Generating Income from Open Source(vadimdemedes.com) |
Generating Income from Open Source(vadimdemedes.com) |
There is nothing wrong with developing software and selling it. It used to be the norm and it is still quite common.
These days "open source" is virtue signaling and expected "to be cool". That people then grump about not being paid shows that they are not particularly interested in the concept of open source.
However, the term open source has been used for so many different things that is has lost most shared coherent meaning. I guess that is why "Libre Software" is now being used by some.
What is plain hypocrisy to me are the people who developed "for profit" open source projects, and are angry when they dont get paid enough, when the projects they have created rely on GNU tools, the Linux kernel (Perhaps Debian), a libre software database on so on.
Their product is built on the shoulders and sacrifice of other open-source folks that often is not getting paid or not paid enough.
For profit open source should have clause of donating X% of profits downstream to the open-source projects they rely on. (Since they would have no product without them)
There are some core and vital open-source projects that have enormous importance for millions. I think supporting them should be a priority.
A lot of modern software is "open source" until it become inconvenient and they introduce "Pro", or "see source", or shift license entirely.
A lot of companies should just be "for profit". There is nothing wrong with it and nothing wrong with making money, or making money off of open source. I do nitpick on people who see open source as a way to get paid and they are angry if they dont get it.
Now, maybe it's not as profitable, but not every wants to be as big as Apple or Microsoft.
Also helping sourcehut is that it's actual deployment process is uh... Interesting? It's expecting to run on Alpine (and not dockerized Alpine) and you need to be able to supply an email server. The latter especially means that the usual problem of "FOSS hosting with a price label" doesn't really pop up; nobody is going to bother cloning sourcehuts deployment setup without the fees attached because very few people want to selfhost email for a personal instance of FOSS software. (Since self-hosting email is rather annoying for maintenance since your server IP can end up in random blacklists and you won't know until the setup stops working, and sourcehut relies a lot on email.)
[0]: https://man.sr.ht/billing-faq.md - seems to mostly be on a "were not gonna be jerks about it" - "if you're so poor you can't pay for sourcehut on the price given, send me an email and we'll work something out" basis - the goal doesn't seem to be to extract profit from every user and "fuck poor people", which is quite respectable if you ask me.
This is a big part of why I find it so crazy that people argue that "Open Source" work isn't open when a commercial license is attached (besides the fact that "available" and "open" are synonyms and it is a clearly reasonable interpretation of the words). It strikes a balance between that "coolness" you mention while not letting big companies take your work and make massive profits while stepping on your shoulders.
I think it is completely reasonable for Open Source (aka Software Available) projects to say "Here's my work. If there's pie to be had, just be cool and share a little. If there isn't, don't fret, we're in this together." It is astounding to me that people attack those asking for a little pie when they so clearly are lifting up the pie eaters. It would be one thing if they were asking for the whole pie, but (like this article) that doesn't seem to be the vast majority of them. And if we're being honest, attacking the scaffolding seems just like it ends up in less pies overall.
Most of the time when I see people grumping about getting paid for open source, it isn't the actual developer. It's someone who thinks the developer should get paid more. Usually it's a blog post to a, "Did you know critical package x is maintained by some 70 year old dude in Dubuque, Iowa... and he's on VACATION RIGHT NOW?"
That said, the article is a pretty good summary.
- They want to be "open source" because they or their VC think it's important for the marketing benefits and it seems to be of interest to customers
- But, they're not really planning to invest to make the open source development model work
- And, they sure don't want to disadvantage themselves from a business perspective (have trouble holding features back/give potential competitors a leg up)
Sure, but the FSF (and others) may have an opinion on the license given to the paying customers, in order to not restrict them.
Otherwise, spot on; Open source is not a business model: <https://news.ycombinator.com/item?id=32287432>
Say what you want about copyright, just revealing the source impedes your ability to commercialize it.
That means the client code can't be your secret sauce. The client code is effectively this table-stakes item that you have to release for free. The hosted server component is where you have to put your differentiators.
Obviously doesn't work with all types of software, but even something like a C++ backend library (like OpenCV) you could open source the C++, but sell access to the python bindings.
I'm going out on a limb here, but what have you done to support them? Apologies if you have, but a lot of people seem to want someone else to support their favorite free thing.
