These are two pretty distinct concepts, and the (traditional) motives for those two things don't merge terribly well.
Over and over we see the same story playing out. Companies need to make revenues to sustain the employees. 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.
This is not an accident- it is baked into the very point of open source. There's a reason that very few people in the bazaar actually make decent money. There's a reason the cathedral has treasures.
My recommendation is this - decide if you want to make a company, or if you want to make Open Source. The number if places that have succeeded in both is vanishingly small.
> My recommendation is this - decide if you want to make a company, or if you want to make Open Source. The number if places that have succeeded in both is vanishingly small.
I would suggest the contrary: if you want to build a company and believe open source is the right way to do it, please do try! We don't have enough open source companies, we need more successful examples of this.
It's hard and there are traps one needs to not fall into. I personally think VC money is one of the biggest traps, it's absolutely critical you keep control of where and how the companies is going. If you don't have investors to feed, you may make enough money to pay the employees with a good strategy. You can even manage without open core, which incentivizes pushing the useful features outside the open core.
This is not very convincing as a reply to the risks pointed out in the previous post. The choice of the word "try" seemingly recognizes the risk of failure, but you seem to ignore the serious consequences of such a failure for founder and employees. I mean, this is not a game.
There are four business models that I believe are sustainable for open source development.
1. Solo developer (or small group), funded via e.g. Patreon
2. Non-profit funded by sponsorships and donations
3. For-profit but the software is free; the company charges for support and/or cloud services
4. Open collective, where donations fund bounties that are paid out to people who contribute patches
In other words, I think the only models that can really work are models where you genuinely don't care about other companies taking your code and using it for their own purposes. (This shouldn't be surprising - that's literally the whole point of open source!)
When I see companies trying something other than these four business models (usually because they desire more money than being a non-profit can offer) I inevitably see companies that might as well just not be open-source companies, because their licensing has to be restrictive for them to compete.
Very few companies compete on code alone. Those that do I understand the model a bit more, but still usually have the velocity to run away from copycats.
Uber, Yahoo, Google, FB, you name it, all provide a ton of work open source. They don't compete on code alone.
Even if they end up not succeeding, OSS companies are a good thing for another reason: they shift the Overton window.
Some companies are getting boo'd on HN and elsewhere for changing their license from OSI-style to source-available, whereas 10 years ago there would be no source available to the public at all. This is, overall, a positive development.
Well, there's no harm to them if they stop sharing their source X months after switching to source-available model.
It'll even be a net benefit since CoPilot will not devour their code, strip its license out and spit (sorry, emit) to anyone who asks the right questions.
An OSI approved license allows building upon that work, more importantly GPL prevents these improvements to go private.
Source-available provides "for your eyes only" view to the code, and is just window dressing when compared to other models.
I fail to see a net benefit, just a staggered backtrack.
Actually what you're referring to is specifically FOSS - free and open source which is used synonymously with 'open source'. Furthermore there is free as in gratis versus libre.
> So the first mover, the program author, is always at a strategic disadvantage.
I disagree with that part. You have the people who know and understand the code, that's worth a tremendous amount. This also applies for new features, who else is gonna be as proficient as your people at building on top of it? Also a lot of enterprise contracts are all about assured support, who is placed better than you to provide it? You have the people that have the best understanding of the code.
The cathedral vs. the bazaar refers to a way of operating an open source project, not open source vs. closed source [0]. Raymond uses the example of Emacs vs. Linux (Stallman vs. Torvalds) where Stallman works in private, only allows a privileged few to contribute vs. Torvalds who accepts changes liberally.
> decide if you want to make a company, or if you want to make Open Source. The number if places that have succeeded in both is vanishingly small.
Google releases Android source code. While people would like even more of it released, I can and have compiled LineageOS and put it on an Android device.
I could even go more in and get a Pinephone (I do not have one, but have a Pine watch). From another open source company.
Google released Angular and more OSS. Facebook released React.
Red Hat built a giant OSS business before being swallowed up by IBM - b2b companies to get swallowed up by big b2b companies.
On a smaller scale, Health Market Science released a generic database Java library I used. They needed to write some generic software to use, then open sourced it if anyone found it useful. Doing so did not threaten their business, LexisNexis acquired HMS.
Open source is released within and without companies. Lots of successful companies release open source. Most of what I use is open source - from my System76 laptop and desktop, to my Android and LineageOS phones, to my Pinetime watch. I suppose for people who don't program it the appeal level is different, but I do program.
> decide if you want to make a company, or if you want to make Open Source.
these companies decided that they can leverage the marketing that is "open source" in order to gain mind share and traction, after which they start looking to charge.
That is what i associate with any open source company.
It's not always pure marketing. There are companies who truly have open source / free software in their DNA. The founders are true open source proponents, and some people joined these companies because of the open source aspect. Because of this, sometimes, a majority of the people in these companies are pro-open source and will skew things towards this. Some don't even have investors pushing against this.
What do you think about WordPress, XWiki, Nextcloud, Passbolt, Univention, LinPhone, Element, Igalia for instance?
I don't know the internals of each of these, but I believe they all manage to make money from open source by keeping the free software / open source spirit.
I’m not convinced that Open Source is a marketing terms that matters much for many industries (SaaS, hardware, probably others). If anything the continued push back against people trying to stay aligned with the Open Source community and have commercial enterprises tells me that in many ways you’re probably better off not trying in the first place. I personally don’t think that’s the right message.
It is hard to be open licensed but find other players in the space who just want to commercialize other’s contributions without giving back. And I have not seen any practical proposals for how to deal with this without changing license terms away from the OSI definition.
There are lots of successful open source companies. How much did IBM buy Red Hat for? How profitable was it? There are businesses making money out of every major open source project and contributing to it.
