprietary software - it's a kind of artificial scarcity (there's no moral judgement there, just an observation). But yes, hosted services and so-on are also rare (it costs more than microcents for someone else to set up an equivalent hosted service).

> Another lens might be: who are you selling to? Classical scratch-your-own-itch open source software is built by programmers for programmers.

Yes, open source works well for software written for programmers as an audience. If programmers aren't using the open source software, then there usually needs to be a business that's supporting the development of it.

And they talk about how it harms the development community...exactly how does it harm the average OSS developer if Amazon or Microsoft or Google are adopting the project and making it easier to deploy in cloud scenarios?

Seems like a 'have your cake and eat it too' scenario. These companies benefitted from massive adoption thanks to open source being developer friendly and free, yet they want it to be just restrictive enough that only their company can monetize it.

In Mozilla's case, I would love to see some sort of cloud hosting for Rust. I also wish I could write front end in Rust but maybe I'll just have to wait for webassembly. I can't imagine that some sort of rust to js compiler (kinda like kotlin) would be too hard and this would be a killer application.

Matei Zaharia (of the Spark fame) and friends are working on weld, https://github.com/weld-project/weld. There's going to be a lot of money there as well. A cloudera competitor of sorts.

But yeah, go ham on Rust based paid solution.

Do not release the source until the fund is raised.

People pay for scarcity and our whole economic system is based on it. Open source unfortunately is the exact opposite.

Note that there are open source projects making money. For example Wordpress is GPL, which forces all plugins and themes to be GPL-compatible as well. And there are plenty of commercial plugins and themes selling quite well, all of them open source.

But that's not necessarily something Mozilla could emulate.

This statement is too general and obviously false. Many artists make money. Of course many or most don’t but “can’t” is a very strong word.

The money involved was not huge but it was great to have. Users did not care that the software was OSS. They cared that is solved a problem and the ebook I sold made it even easier. That is how OSS makes money - by solving problems.

OSS is not a business strategy, it is just a software license. You do not "make money" by adopting a software license, whatever it's moral merits. You need a business that solves a problem and OSS can be part of that.

Instead we have smartphones, computers that function as pocket strip-malls with addictive games and back-alley scams.

Agree / disagree?

In my last 3 gigs, one of them was fully opensource (the Madrid City Hall decided to build a direct democracy platform) and the two others have been "open core", where there was an open-source component accompanied by an "enterprise offering" which was closed source, with extra features. Money was made by selling this "enterprise version", plus support, to other companies. This money pays for my salary, which involves mostly the open source side of things. The Open Source version is a "foundation" and a very good "marketing tool" amongst other developers. There's an ecosystem built on top of the open source version and we have some pull requests from the community. The main drive behind new features is still our internal dev team.

The target-audience of the adds would of course be coders because who else reads source-code. Maybe not a big audience but the adds could be permanent.

And what would the donors be advertising? Maybe development tools but also themselves as a potential employer. It would be basically VERY targeted marketing because only coders would see it.

The reason this works is that once you have a large user-base you can get people to buy your services because you have now developed a stickiness with your customer-base.

Exhibit A: Hashicorp. People started using Vagrant, the user-based expanded. When they moved onto other pay-services, people followed.

The reality is that developers want 'free' stuff they can build with but IT/Ops wants 'pay' services where they get 24/7 support, consulting, etc etc.

No company will be fully OSS because no money = no company. However I can see many orgs support OSS as a customer engagement tool to drive their brand and get a user base that they can drive towards other products.

Especially business people who resell the code to their customers without ANY investment into actually creating something or paying people with talent who do things that make the world a better place.

Why would the world properly value our contribution if WE don't value our contribution?

Most OSS business models don't try to sell the actual OSS. Usually a separate, commercial product is built on top of the actual OSS.

> You can’t make a living making music. Or art. You can’t even make a living taking care of children.

For everyone else smashing their keyboards, I think his point is: It's hard to make good money as an average artist, whereas it's easy to make good money as an average software engineer.

> This is where the Free Software mission has faltered despite so many successes: software that people actually touch isn’t free or open.

The point is that those commercial applications almost always depend on FOSS. If you have a better idea, go and build a better commercial product. You'll probably depend on similar FOSS.

The problem with this mechanism is that now each project is out on their own to advertise their needs and get donations. I'm actually experimenting in this space with SeedAndDew (https://www.seedanddew.com). My hypothesis is that if you can give people the ability to set a specific amount to open source that automatically goes to the right places, more people will contribute. Just went full-time to see this through so early days but I'm optimistic.

Even if there was 100 free forks readily available. People would pay for the brand :)

Honestly, I prefer bitwarden over "password store" because I pay and, thus, I know it's unlikely to go away or go unsupported.

Basically, you can increase the demand for a product by decreasing the cost of it's complements. You can increase the demand for managed Redis services by investing in making Redis better. The caveat is that you can't control where that demand goes and you might end up helping out your competitors more than yourself.

I disagree. Some of the very successful FOSS projects are backed by businesses that benefitting using those. Linux, RocksDB, React, Tensorflow would be good examples.

When going from free to paid, consider the probability that the only reason people use your service is because it's much cheaper then the alternatives.