I’m not taking any sides here, but GitHub’s sponsor feature makes this _very_ easy to do, and that’s valuable to me.
Edit to add: Agility in contributing money to FOSS projects is not where it could be, and that’s absolutely a place we could do better in to drive more direct funds into the FOSS ecosystem.
Also, it's not helpful to go out of your way this much to disqualify someone from having an opinion. It's essentially an accusation of hypocrisy, made absolutely blindly and randomly at a person you don't know anything about who expressed an opinion you didn't like.
> There is nothing wrong with developing software and selling it. It used to be the norm and it is still quite common.
Yes, agreed.
> These days "open source" is virtue signaling and expected "to be cool".
Yes, agreed.
> That people then grump about not being paid shows that they are not particularly interested in the concept of open source.
Well, yes, they got pulled in by the hype and then realized that they needed some kinds of cashflows to make it sustainable.
> I do nitpick on people who see open source as a way to get paid and they are angry if they dont get it.
Why? Because they made one mistake up front and picked the wrong model and didn't realize until they were several years into their project that open source was wrong for them? So you decide to crucify them for that, even though if they'd made that decision right from the start you would have 100% fully supported them?
I didn't realize that open sourcing something was a perpetual indefinite infinite-year long contract that requires endless toil and slavery to the cause. Perhaps we should be a bit more clear about that up front.
I do think moving the goalposts or being dishonest about your intentions are wrong.
What I’d like to see when this happens is honesty about relicensing with a view toward profit (or whatever you’re after), and not attempting to scrub your more liberally licensed code from the face of the planet.
I’m not asserting that everyone does this, but it’s certainly been done many times.
If you think you might one day make money on it, just say that up front and don’t be greedy when you finally do.
Why do you exclusively point out "for profit" open source projects here? What about basically any commercial software that leeches on open source software without donating anything?
We're all standing on the shoulders of giants. I don't see the hypocrisy here.
I propose something like this: When I release code, it's rights-reserved for 5 years, then open-source (and this particular clause would be irrevocable). Anyone may use the software for non-commercial purposes. Anyone may contribute, those who contribute will be granted permission for commercial use if I deem their contributions significant enough. Anyone may distribute the software under these terms.
If such a model became popular, I have a hard time imagining it could make things any worse. It might even accelerate open-source development by shifting societies resources to those who are actually producing this valuable software.
You might say, "but it's not open-source", fair enough, but we can view it as open-source contribution with a delay. For example, if this model became widely popular this year, and we saw great progress with this model, then come 2028 we would be flooded with new open-source software and ultimately might be better off than we would have been without this model.
Ultimately people choose to make their software open-source for moral reasons, and because they hope that by giving they might receive in turn--I contribute, you contribute, and we share freely (and we're both poor). This model I've talked about still achieves similar goals. People might still be willing to contribute, even for free, because I have given them an irrevocable legal promise that their contributions will be made available to all at a specific time.
(This whole thing makes me rethink copyright and patents and how much they really contribute to society. Perhaps their terms should be shortened?)
[1]: https://github.com/dragonflydb/dragonfly/blob/main/LICENSE.m...
I’m not making any comment about the merits of this model. I strictly make and sell closed-source software (SaaS). But it’s funny to me that this is essentially the same business model, but you use the threat of legal action rather than API keys to enforce your rights. “Open Source” is in the name only.
I definitely agree that 1) we need to convert companies more, but 2) it won’t happen with the traditional sponsorship/donations model. They need to be able to quantify the value of their investments.
With the first version of Polar, we aim to provide that while giving maintainers better insight on needs & funding to develop efforts that align with their vision.
Our tax law (in Canada) is so opaque, and I feel like finding ways for these monetary contributions to be deductible as business expenses would make it so much easier for businesses to justify the expense.
If donations can't be written off, OS projects would then do well to add a paid tier with some useless exclusive functionality, just so donors can make it a software expense.
Completely open source software, but you can pay a reasonable amount for managed hosting and support. Plus obviously it supports the devs and new development.
A simple line to separate DIYers from anyone who wants the convenience, peace of mind, or to support the project itself. Would love to see more of this.