The business model fails for reasons other than it just being open source:
1. trying to sell open source as a product rather than as a way to sell something else - services, hardware, whatever.
2. being the main developer - and therefore losing the benefit of sharing development costs. How is it a bazaar if there is only one organization developing it?
> My recommendation is this - decide if you want to make a company, or if you want to make Open Source. The number if places that have succeeded in both is vanishingly small.
The issue at hand is more about deciding how OTHER people/companies may or may not use your source-code to suit their own needs. If you decide Open Source, someone else can easily decide: actually, Company (and free labor to boot).
But to your point, in the long term, Open Source is a better prospect (for the people not the companies). Think of all the problems with crappy IoT devices and bad security practices. Many devices _could_ have decades of life left in them, only to be bricked because the companies want to sell newer crap instead (for the companies not the people). And on-and-on.
I think the whole idea of a new licensing model is a good starting point—like a forced NDA to keep a head-start. However, the same problem holds: nothing will prevent someone from stealing the source code either way—one license or otherwise—if it is freely available. But if there is a large community of contributors, the value prospect for everyone is huge.
And, to top it off: now with AI—how can we prevent an AI from rewriting the sources in a way that appears "nothing like the original"?
A successful company needs to sell something scarce. Open source software is by definition not really a scarce good, so you need to find something adjacent to it that is, and that can be a lot trickier.
I don't thing that is necessarily the case. If your model is to sell the software as a SaaS, then yes it gives you a disadvantage, because someone else can just take it and sell the same thing without paying for development. Same thing if you sell a packaged product. If you sell support, you have a slight advantage, because you have more expertise in your product since you created it, and can make changes directly to it. But then it can be hard to find customers willing to pay, and you have an incentive to make the product difficult to use so more users require support.
I think the ideal situation would be where you are paid for the development itself, rather than selling the resuling software. Maybe with a pool that interested parties can invest in for initial development, new features or continued maintenance, like kickstarter or bug bounties. But that probably wouldn't have the high margins that the software industry has become accustomed to.
Actually this is exactly how many, if not most, software companies are set up today.
The interested parties in this case are users. They pay a small subscription each month. That subscription pays for development, support, hosting, and so on.
Its the very definition of sustainable because it doesn't rely on "new sales" to fund existing customers. The overall cost is spread very thinly across the ones who are benefiting.
I think this applies to Product Companies. If you look at Europe there is very large number of Service Companies developing Open Source and they are doing very well.
The key is - they are Not relying on having some form of monopoly in regards to such software for their success
> This is not an accident- it is baked into the very point of open source. There's a reason that very few people in the bazaar actually make decent money. There's a reason the cathedral has treasures.
Then it's good that "bazaar" vs. "cathedral" is completely orthogonal to "free software/open source" vs. "non-free".
"My recommendation is this - decide if you want to make a company, or if you want to make Open Source"
I don't think this is true in general, but maybe indeed in many cases.
And a non profit organisation would be a better fit(tax reductible donations!), but those are not exactly easy to set up and a project in itself.
valid point but reductionist by narrowing the terms that are considered at all before reaching a broad conclusion. Just as the essay mentions that software distribution is no longer a bottleneck, the "social ecosystem of inventors-builders-makers-users" is a far different proposal in 2024. Some here recall when "using a computer" was a niche activity, now "all of China, yes or no" is ordinary talk? Fish need to see the water - coders need to see the social ecosystems. In this small post, markets are a subset of social ecosystem, but a powerfully charged one.
1. Open source is about the code belonging to the community
2. Open source projects can not survive without funding - as you develop
3. So, both (open source and company) can be compatible if and only if funding belong to the community. In other world, if the open source company is decentralised and democratic. And that is what we need to fight for.
You could have said "company" and "made in EU" and it wouldn't have been more or less true.
Open source itself is not necessarily a problem but one over many factors that have there pros and cons. And for one struggling company doing open source there are hundred or thousands of struggling companies not doing anything opensource or benefiting from opensource without contributing.
The tragedy with any form of software development is that it becomes a commodity very quickly. That's why open source works so well for commodity software. The challenge is that while everybody needs commodity software, they aren't necessarily willing to pay for it. People pay for other things like managed services, support, etc. But not for software directly.
Take operating systems as an example. Unless you have very specific needs, Linux is probably good enough. That's true across most of the industry. Most device manufacturers at this point default to using Linux. Why bother building your own OS when the commodity option does the job. It doesn't make any sense to want to compete with free and OSS. Apple and MS seem to continue to resist this notion but they are increasingly exceptions.
So, open source companies are a bit of a contradiction in terms. You invent something, and then immediately turn it into a cheap commodity by releasing it for free. And then you expect to get money for that. Investors get attracted by rapid growth. And giving something for free can produce some rapid growth.
So, they've repeated fallen into the trap of investing in what proved to be worthless commodities. Some outlasted their IPO at least, which makes for happy investors. But those quickly turn into niche products. Because they are at that point commodity software providers competing with perfectly good OSS ones.
Most new closed source database products out there only have a short while before their differentiating features are absorbed by open source ones. Twenty years ago there were lots of new database products. Most of those were OSS. Several of those then transitioned to closed source. But at this point postgresql does most of what used to make those products interesting. And if it doesn't, just wait five years.
The finance model for OSS is very simple. You need OSS to produce closed source software. It's not optional. You can't compete by doing everything in house. So, most OSS software projects are financed by those companies using them the most. Those projects don't fix themselves. So, there are a lot of companies spending money on the commodity stuff they've built their business on. For example, Oracle keeps on pooring money into Java. Even though it is open source. They make money from it elsewhere. Same with IBM. MS is a big OSS contributor. Google too. Every big software company out there. None of them are driven by idealistic motives. This is economically necessary for them to do.