Our current economic system focuses on money as a medium of "equivalent exchange", as in there's no free lunch and a dollar you spend on buying a piece of butter necessarily means you can't spend that same dollar on a gun (same thing for the butter merchant). This exchange is based on a contractual agreement that binds for a finite time, and stuff that happens outside that window is necessarily considered an "externality". In other words, money facilitates a social interaction that prioritizes focus on the costs/benefits of individual(s) involved while de-prioritizing the costs/benefits of the collective at large.

But contributing to open-source is not that sort of social interaction, instead, it's one that prioritizes contributing to the collective over individual gain - i.e. a social interaction that focuses on externalities instead of on contractual time-based exchange. Money, with its focus on being loosely "conserved" (e.g. you can't double-spend the same dollar), is simply not a good medium for encouraging cooperation and discouraging free-riding to open-source (and other public-goods)

In traditional villages, public-goods contributions were facilitated with some combination of reputation and gossip. Later, cities and states managed to come up with taxation, courts, and prison when reputation and gossip failed to scale. But the super-connected capitalist "globopolis" of today demands a new form of organizing cooperation and eliminating free-riding when it comes to public goods.

To properly solve the problem of the economics of open-source, we need to solve this more general organization problem.

Implemented as a charge added to cloud usage costs.

Terms of use and terms of agreement machine readable and included in public available source code.

Code publically available, publically built and cryptographically tagged by usage and licensing Metadata. Like a smart contract.

Anyone who used the code would automatically be billed by the cloud provider on the license holder behalf.

This would align incentives better and enforce the openness of the source as a byproduct.

(the licensing should ideally be built into the platform, which should be vendor neutral, e.g. K8s and maybe the license could waive cost collection for self hosted use on own hardware)

NB: I don't mean more hours are required; I mean workers get paid more per hour.

Labor prices only seem to change with supply and demand. If you're making lots of money, and labor is cheap, why pay more? That's where a minimum wage helps shift the allocation.

You could shrink the Labor supply, perhaps artificially; or you could increase demand for Labor.

You could directly legislate by license: the license available only to businesses that e.g. pay at least a minimum wage.

Of course, businesses won't like that, so your software had better be something very useful.

Services > Support > Packaging > Subscriptions > Open Core > Managed > Cloud > SaaS

Basically they open source a core component that doesn't compromise the business model. For example, only open sourcing an engine but not plugins, modules, or dashboards around it. They also open source the core once they've already gained traction which puts them way ahead of anyone trying to do the same thing they are.

I did this with https://vcvrack.com/ and it has been successful IMO.

Of course there are bizillion of projects but they are just not as popular or not used for anything that generates a lot of money.

I'm sure when you are a core contributor of some popular Open Source projects, you'll have no problem doing well-paied consulting work or get a good job that allows for Open Source work within working hours.

OSS exists in the spectrum of a gift economy. Once you give it away, you don't get to expect something in return, or expect them to use it in a particular way. You gifted it away. In fact, the original OSS definition explicitly stated you could not discriminate against fields of endeavor. But I guess envy is stronger than idealism.

I see way too much focus on the value of whatever they're selling and no focus at all on what else I could do with that money. I'd gladly pay $8 or $10 a year for a good bookmarking service, but not $45.

It's like complaining that insulin isn't fun enough. Nobody taking insulin was trying to have fun.

The best we can do is to work within the dilemma. That's what https://snowdrift.coop is aiming to promote and support.

What?

RedHat - $34B

MuleSoft - $6.5B

Elastic - $5B

Cloudera + Hortonworks - $5B

Pivotal Software - $4.7B

MongoDB - $3.5B

Then you could make money from big companies, and let individuals and small startups use your software for free.

Commercial software that is lousy and buggy is not.

Free, open source software the requires an army of consultants to make use of it - is not.

OS Hardware can very much make money (hear Adafruit, SparkFun, Pololu, Seeed Studio, of course Arduino and to an extent Raspberry Pi).

Great quote; tucking it away for later. This is why democratic socialism is essential. Money is a lossy codec for true value.

If APIs are copyrightable that's a very unfortunate thing, but it would also make it possible to protect FOSS IP much more aggressively. Mongo wouldn't be getting destroyed by Amazon, for example.

Some examples include Google with Kubernetes, Netflix with their cloud tooling (Eureka, Ribbon, etc.), LinkedIn with Kafka, Yahoo! with Hadoop, and many others.

It's really annoying to make a typo, realize it, try to correct it, and ending up hitting return on a malformed url AGAIN because the site finished loading in the meantime.

They've been playing serious business for some time now, but it's pretty much been borrowed feathers.

Firefox got big because of early enthusiasts who believed in a better web. That got them enough clout to get google interested, and google basically gave them large piles of money.

Somehow that gave them the idea they were a "real business" now, and they started spending and reorganising themselves accordingly.

Now it turns out they probably weren't a real business after all. Those piles of google money in the bank could have sustained pure Firefox development perpetually, probably even allowed for fun and useful side-projects like Rust, but no, Mozilla wanted to play serious business, except, of course, now they have to compete against actual serious businesses, which in this case would be Google and Microsoft, and so far their efforts haven't been a unanimously successful.

> This is where the Free Software mission has faltered

No, this is 1. where Mozilla has faltered, and 2. Focusing on money, marketability, and business friendliness is pretty much an open source objective. That's fine, but don't drag Free Software into this.