In other words, I don't think it's a sustainable way to support open source projects
Then there's companies that write open-source software and are not only successful in charging for it, but also have funding. I run Fossfox [0] that indexes companies that have either open-sourced their main products (eg ClickHouse) or that heavily contribute to open-source (eg OpenAI). The ones that open-source their main products: some of them do use copyleft licenses, but most surprisingly go with Apache 2.0 (and of course use some dual licenses; with "ee/" directories). The one thing a lot of them have in common is providing hosted versions of their software. A lot of customers don't want the hassle of running it themselves, and do pay for hosting and support.
https://github.com/fossjobs/fossjobs/wiki/resources https://www.fossjobs.net/
But it hasn't been updated in years: https://github.com/metafizzy/isotope
Does the commercial license use a different repo?
> Purchasing takes a minute. Our purchasing form accepts credit cards or PayPal. Once purchased, you’ll receive a commercial license PDF and you will be all set to use Isotope in your commercial applications.
https://isotope.metafizzy.co/license.html#purchasing-a-comme...
https://github.com/whyboris/Video-Hub-App
I also sell it for $5 and have sold just over 5,000 copies last month (5 years old app). Importantly, $3.50 of every purchase goes to a cost-effective charity, GiveWell recommended Against Malaria Foundation (see website for details).
I call this charityware and I wrote about it years ago: https://medium.com/@whyboris/charityware-doing-good-with-pro...
For instance: Here is one such company (https://opsverse.io) that specialises on both #1 and #2.
Community and OSS are 2 things that are at the center of everything that OpsVerse does. OpsVerse provides "few clicks and you are up and running" sort of suite of devops tools entirely powered by OSS tools, and without the trouble of managing/scaling the tools.
I see these donation numbers where they’re netting $200/mo for some wildly successful tool. Closed sourcing it + marketing and you’ll be at $2-$3k/mo in 6 months with moderate effort.
Open source is cool for a communal app with a lot of different vectors. Openpilot for example is smart for me. For everything else though? It’s an easy way for software to become unmaintained. Good luck finding someone to work for free in perpetuity.
https://codecodeship.com/ (JS, Ruby, .NET, Rust, Elixir)
https://anystack.sh/ (PHP, Ruby, JS, Python)
Full disclosure: I built Code Code Ship
People don't understand that in these constellations the project is commercially released first and then the open source license allows things that have not been covered by the commercial license. This is the opposite of restriction.
In fact all of the big name software companies have people on staff that represent them in various mission critical open source projects. There are also a lot of smaller companies that are very successful that are built on top of open source projects that they either pioneered or contribute to heavily. Commercial contributions to open source are the back bone to the open source communities.
Any such business is a great place to start if you are looking to make money from open source. They might employ you if you are good enough and having a track record as a top contributor is a great way into such companies. Getting involved with such projects is a great way to bootstrap your professional career in open source development.
Where a lot of amateur run projects fail to make money is mis-perception of their value, unprofessional behavior by developers, or just plain poor business and communication skills. That happens a lot. Computer scientists are not great at this stuff. Also they tend to be a bit spoiled because they are constantly being pampered by companies that need their technical skills. So, not necessarily the best entrepreneurs.
Built it and they will come is not how things work. It's not a business model. Making money is hard work and you have to be smart about it. Just because you think that your source code is worth a lot doesn't mean it actually is. E.g. a lot of javascript packages are a combination of tiny and a bit of a commodity. What's a commodity: something with a lot of alternate implementations. If you are cost conscious consumer of such things, you are going to be optimizing for cost and convenience and pick the cheapest (i.e. free) good enough thing. Commodities are by definition not worth a lot. And in OSS that typically means 0$.
Mistake #1 with OSS companies: most of them provide commodity functionality that just isn't worth a lot to many people. Set your expectations accordingly. There can still be valid reasons to work on them.
I maintain some small open source projects and I've set my expectations accordingly. I don't expect or ask for donations. Below a certain threshold the admin work on my side would cost me more than the amount I'd receive. It makes no economical sense to me to bother with low $ amount tips unless I magically can get a lot of them. For the same reason, I don't donate to OSS projects either. I reserve donations for more worthy causes.
I make money indirectly via consultancy. The libraries are just there to make my life easier and they are a nice talking point that helps my clients understand that I might know what I'm talking about. Most of my clients don't actually use my OSS even. Mainly they just make my life easier. And I enjoy working on them. Nothing wrong with that.
You don't need to propose it. If its your code you can just do it.