>The tragedy with any form of software development is that it becomes a commodity very quickly.
I can't think of anything that is less commodity like than software. Commodities are raw materials than can be easily substituted. For example Ukraine is invaded so we all switch to American grain. Try switching SQL Server for mysql or changing your Python code base to Ruby. Software is sticky.
What you are describing is lockin, which is a mechanism companies use to ensure you keep on using their software.
But think about libraries. Imagine a library that handles dates and times correctly. You probably need one for whatever you are doing. Would you pay for one? No of course not because this is a solved problem and just about any language has this built into their standard library. You could build your own but it would have no economical value. Because it is a commodity. There are countless examples like this.
SQL server is a great example. Buying SQL Server for a new project doesn't make any sense for most people. The only reason to use it would be that you are already using it or somehow stuck with it (because you work in some place with lots of legacy software). Otherwise, there are plenty of perfectly fine and free alternatives. Databases are a commodity. If you need one, there are dozen or so free mainstream alternatives. The economic value of new ones is very low unless they do something particularly novel. Which then promptly gets imitated by others. And I'm sure SQL server has a few nice bits and bobs that are amazing. But by and large most people would ignore it as an option at this point. I haven't touched it in twenty years myself. At the time we had a product that could use either mysql or sql server (and Oracle even). Mysql worked fine and there was little technical reason to prefer the other options other than some imaginary value that companies associated with it (like getting support that they would never use).
> You invent something, and then immediately turn it into a cheap commodity by releasing it for free.
Exactly. A 71-line python script https://github.com/getsentry/sentry/commit/3c2e87573d3bd16f6... was groundbreaking when it came out and the fact that it springboarded into a startup is commendable.
But 15 years later and Functional Software Inc. d/b/a Sentry has to support 100+ (according to their marketing material) platforms as well as adding new verticals in UX session replay, web performance, backend profiling, CORS error monitoring, code coverage etc. just to keep the growth machine going. That's not OSS' fault.
software has zero marginal cost, so therefore, if a piece of software can be used by the majority of humans, it makes sense that it ends up costing very little per instance of it.
It's not a tragedy. It's actually a good outcome imho.
> Apple and MS seem to continue to resist this notion but they are increasingly exceptions.
I din't think resisting is the correct word: Linux could be perfect from the user perspective but not the device one: battery handling in a notebook, good support of certain devices, etc. Linux does not replace the whole needs of users or organizations and that is why MS and Apple continue to exist. They existed before Linux either! The business is not the OS.
I happen to be in a special situation here. One of my softwares, Redis, has a very permissive license that was used by many vendors to sell it. At the same time, I'm a lot into 3D printing. So I guess this gives me some perspective on both of the sides of the matter, and what I think, TLDR, is: Bambulab effect on the 3D printing ecosystem is going to be very positive. A few considerations:
1. Prusa is blaming the Bambulab OSS situation, but the reality is that they struggle a lot in innovation, recently. I'm a Prusa Mini and MK4 owner, so I'm kinda of a fan of them, but still... They had struggled to move raise the bar, happy enough to innovate very slowly, with their huge delays with the Prusa XL, and exploiting the fact that other companies were just doing terrible copies. Bambulab before anything else saw that from the POV of technology there was too much left on the table.
2. Prusa itself based its work for PrusaSlicer on an existing open source slicer (written by an Italian guy, btw). While they continue to release PrusaSlicer as a free software, as the license requires to do, what they did renaming the software "Prusa Slicer" is a form of ownership appropriation. PrusaSlicer is a great software that can be used with many non-Prusa printers, my point is just that different ethical perspectives may lead to different conclusions. One could argue that Bambulab created their own slicer instead of just contributing to PrusaSlicer because the naming and the setup of the project makes it hard to do so.
3. Bambulab is also being cloned. The fact that their hardware is not open source does not mean people can't copy the design. The general ideas that make Bambulab printers great can't be patented or copyrighted. See the Creality K1... it resembles Bambulab printers a lot.
4. After Bambulab showed the 3D printing community that it was possible to build better printers, the whole 3D printing landscape became immediately much, much better. They raised the bar. Maybe Prusa will struggle and even fail in the next years (that's my prediction: they will fail, but slowly, since the management is too lacking), but the 3D printing world will be overall much better. In just two years, the average 250$ FDM printer jumped from terrible to totally ok, and this is some kind of value that Bambulab provided, regardless of licensing.
5. One thing that is killing Prusa is in some way they don't use OSS software enough! The future is Klipper, but still they continue with their not-invented-here syndrome to develop their Marlin twist.
6. Finally: using open source does not make you ethically required to open source your stuff. Open source licenses have terms: you just need to follow those terms. If you open source stuff, you are great, but failing to do so does not make you terrible. Remember that in the 3D printing ecosystem there isn't any cloud-companies situation like in the SAAS scenario, where there is a monopoly that allows only a few to exploit OSS value.
I haven't ever used a 3d printer. But your comment made me realize that if PrusaSlicer is based slic3r, it's actually also using software that I wrote many, many years ago.
That's another side of open source: if you don't rely on it to make a living (though it did help in getting my first job as a developer!), there's that pure joy in seeing your software get picked up and used by others. This little discovery made my day.
My perspective is that Bambulab can be a shady company and I have seen behaviors that's concerning such as using bots and uploading designers' prints without permission as well as patents.
Prusa is a standard bearer of the open source 3D printing community and has done a lot of things right and has a lot of goodwill. But they required a competitor to shake things up. Prusa appeared to not be very good at incorporating community contributions, but I haven't really investigated that in-depth, so I may be wrong. In any case, I hope Prusa step up their game in competing and do not fail. They are an example of an open hardware company that shows that Free and open source and commercial activities are not inherently incompatible.
