It's already hard to build a successful business. Odds are already thin for you to succeed. Why choose a model that worsens it?
Open Source was not created as a funding source to for-profit ventures. That's what these businesses are doing: raising labour funds for free to fund a profitable product. That's not the spirit and the purpose of Open Source.
You are misguided. Here's what the company that employs me does:
- develop an open source product. Open to contributions, but most of the dev is done by employees.
- sell support and consulting on this product
- sell pre-packaged open source extensions to this product, with support
Totally withing the free software / open source spirit. The world gets great open source software for free (not open core: truly and completely open source), and the business behind it is sustainable as well. We also donate to some open source projects that we use, chosen by employees.
At some point, if you want open source to take over, you need it to grow and strive, and having it supported by business is a great way of getting this. The world probably cannot be run only by side projects.
Not saying that all open source companies behave this well, but it's not an impossible outcome.
1. Open source product
2. ???
3. Profit!
Open source isn’t a magical bandaid though it may get you some early traction, the same you can get by having a free plan.
If you develop it appropriately and plan it out, it CAN be successful. But you have to think it out and choose an appropriate license, etc.
Most successful open source companies would have been successful as closed-source; and you rarely can take a closed-source product and make it successful by going open source.
I'm with you on this. You need to carefully plan. I also strongly believe you should not go with VC money. That excludes "burning money and we'll figure it out later". Your open source project will therefore probably start as a side project and/or with government help dedicated to business creators.
> Most successful open source companies would have been successful as closed-source
I believe this too. Now, there are more and more things that require open source products. Open source companies, including the one I work at, have received funds from openDesk [1] to develop a complete suit of open source tools. More and more research labs, universities and public institutions require open source. So there's a dedicated market that's growing.
You may succeed as a closed source company, probably more easily than as an open source company. But the story is not that clear cut, and it also depends on what you want to build and whether you really want to build that awesome product as closed source.
At some point the incentives become misaligned, though, we've seen this happen over and over.
Someone proposes and sends a PR making huge quality of life improvements, it's shot down because it would reduce consulting revenue (the rejection reason is officially not this or not even stated at all).
Someone proposes and sends a PR implementing advanced functionality that the core company sells as part of a paid extension, the PR is rejected (for obvious but not mentioned reasons).
A cloud provider starts offering a service based on the software, the license is changed to non open source.
Plus, these kinds of businesses generally only scale to mom and pop store or maybe 100 consultants.
Of course I would indeed expect a contribution that:
- tries to disable our licensing mechanism in our paid extensions
- tries to copy our paid extension code to our product or to a community extension
To be rejected upstream. But isn't it fair game? They can still fork the code and go start something on their own, it's one of the important features of open source. One can start a new extension repository and convince users to add it to their config, or even fork the product and provide this by default. But this is work, they'll have to maintain this, and will probably depend on us, in the end.
Most likely, we would welcome significant contributions and QoL improvements, we are only so many people for that much work to do, improvements would allow us to focus on something else. If someone outside the company is willing to maintain some feature that we currently offer for a price, it probably actually grow / improve the ecosystem and it would probably be welcomed, letting us focus on other things our customers want. People maintaining extensions or feature for our product would likely be friends. Of course you can't just dump some code and go away, that's not sustainable.
> Plus, these kinds of businesses generally only scale to mom and pop store or maybe 100 consultants.
That may be true. My company reached 60 people this year. Nextcloud is around 100. But not everyone wants to become too big. There are a lot of advantages in staying small-ish. For the CEO, a big company is also not the same fun as a 50 people company.
for good software consulting revenue is not a function of how many bugs/bad UX there are- you consult on setup, integration,maintenance,etc. If the software improves the need for this does not disappear.
If you communicate clearly what the proprietary parts of the software are, it also shouldnt be an issue. If someone wants to implement and especially maintain and test them themselves without paying, they are free to fork and do that.
Income is mostly from support and consulting. These two things still pretty much rely on the actively developed product, so a big part of the company is dedicated to this. I'd say we are both a software and service company. I believe this is one nice way of developing open source software.
No insult taken :-)
The thing is that support services aren't scalable. Many software businesses want a scalable source of revenue, that's why they go for cloud services.
They don't want to compete with copy cats, though, because they want monopoly-level margins.
Then they complain that Open Source won't let them have those astounding margins.
This is not an issue with Open Source licenses.
We also have this.
> just that Open Source was not designed with your employer use case in mind
Yep. Free software was designed with user rights in mind. Open source was then designed to sell the idea of free software to companies, removing the "user rights" parts, which sounded frightening to businesses, focusing on the development model, targeting developers and not users. Free software is defined with the famous 4 rules, open source is based on the open source manifesto, basically a copy of the Debian Free Software Guideline (from the Debian project, which, coincidentally, was closer to the Free Software Spirit at the time, probably still is). They are both about the same set of software and licenses, for the most part.