I would recommend that you clearly document (for yourself, if not publically) what your goals are. Is it money? Fame? Github stars? Users? Contributers? Something else? In order to determine if the license is a success you need to set goals, and then measure the results.
Frankly, while this topic comes up frequently on HN, it's getting somewhat boring. I'm starting to come around to the viewpoint that you can either give your work away for free, or make a living, [1], but fundamentally those two goals are against each other.
Sure you can make a living tangential to the work (docs, support et al), but of course that's taking time away from the code (so kinda just like a day job.)
[1] as with lots of fields, like authors and music, there is a tiny sliver of people who manage it, and are held up as examples that it can be done.
But I do applaud your creativity here, and i encourage you to experiment. My first thought is why bother to go Open at all? Presumably to meet some goal and it will be interesting to see if it meets that goal.
I suspect you are looking for contributers (but I'm putting words in your mouth here) and I would suggest that the vast majority of projects get no contributers, regardless of license.
=> https://mariadb.com/bsl-faq-adopting/ MariaDB - Adopting and Developing BSL Software
I've also noticed Sentry using it. I think it address one of the the main criticisms of proprietary licenses, while still allowing a company to form a considerable moat around their software. (I think we can agree, death of author + 70 years is way too long for code source, but 5 years of copyright seems perfectly reasonable.)
Too long copyright durations and too stringent distribution restrictions are undone by open source projects. This does not result in an equilibrium because there is a vast space of projects that need to be proprietary in some dimension and open source in another. The equilibrium solution has to be a hybrid of the two approaches.
Side none: An equilibrium is like a limit of a converging series. You don't have to reach it, just get close enough until you are happy aka |x_i - x|< epsilon
https://www.fsf.org/blogs/rms/selling-exceptions
He's got mixed feelings, but says "I consider selling exceptions an acceptable thing for a company to do, and I will suggest it where appropriate as a way to get programs freed."
If of all people rms considers something acceptable, then whatever that thing is, it certainly can't be described as "virtually closed source".
If you release something as open source software, then it is open source software. The end. That's what the term means. Also licensing it under a proprietary license doesn't change that. Free software licensing is not "try before you buy", and it is not "free usage for non-profits". Open source and "not for profit" are not synonyms de facto or de jure. You can use GPL'd software in commercial products and in fact many people do, frequently. You probably have a dozen copies of Linux in your house in various appliances that you don't even know are there. There are many other examples.
If you publish a piece of software under two licenses, one open source and one commercial, why would anybody pick the commercial one?
You cannot say "corporations have to pick the commercial one". The only way to do that would be to include a term in the open source license so it cannot be used by corporations. But then it would no longer meet the open source definition. "Free for private use but payed for commercial use" is not open source!
But if the open source license is copyleft (e.g. GPL), any software that uses it would also have to be open source. So dual licensing allows corporations to pay to keep their own code a secret.
This implies that dual licensing is pointless for non-copyleft licenses like MIT, and also for developer tooling that is not directly used in the final product.
That is not what is proposed in the article:
----
Isotope is open source, but there are different licenses depending on how you intend to use it:
Open source license.
Commercial license.
----
That is not open source. It's proprietary software offered as freeware for non-commercial use.
__Stop gatekeeping__
Source available software with commercial licensing still helps the main intended target: hackers, builders, inventors, creators, learners, researchers, pen testers, and so on. Opened source code makes programs safer and encourages creativity. The article is right that you can't just live on donations and pure good will, as much as we'd love to see this you gotta pay the bills. Accountability is low in large societies and social pressure means little. SAS still helps the people, but it also puts restrictions on those that would profit off of your work and leave you high and dry. Comments like these only encourage closed source hardware as well as the lack of free software.
The fucking name OSS confuses people and this has been an utter disaster. Your name should clearly indicate what is being advocated and should not conflict with a reasonable but distinctly different definition (there's a lot of this for some fucking reason). Especially when the reasonable interpretation makes more sense! Even you say that it is in the name only and we see these fights almost weekly here. Words mean what people use them as. I don't care what OSI says, OSS and SAS are the same thing. Corporations lose trademarks as words/phrases enter the lexicon, I don't know why we should defend OSI's inane definition. They have not won the culture war on what the term means.