Nonetheless, I bought Bambu Lab's X1C and AMS combo for my business, because I can't really afford to wait on Prusa to deliver their MK4 MMU3 kit. I still purchased two MMU3 kits from them. The irony is that the Bambu Lab printer failed to work out of the box and hadn't worked out a solution thus far, but that's just my individual experience and partly my personal failing.
In the future, I hope my business becomes successful so that I can fund true free software and open source projects so that more of what I used for my business are open source. I don't really want to use a Bambu Lab product any more than I have to. I'll readily jump on open hardware solutions that meet my business requirements.
Bambu Lab may deliver a better product, but at what cost? Always be mindful that these companies are not our friends, especially business that are more transactional. They don't necessary care about the ecosystem at large. If they can get a monopoly, they will.
If you're an open hardware company, be hungry. Don't rest on your laurel. The open source community can only take you so far if you have an inferior product. Not everybody cares about you being open source.
Now, you will hopefully have a lot of community goodwill from being open source. They're your biggest fans and often your biggest customer. If you are open source without being an open source company, you're not really taking advantage of the resources available.
We should fight hard against patents and anything that might threaten democratic 3D printing and manufacturing. Support companies that support the OSS spirit embodied by RepRap. Tell people why this is important.
Let us not forget that the 3D printing business and hobby got started by the expiration of patents and the work of RepRap, the original gangster 3D printer.
It's funny how they lied about being open source twice and got the exact same reaction from it both times, but still haven't learnt their lesson. All they need to do to avoid backlash is call it source available (like it is) but instead they chose to mislead people for no real benefit to themselves.
Well, I also understand not wanting to call it source available, because that is used to describe software where the source code is available, but the license restricts basically any usage of the source.
I think we really need one or more labels for licenses somewhere in between open source and source available.
We should come up with a name for this new generation of licenses. They are not "open source" or "free software", and using those words would be both confusing and immediately polarizing. But having nothing more specific than "source available" seems very insufficient, since most of them turn open source quickly or only deny rights that do not usually matter to the end-user.
(licenses in that family: BUSL, FSL, Common Clause)
It's pointless to call this anything other than source available. This license is just a no-compete license with extra steps. I see little practical difference between "you can't compete with us" and "you can't compete with us for as long as we're developing the software". Maybe we should call this a "you can have the scraps" license because the project only becomes open source when the developers stop caring about it.
> It's pointless to call this anything other than source available.
I obviously strongly disagree with this. An FSL licensed project turns into full, undeniable open source after two years.
> Maybe we should call this a "you can have the scraps" license because the project only becomes open source when the developers stop caring about it.
Two years isn’t a lifetime. If there is value left the community has full rights to do something with it. No legal worries can stand in that way. You can even rebirth a new company from that if you so desire.
It depends who "you" refers to. If you are a company offering a related product, then yes you have to start over and can't use any part of the existing ecosystem. If you are an end-user or a company wanting to use this internally (usually my situation), then the difference is very important.
Maybe it doesn't matter until there are more licenses of this kind though. For now, maybe "BUSL and FSL software" is simple enough, since it's just those two.
The Open Source is about freedom for users not about making life easy for developers or the Company. Open Source is not for everyone and many those crying how horrible Open Source is simply should not gone with Open Source to begin with :)
>The Open Source is about freedom for users not about making life easy for developers or the Company.
That's exactly what open source is about. Open Source was explicitly invented as a more business-friendly version of Free Software, without the whole "freedom" spiel.
Realistically "Free Software" will never actually mean "Free as in Freedom", but instead "Free as in $0".
I'm fully supportive of the Free Software movement but if you have to explain the name actually means a different thing than what it says in plain English then you are going against the grain and being unreasonable. You aren't going to succeed in changing the language but rather, you are just going to confuse the majority of people who hear the term in passing.
Libre can work because it shares roots with liberal, but realistically "Open Source" gets the point across quicker, even if it lacks the nuances of the "Freedom" aspect.
> Why should someone start developing open source if there is no way to make a living from it long term ?
If you cannot answer this question for yourself you definitely should not start or participate in an open source project. Is there someone saying you should? GP isn’t.
If it adds any context to this, Peter is one of the founders of Percona, a company built entirely around open source software and associated services.
That doesn’t mean you should agree with him, but when he says “ those crying how horrible Open Source is simply should not gone with Open Source to begin with” it does have a different meaning if you know where he comes from.
That’s the point. OSS is generally not a good business model. And that’s okay.
Not all software is easily monetized or particularly valuable. The value of standardizing on an interface and outsourcing the maintenance and security to others provides more value than the software itself.
My point of view, as a linux user from my teens to now, early thirties, I'm just some random, so don't take this too seriously.
Are we totally out of touch? Let my explain myself, the idea that if the drivers/firmware were open source it will make the bet on some hardware safer sounds like this to me, "Someone could get it working and share the code in case the company goes out of business in a few years", emphasis on the someone part, here is the thing, the people capable of doing that is pretty small, I ask myself, for the time it would take to deal with that, I could earn more money in that time that I would save 'fixing' the drivers instead of just buying a new one from another brand or the second hand market.
I know people love the romantic idea of free software, open source, etc, I just want my shit to work well, which usually is open source stuff because in most cases it has a zero monetary cost and people use it, so it gets bug fixes from their own users that are trying to make their thing work. I'm equally happy when the software is propietary but works well, case in point the nvidia drivers, it just work well for my use case, the amd drivers are open, well, I can't use their gpus without ending with a headache every few days, doesn't matter if is open source or not, it needs to work well.
Back to my point, even for the power users, how many people will actually fix the drivers/firmware of some hardware they bought? We are in a position in life that is cheaper to just buy new hardware and use that time to learn new things that will pay more in the long term and I personally find we, the linux community are out of touch most of the time with what most people care about.