There's nothing in those texts / philosophies about doing business, for or against (only that free software was explicitly designed to allow selling software, because why not - and at the time, to distribute software, you probably had to copy it to floppy disks and give them, so you had some copy cost to absorb). It's indeed up to businesses to figure out the business model around FLOSS.
But that's true of many things, isn't it?
Presumably jraph's goal in writing their comment is to convince us that starting an Open Source company is a good idea -- but they haven't succeeded, for the reasons mentioned. If they want to convince more people, then they should come up with counterarguments to the top level post's arguments (and if they don't want to convince more people, that's fine too).
No, I haven't succeeded because I haven't tried yet. However, I am surrounded by open source companies that succeeded.
The comment I responded to obviously reasons theoretically, and confuses some things. I will answer two aspects of it:
> Open Source makes "competing" with an existing company trivial, but with none of the invested costs. So the first mover, the program author, is always at a strategic disadvantage
That's not true. You can't just take some open source program developed by someone else and make money out of it. Let's say I develop open source product P. I offer support and consulting. Let's say you want to compete with me on P. You initially lack the expertise and the notoriety. If Alice and Bob want to get support or consulting for product P, they'd better turn to P experts, which is the company that builds P, me. For you to get customers, you need to prove the world that you are an expert. It takes spending time with the code, and also on marketing. But while you are spending time on this, I am too. You will also fatally need to contribute improvements to the product you are selling to make your customers happy. Actually, if you start contributing, it's a win for both of us and our companies can even be friends.
I'm not making this up. That's actually where I work.
> This is not an accident- it is baked into the very point of open source
So we just countered this. I'm not saying there's no risk. but that's not "by construction".
> There's a reason that very few people in the bazaar actually make decent money. There's a reason the cathedral has treasures.
And now we are confusing development methods, not "open source vs not open source". You don't need to be organized in "bazaar" to build free software. Look how, for instance, SQLite, as open as you could dream, manages very well with their non contribution policy. They are profitable. They have a "first mover advantage".
> These are two pretty distinct concepts, and the (traditional) motives for those two things don't merge terribly well.
Well, they can.
But my biggest argument is that you don't need to believe me. Many "actually open source" companies have succeeded. So, what gives?
I don't need to actually build an open source company to prove my point. Others have been doing well. Maybe me in the future, that's not completely excluded. Thing is, being employed also has advantages.
>For you to get customers, you need to prove the world that you are an expert. It takes spending time with the code, and also on marketing.
I agree with this, and that the makers of P likely enjoy a big (and genuine) advantage in being considered the authority in all matters P. But for a competitor (me, in your example), spending time with your existing, working code is much less time and effort than writing brand new code, and that difference in my effort levels is the disadvantage, for you, of going Open Source vs. closed-source.
>SQLite, as open as you could dream, manages very well with their non contribution policy
I wasn't aware that SQLite was a profitable enterprise. I have to admit that this is a very strong example for your case, given the competition in this area from other FOSS software (PostgreSQL, MySQL).
There's also possibly a small % of users who end up knowing enough to end up doing "mostly supporting others". They can end up becoming competitors as well.
So yes, this model can work, as its currently doing for you (and others). But it certainly is a lot harder to build, and keep such a business going.
Being an employee of such a business is great. You still get a paycheck every month, so the model is irrelevant. Owning such a business though is precarious - basically the staff can walk anytime, and effectively take all the customers with them.
indeed
> just try
I did say "please" try, the nuance is quite important :-)
You know, I think the views differ because we have two problems:
1. How can I create a profitable company
2. How can I make my open source work sustainable
If you only target 1, open source might not be the more immediately easy solution.
I'm aiming at 2, and 1 is one way to do it: given I want to live from open source work (because of my convictions), can I do it by building a company?
The answer is: it might be harder, but yes.
If you don't care about open source, my comments here are just mostly annoying.
Of course there are successful examples of just about every model under the sun. If you can find a model, and a business which allows externalities like Open Source then that's fantastic.
And yo be clear I'm not saying companies fail because they are Open Source. Mostly they'll fail for the same reasons every other model folks- they don't generate enough cash to pay the bills.
Indeed Open Source can be an attractive start. There's at least a chance of free labor. There's some marketing milage.
The problem is not the start. The problem is when the money starts flowing. When you're "surviving" no-one is really incentived to muscle in. When you start "thriving" then suddenly your space looks really interesting to others.
And licensing one's product seems to amplify that risk.
> based on a few opinions of HN comments
What if someone does do that? As a HN commenter I can't bypass the possibility that somebody takes my comment at face value without conducting proper research and analysis.
HN comments (and really all content on the Internet in general) should be written responsibly. I am not saying that GP didn't write their comment responsibly.
> nor it's their responsibility to outline the full extent of risks involved
But then you can't raise complaints when somebody does outline the full extent of risks involved.
Someone not conducting proper research is doomed to fail at building a company anyway.
You should be careful in writing HN comments but you should not feel responsible for a business failing. Someone reading should also read comments from anyone, one's writing does not go unchecked for long on HN.