It doesn't matter one whit whether people can "pay the bills". It's totally irrelevant to the discussion. If you can't make money making open source software, then go get a job. This is like people complaining they did a degree in underwater basket weaving and can't get a job, or people complaining that they can't get a job as an artist or a guitarist. Guess what, most people can't make money from their hobbies. Some programmers get pretty lucky and turn their hobbies into jobs. If you want to give it a go, good on you. Most people don't even want to. Of those that do want to, most do not succeed. You have no entitlement to be successful in your attempts to do so.
And there's no rule that says that you have to release things you make as open source software. If you want to make money, you might find that the best way to do that is not to release your software as open source. That's your prerogative. But you can't make that choice and then also choose to ride on the reputation of the "open source" label. That label means something, and has meant the same thing for decades. You can choose: release your software as open source and call it that, or release it as "source available" and be honest about it.
And you're completely wrong when you say "They have not won the culture war on what the term means". They absolutely have. The only place I've ever seen anyone pretend that "open source" can legitimately mean something other than the OSI's definition is on this forum.
I completely agree. I’ve avoided using the term “open source” where possible for my SAS, instead opting for word-play like “open, source-available”, but it was frustrating that I felt I even needed to do that to avoid backlash from the zealots. And I’ve loosened my reigns there a bit as well as time has gone by. I think it’s silly. I want to use “open source” because that’s what it is.
As an example, there’s no reason Elastic should have had to cave to OSI and start using “free & open” everywhere in place of “open source.” The reality is SAS licenses like ELv2 may not be “Open Source (tm)” but they sure as hell are “open source” to most people in our industry.
FOSS !== OSS.
It is open source, yes. But "stop gatekeeping" is a bad phrase. "Gatekeeping" is now a negative spin on "telling the difference between things", and thus worse than useless.
Every law system in the world distinguishes natural personhood from legal personhood. If you've heard somewhere that Citizens Unite ruling said that corporations and people are the same, you've been lied to and disinformed.
You really do not understand what rights Open Source licenses give you. The GPL in particular really only affects you if you distribute it (a js library vs nginx). There is no 'non-profit' exception in the GPL.
> but you use the threat of legal action rather than API keys to enforce your rights. “Open Source” is in the name only.
Every software company has a license that governs it's use. I am pretty certain your company will use legal threats if it's source code got out or it's frontend code was being abused. There are likely legal clauses in your employment contract over the source code that you have signed.
of course not.
But nobody has successfully been able to charge for it, without additional things such as services/support etc.
Hard to imagine this can be upvoted to this high.
"Open Source" just means that the source code is open, not closed.
Some people just like the source code to be open (e.g. for security reasons to examine the source code.).
It is just as simple as that.
Some licenses favor the freedom of the user, and some the freedom of the developer, in different degrees. It is not a "name only" thing, unless we insist on misinterpreting it.
It is deductible in India.
Disclaimer: I'm not a lawyer, chartered accountant. Nothing I ever say should be construed as legal advice etc.
> Our tax law (in Canada) is so opaque
I was going to dismiss it as something people in every country complain about but then I tried to look it up myself:
Depending on where you're advertising, you might be entitled to either a 100% deduction, a 50% deduction or no deduction at all. And you need to take into account if the publication has atleast 80% original content to figure out the deduction. Seriously, WTAF is that?!
You can deduct expenses for advertising, including advertising in Canadian newspapers and on Canadian television and radio stations. You can also claim any amount you paid as a finder's fee.
Certain restrictions apply to the amount of the expense you can deduct for advertising in a periodical. You can deduct all the expense if your advertising is directed at a Canadian market and the original editorial content in the issue is 80% or more of the issue's total non-advertising content.
You can deduct 50% of the expense if your advertising in a periodical is directed at a Canadian market and the original editorial content in the issue is less than 80% of the issue's total non-advertising content.
You cannot deduct expenses for advertising directed mainly at a Canadian market when you advertise with a foreign broadcaster.
To be deductible as advertising, I would think the payment would need to be be explicitly for advertising services, not a dontation with the side-effect that you are recognized and thanked on a list of donors. That's just a courtesy, not an advertisement.
I can't imagine this is too different from paying a youtube or social media influencer to promote your product, or paying for an ad spot on a podcast.
I don't understand what you mean from this. What service doesn't charge? Either demos/trial versions, or they are actually charging someone else for your activity.