Also, I find amusing when people get the pikachu face when some big corporation uses OSS and makes profit off of it, isn't that the whole point of the OSS? When I use it on my pc I'm also making profit for myself, the corporations are just doing the same but at scale, and if your goal is to make profit, who wouldn't use all the tools available? They don't have to like OSS, don't have to share the values of the OSS, don't have to 'give back', but OSS allows anyone to use it, so why wouldn't they? Do you want to be a company that is 100% OSS? That's cool, just don't make the pikachu face when another company enters the market and use what have build and shared.
> Also, I find amusing when people get the pikachu face when some big corporation uses OSS and makes profit off of it, isn't that the whole point of the OSS? When I use it on my pc I'm also making profit for myself, the corporations are just doing the same but at scale, and if your goal is to make profit, who wouldn't use all the tools available? They don't have to like OSS, don't have to share the values of the OSS, don't have to 'give back'
That is the point of Open Source, but it is not the point of Free Software. Specifically -- and I know this was addressed in the article -- this is why AGPL should be what programmers are using. It doesn't discriminate, but it does make sure that any entities that use Free Software contribute back to the commons.
This bears repeating again and again; this is why copyleft exists.
> The terms “free software” and “open source” stand for almost the same range of programs. However, they say deeply different things about those programs, based on different values. The free software movement campaigns for freedom for the users of computing; it is a movement for freedom and justice. By contrast, the open source idea values mainly practical advantage and does not campaign for principles. This is why we do not agree with open source, and do not use that term.
> this is why AGPL should be what programmers are using. It doesn't discriminate, but it does make sure that any entities that use Free Software contribute back to the commons.
A competing vendor could offer the software verbatim and still profit due to marketing and network effects (AWS has multiple products that work better together than using disparate vendors for each product), and AGPL would not help in that case.
> Back to my point, even for the power users, how many people will actually fix the drivers/firmware of some hardware they bought? We are in a position in life that is cheaper to just buy new hardware and use that time to learn new things that will pay more in the long term and I personally find we, the linux community are out of touch most of the time with what most people care about.
Assuming you meant Linux kernel specifically, I think you are out of touch how "the linux community" works. Absolute majority of participants are paid by corporations and do what their managers tell them to do. Nobody is just looking to randomly (and for free) improve things average Windows normie would care about.
The FSL is interesting because it correctly assumes the existence of parasites, ie 'harmful free-riders'.
There are ofc too many projects doing what is in effect free R&D for larger entities... these free riders having zero interest in the health or longevity of open technical ecosystems, so game theory is the right lens, I think.
I'd be very interested to learn of any other proposed structures along the lines of the FSL: particularly if anyone here has opinions on/ direct experience with them!
---
Earlier incarnation/s of this approach are mentioned on the FSL page.
Readit News
Open Source. Company.
These are two pretty distinct concepts, and the (traditional) motives for those two things don't merge terribly well.
Over and over we see the same story playing out. Companies need to make revenues to sustain the employees. 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.
This is not an accident- it is baked into the very point of open source. There's a reason that very few people in the bazaar actually make decent money. There's a reason the cathedral has treasures.
My recommendation is this - decide if you want to make a company, or if you want to make Open Source. The number if places that have succeeded in both is vanishingly small.
I would suggest the contrary: if you want to build a company and believe open source is the right way to do it, please do try! We don't have enough open source companies, we need more successful examples of this.
It's hard and there are traps one needs to not fall into. I personally think VC money is one of the biggest traps, it's absolutely critical you keep control of where and how the companies is going. If you don't have investors to feed, you may make enough money to pay the employees with a good strategy. You can even manage without open core, which incentivizes pushing the useful features outside the open core.
This is not very convincing as a reply to the risks pointed out in the previous post. The choice of the word "try" seemingly recognizes the risk of failure, but you seem to ignore the serious consequences of such a failure for founder and employees. I mean, this is not a game.
1. Solo developer (or small group), funded via e.g. Patreon
2. Non-profit funded by sponsorships and donations
3. For-profit but the software is free; the company charges for support and/or cloud services
4. Open collective, where donations fund bounties that are paid out to people who contribute patches
In other words, I think the only models that can really work are models where you genuinely don't care about other companies taking your code and using it for their own purposes. (This shouldn't be surprising - that's literally the whole point of open source!)
When I see companies trying something other than these four business models (usually because they desire more money than being a non-profit can offer) I inevitably see companies that might as well just not be open-source companies, because their licensing has to be restrictive for them to compete.
Uber, Yahoo, Google, FB, you name it, all provide a ton of work open source. They don't compete on code alone.
Some companies are getting boo'd on HN and elsewhere for changing their license from OSI-style to source-available, whereas 10 years ago there would be no source available to the public at all. This is, overall, a positive development.
It'll even be a net benefit since CoPilot will not devour their code, strip its license out and spit (sorry, emit) to anyone who asks the right questions.
An OSI approved license allows building upon that work, more importantly GPL prevents these improvements to go private.
Source-available provides "for your eyes only" view to the code, and is just window dressing when compared to other models.
I fail to see a net benefit, just a staggered backtrack.
https://en.wikipedia.org/wiki/Gratis_versus_libre#%22Free_be...
> and the (traditional) motives for those two things don't merge terribly well.
So depending on whether your source is free as in beer or free as in speech, dictates what business strategy you might have.
I disagree with that part. You have the people who know and understand the code, that's worth a tremendous amount. This also applies for new features, who else is gonna be as proficient as your people at building on top of it? Also a lot of enterprise contracts are all about assured support, who is placed better than you to provide it? You have the people that have the best understanding of the code.
Yes, today a bunch of highly skilled, highly trained people happen to work for you.
Tomorrow those same people get hired by your customers, strike out on their own, or join the competition.
All the points you raise are true. But the company doesn't "own" the people, it can only exploit them for as long as they hang around.
[0] https://en.wikipedia.org/wiki/The_Cathedral_and_the_Bazaar
Google releases Android source code. While people would like even more of it released, I can and have compiled LineageOS and put it on an Android device.
I could even go more in and get a Pinephone (I do not have one, but have a Pine watch). From another open source company.
Google released Angular and more OSS. Facebook released React.
Red Hat built a giant OSS business before being swallowed up by IBM - b2b companies to get swallowed up by big b2b companies.
On a smaller scale, Health Market Science released a generic database Java library I used. They needed to write some generic software to use, then open sourced it if anyone found it useful. Doing so did not threaten their business, LexisNexis acquired HMS.
Open source is released within and without companies. Lots of successful companies release open source. Most of what I use is open source - from my System76 laptop and desktop, to my Android and LineageOS phones, to my Pinetime watch. I suppose for people who don't program it the appeal level is different, but I do program.
these companies decided that they can leverage the marketing that is "open source" in order to gain mind share and traction, after which they start looking to charge.
That is what i associate with any open source company.
What do you think about WordPress, XWiki, Nextcloud, Passbolt, Univention, LinPhone, Element, Igalia for instance?
I don't know the internals of each of these, but I believe they all manage to make money from open source by keeping the free software / open source spirit.
It is hard to be open licensed but find other players in the space who just want to commercialize other’s contributions without giving back. And I have not seen any practical proposals for how to deal with this without changing license terms away from the OSI definition.
The business model fails for reasons other than it just being open source:
1. trying to sell open source as a product rather than as a way to sell something else - services, hardware, whatever. 2. being the main developer - and therefore losing the benefit of sharing development costs. How is it a bazaar if there is only one organization developing it?
Successful tech companies don't sell themselves, particularly to has-beens like IBM.
The issue at hand is more about deciding how OTHER people/companies may or may not use your source-code to suit their own needs. If you decide Open Source, someone else can easily decide: actually, Company (and free labor to boot).
But to your point, in the long term, Open Source is a better prospect (for the people not the companies). Think of all the problems with crappy IoT devices and bad security practices. Many devices _could_ have decades of life left in them, only to be bricked because the companies want to sell newer crap instead (for the companies not the people). And on-and-on.
I think the whole idea of a new licensing model is a good starting point—like a forced NDA to keep a head-start. However, the same problem holds: nothing will prevent someone from stealing the source code either way—one license or otherwise—if it is freely available. But if there is a large community of contributors, the value prospect for everyone is huge.
And, to top it off: now with AI—how can we prevent an AI from rewriting the sources in a way that appears "nothing like the original"?
https://journal.dedasys.com/2007/02/03/in-thrall-to-scarcity...
I think the ideal situation would be where you are paid for the development itself, rather than selling the resuling software. Maybe with a pool that interested parties can invest in for initial development, new features or continued maintenance, like kickstarter or bug bounties. But that probably wouldn't have the high margins that the software industry has become accustomed to.
The interested parties in this case are users. They pay a small subscription each month. That subscription pays for development, support, hosting, and so on.
Its the very definition of sustainable because it doesn't rely on "new sales" to fund existing customers. The overall cost is spread very thinly across the ones who are benefiting.
The key is - they are Not relying on having some form of monopoly in regards to such software for their success
Then it's good that "bazaar" vs. "cathedral" is completely orthogonal to "free software/open source" vs. "non-free".
I don't think this is true in general, but maybe indeed in many cases. And a non profit organisation would be a better fit(tax reductible donations!), but those are not exactly easy to set up and a project in itself.
Companies also hold their cards close tending to not share their insights, information, and plans.
1. Open source is about the code belonging to the community
2. Open source projects can not survive without funding - as you develop
3. So, both (open source and company) can be compatible if and only if funding belong to the community. In other world, if the open source company is decentralised and democratic. And that is what we need to fight for.
You could have said "company" and "made in EU" and it wouldn't have been more or less true.
Open source itself is not necessarily a problem but one over many factors that have there pros and cons. And for one struggling company doing open source there are hundred or thousands of struggling companies not doing anything opensource or benefiting from opensource without contributing.
Take operating systems as an example. Unless you have very specific needs, Linux is probably good enough. That's true across most of the industry. Most device manufacturers at this point default to using Linux. Why bother building your own OS when the commodity option does the job. It doesn't make any sense to want to compete with free and OSS. Apple and MS seem to continue to resist this notion but they are increasingly exceptions.
So, open source companies are a bit of a contradiction in terms. You invent something, and then immediately turn it into a cheap commodity by releasing it for free. And then you expect to get money for that. Investors get attracted by rapid growth. And giving something for free can produce some rapid growth.
So, they've repeated fallen into the trap of investing in what proved to be worthless commodities. Some outlasted their IPO at least, which makes for happy investors. But those quickly turn into niche products. Because they are at that point commodity software providers competing with perfectly good OSS ones.
Most new closed source database products out there only have a short while before their differentiating features are absorbed by open source ones. Twenty years ago there were lots of new database products. Most of those were OSS. Several of those then transitioned to closed source. But at this point postgresql does most of what used to make those products interesting. And if it doesn't, just wait five years.
The finance model for OSS is very simple. You need OSS to produce closed source software. It's not optional. You can't compete by doing everything in house. So, most OSS software projects are financed by those companies using them the most. Those projects don't fix themselves. So, there are a lot of companies spending money on the commodity stuff they've built their business on. For example, Oracle keeps on pooring money into Java. Even though it is open source. They make money from it elsewhere. Same with IBM. MS is a big OSS contributor. Google too. Every big software company out there. None of them are driven by idealistic motives. This is economically necessary for them to do.
I can't think of anything that is less commodity like than software. Commodities are raw materials than can be easily substituted. For example Ukraine is invaded so we all switch to American grain. Try switching SQL Server for mysql or changing your Python code base to Ruby. Software is sticky.
But think about libraries. Imagine a library that handles dates and times correctly. You probably need one for whatever you are doing. Would you pay for one? No of course not because this is a solved problem and just about any language has this built into their standard library. You could build your own but it would have no economical value. Because it is a commodity. There are countless examples like this.
SQL server is a great example. Buying SQL Server for a new project doesn't make any sense for most people. The only reason to use it would be that you are already using it or somehow stuck with it (because you work in some place with lots of legacy software). Otherwise, there are plenty of perfectly fine and free alternatives. Databases are a commodity. If you need one, there are dozen or so free mainstream alternatives. The economic value of new ones is very low unless they do something particularly novel. Which then promptly gets imitated by others. And I'm sure SQL server has a few nice bits and bobs that are amazing. But by and large most people would ignore it as an option at this point. I haven't touched it in twenty years myself. At the time we had a product that could use either mysql or sql server (and Oracle even). Mysql worked fine and there was little technical reason to prefer the other options other than some imaginary value that companies associated with it (like getting support that they would never use).
Exactly. A 71-line python script https://github.com/getsentry/sentry/commit/3c2e87573d3bd16f6... was groundbreaking when it came out and the fact that it springboarded into a startup is commendable. But 15 years later and Functional Software Inc. d/b/a Sentry has to support 100+ (according to their marketing material) platforms as well as adding new verticals in UX session replay, web performance, backend profiling, CORS error monitoring, code coverage etc. just to keep the growth machine going. That's not OSS' fault.
It's not a tragedy. It's actually a good outcome imho.
I din't think resisting is the correct word: Linux could be perfect from the user perspective but not the device one: battery handling in a notebook, good support of certain devices, etc. Linux does not replace the whole needs of users or organizations and that is why MS and Apple continue to exist. They existed before Linux either! The business is not the OS.
1. Prusa is blaming the Bambulab OSS situation, but the reality is that they struggle a lot in innovation, recently. I'm a Prusa Mini and MK4 owner, so I'm kinda of a fan of them, but still... They had struggled to move raise the bar, happy enough to innovate very slowly, with their huge delays with the Prusa XL, and exploiting the fact that other companies were just doing terrible copies. Bambulab before anything else saw that from the POV of technology there was too much left on the table.
2. Prusa itself based its work for PrusaSlicer on an existing open source slicer (written by an Italian guy, btw). While they continue to release PrusaSlicer as a free software, as the license requires to do, what they did renaming the software "Prusa Slicer" is a form of ownership appropriation. PrusaSlicer is a great software that can be used with many non-Prusa printers, my point is just that different ethical perspectives may lead to different conclusions. One could argue that Bambulab created their own slicer instead of just contributing to PrusaSlicer because the naming and the setup of the project makes it hard to do so.
3. Bambulab is also being cloned. The fact that their hardware is not open source does not mean people can't copy the design. The general ideas that make Bambulab printers great can't be patented or copyrighted. See the Creality K1... it resembles Bambulab printers a lot.
4. After Bambulab showed the 3D printing community that it was possible to build better printers, the whole 3D printing landscape became immediately much, much better. They raised the bar. Maybe Prusa will struggle and even fail in the next years (that's my prediction: they will fail, but slowly, since the management is too lacking), but the 3D printing world will be overall much better. In just two years, the average 250$ FDM printer jumped from terrible to totally ok, and this is some kind of value that Bambulab provided, regardless of licensing.
5. One thing that is killing Prusa is in some way they don't use OSS software enough! The future is Klipper, but still they continue with their not-invented-here syndrome to develop their Marlin twist.
6. Finally: using open source does not make you ethically required to open source your stuff. Open source licenses have terms: you just need to follow those terms. If you open source stuff, you are great, but failing to do so does not make you terrible. Remember that in the 3D printing ecosystem there isn't any cloud-companies situation like in the SAAS scenario, where there is a monopoly that allows only a few to exploit OSS value.
That's another side of open source: if you don't rely on it to make a living (though it did help in getting my first job as a developer!), there's that pure joy in seeing your software get picked up and used by others. This little discovery made my day.
Prusa is a standard bearer of the open source 3D printing community and has done a lot of things right and has a lot of goodwill. But they required a competitor to shake things up. Prusa appeared to not be very good at incorporating community contributions, but I haven't really investigated that in-depth, so I may be wrong. In any case, I hope Prusa step up their game in competing and do not fail. They are an example of an open hardware company that shows that Free and open source and commercial activities are not inherently incompatible.
Nonetheless, I bought Bambu Lab's X1C and AMS combo for my business, because I can't really afford to wait on Prusa to deliver their MK4 MMU3 kit. I still purchased two MMU3 kits from them. The irony is that the Bambu Lab printer failed to work out of the box and hadn't worked out a solution thus far, but that's just my individual experience and partly my personal failing.
In the future, I hope my business becomes successful so that I can fund true free software and open source projects so that more of what I used for my business are open source. I don't really want to use a Bambu Lab product any more than I have to. I'll readily jump on open hardware solutions that meet my business requirements.
Bambu Lab may deliver a better product, but at what cost? Always be mindful that these companies are not our friends, especially business that are more transactional. They don't necessary care about the ecosystem at large. If they can get a monopoly, they will.
If you're an open hardware company, be hungry. Don't rest on your laurel. The open source community can only take you so far if you have an inferior product. Not everybody cares about you being open source.
Now, you will hopefully have a lot of community goodwill from being open source. They're your biggest fans and often your biggest customer. If you are open source without being an open source company, you're not really taking advantage of the resources available.
We should fight hard against patents and anything that might threaten democratic 3D printing and manufacturing. Support companies that support the OSS spirit embodied by RepRap. Tell people why this is important.
Let us not forget that the 3D printing business and hobby got started by the expiration of patents and the work of RepRap, the original gangster 3D printer.
I think we really need one or more labels for licenses somewhere in between open source and source available.
(licenses in that family: BUSL, FSL, Common Clause)
Anyone has suggestions?
I obviously strongly disagree with this. An FSL licensed project turns into full, undeniable open source after two years.
> Maybe we should call this a "you can have the scraps" license because the project only becomes open source when the developers stop caring about it.
Two years isn’t a lifetime. If there is value left the community has full rights to do something with it. No legal worries can stand in that way. You can even rebirth a new company from that if you so desire.
Maybe it doesn't matter until there are more licenses of this kind though. For now, maybe "BUSL and FSL software" is simple enough, since it's just those two.
Commons Clause is not in that group, as it's not time-limited.
Deleted Comment
That's exactly what open source is about. Open Source was explicitly invented as a more business-friendly version of Free Software, without the whole "freedom" spiel.
I'm fully supportive of the Free Software movement but if you have to explain the name actually means a different thing than what it says in plain English then you are going against the grain and being unreasonable. You aren't going to succeed in changing the language but rather, you are just going to confuse the majority of people who hear the term in passing.
Libre can work because it shares roots with liberal, but realistically "Open Source" gets the point across quicker, even if it lacks the nuances of the "Freedom" aspect.
Should we just expect OSS devs to work for free (or scraps via sponsors/kofi)?
I guess there are a lot selfless individuals who would still use a significant portion of their lifetime to build OSS.
But what about software that needs significant effort to build and constant updates?
I find your view very cynical and I exactly the attitude why I wouldn’t start an ambitious OSS project for something more complex.
If you cannot answer this question for yourself you definitely should not start or participate in an open source project. Is there someone saying you should? GP isn’t.
That doesn’t mean you should agree with him, but when he says “ those crying how horrible Open Source is simply should not gone with Open Source to begin with” it does have a different meaning if you know where he comes from.
Not all software is easily monetized or particularly valuable. The value of standardizing on an interface and outsourcing the maintenance and security to others provides more value than the software itself.
Are we totally out of touch? Let my explain myself, the idea that if the drivers/firmware were open source it will make the bet on some hardware safer sounds like this to me, "Someone could get it working and share the code in case the company goes out of business in a few years", emphasis on the someone part, here is the thing, the people capable of doing that is pretty small, I ask myself, for the time it would take to deal with that, I could earn more money in that time that I would save 'fixing' the drivers instead of just buying a new one from another brand or the second hand market.
I know people love the romantic idea of free software, open source, etc, I just want my shit to work well, which usually is open source stuff because in most cases it has a zero monetary cost and people use it, so it gets bug fixes from their own users that are trying to make their thing work. I'm equally happy when the software is propietary but works well, case in point the nvidia drivers, it just work well for my use case, the amd drivers are open, well, I can't use their gpus without ending with a headache every few days, doesn't matter if is open source or not, it needs to work well.
Back to my point, even for the power users, how many people will actually fix the drivers/firmware of some hardware they bought? We are in a position in life that is cheaper to just buy new hardware and use that time to learn new things that will pay more in the long term and I personally find we, the linux community are out of touch most of the time with what most people care about.
Also, I find amusing when people get the pikachu face when some big corporation uses OSS and makes profit off of it, isn't that the whole point of the OSS? When I use it on my pc I'm also making profit for myself, the corporations are just doing the same but at scale, and if your goal is to make profit, who wouldn't use all the tools available? They don't have to like OSS, don't have to share the values of the OSS, don't have to 'give back', but OSS allows anyone to use it, so why wouldn't they? Do you want to be a company that is 100% OSS? That's cool, just don't make the pikachu face when another company enters the market and use what have build and shared.
That is the point of Open Source, but it is not the point of Free Software. Specifically -- and I know this was addressed in the article -- this is why AGPL should be what programmers are using. It doesn't discriminate, but it does make sure that any entities that use Free Software contribute back to the commons.
This bears repeating again and again; this is why copyleft exists.
> The terms “free software” and “open source” stand for almost the same range of programs. However, they say deeply different things about those programs, based on different values. The free software movement campaigns for freedom for the users of computing; it is a movement for freedom and justice. By contrast, the open source idea values mainly practical advantage and does not campaign for principles. This is why we do not agree with open source, and do not use that term.
Source: https://www.gnu.org/philosophy/open-source-misses-the-point....
A competing vendor could offer the software verbatim and still profit due to marketing and network effects (AWS has multiple products that work better together than using disparate vendors for each product), and AGPL would not help in that case.
Assuming you meant Linux kernel specifically, I think you are out of touch how "the linux community" works. Absolute majority of participants are paid by corporations and do what their managers tell them to do. Nobody is just looking to randomly (and for free) improve things average Windows normie would care about.
The FSL is interesting because it correctly assumes the existence of parasites, ie 'harmful free-riders'.
There are ofc too many projects doing what is in effect free R&D for larger entities... these free riders having zero interest in the health or longevity of open technical ecosystems, so game theory is the right lens, I think.
I'd be very interested to learn of any other proposed structures along the lines of the FSL: particularly if anyone here has opinions on/ direct experience with them!
---
Earlier incarnation/s of this approach are mentioned on the FSL page.
Open Core https://en.wikipedia.org/wiki/Open-core_model BUSL https://mariadb.com/bsl11/ FSL https://fsl.software/