> Interesting? It's expecting to run on Alpine (and not dockerized Alpine)
there's still pleaty of admins who do preferer to not use containers. Conterners shine on "cattle" deployments, if you can get away with pets, that's not really something weird out of the enterprise or new age stuff.
> and sourcehut relies a lot on email.
well, it's literally a service to contribute by email, it would be weird if it wasn't like that.
> If you think you might one day make money on it, just say that up front and don’t be greedy when you finally do.
You're implicitly assuming that they're doing a deliberate bait and switch instead of just realizing after the fact that they screwed up.
There seems to be an open source fundamental attribution error here where someone changing their mind is evidence that they plotted all along to do evil.
It is also a failure to apply Hanlon's Razor--the original choice to open source license probably was ignorant/naive/stupid and not done with malice in mind.
There are certainly cases where there has been an intentional (or effective) bait and switch, but I’m not laboring under the assumption that this is always the case.
I’m not applying any assumption of malice to initial licensing decisions.
The idea I was communicating was that in some cases developers _do_ have intentions that are incompatible with FOSS licenses from the start. Not that this is always the case.
E.g. Elastic search vs. Amazon (elastic search does run upon the Linux kernel after all)
In my business we derive fantastic value from FOSS, and a huge part of that is the absence of countless other layers of profit between us and the upstream developers. So it makes sense for us to give a piece of that business back. More or less, the core of the business is all sitting atop FOSS projects.
But we’re small, so any contributions reflect that. And we do cumulatively use more FOSS software in our stack indirectly than we can sanely evaluate, so we probably won’t be seen contributing funds to every single one.
The Amazon ES shenanigans is a good example of a different “grade” of this situation, where they’re just money grubbing.
The OpenSource license is typically GPLv3 that imposes that your whole application has to be GPLv3. This is enough for corporations to avoid it.
idk. it reaaaaally _isnt_ that big a selling point to see all the open source projects one has used in a commercial application as you might be thinking
https://translate.google.com/?sl=hi&tl=en&text=%E0%A4%B2%E0%...
Isn't that mostly funded through the proprietary jetpack SaaS?
A lot of that sponsored core development comes straight from Automattic, which gets revenue from WordPress.com, WooCommerce paid services, Tumblr advertising, and Jetpack too.
> The open source license is designed for you to use Isotope to build open source and personal projects. The Isotope open source license is GPLv3. The GPLv3 has many terms, but the most important is how it is sticky when you distribute your work publicly. From the GPL FAQ:
>> If you release the modified version to the public in some way, the GPL requires you to make the modified source code available to the program's users, under the GPL.
> Releasing your project that uses Isotope under the GPLv3, in turn, requires your project to be licensed under the GPLv3. If you are okay with this, feel free to use Isotope under the GPLv3, without purchasing a commercial license.
https://isotope.metafizzy.co/license.html#open-source-licens...
Isotope's code is available under GPLv3 with no further restrictions. Although the page says that the project's use of GPLv3 is "designed for you to use Isotope to build open source and personal projects", it does not say that Isotope cannot be used for commercial projects under GPLv3.
> Although the page says that the project's use of GPLv3 is "designed for you to use Isotope to build open source and personal projects", it does not say that Isotope cannot be used for commercial projects under GPLv3.
Yeah, it's unfortunate they misrepresented the GPL and I wonder if that is how the author became confused.
I shudder at these confident, definitive statements when they're off the mark this far. It's not just "freeware" when you can see, modify, sell and redistribute under the same licence. It's open source.
As the author described it, you could _not_ sell it:
"This license allows Isotope to be used in personal or open source projects for free."
It turns out that the author was incorrect about Isotope because their software is GPL, but what they were proposing was not open source:
"There are no Pro versions, no license keys and there’s nothing else to maintain. Individual developers enjoy the same tool for free, while companies pay a reasonable price."
I quite literally grew up with the specific definition. If people decided to redefine open source software, then I'll need a new specific definition, because it's no longer useful or fit for the purpose.
And if you then object to conditions being attached to how you use it, perhaps you should also consider the onerous conditions imposed by the GPL license.
I think the word "open" belongs to OSS, not proprietary code, and I would politely suggest and recommend, you never use that word to describe your offering. (Ditto "free" :)
Just like how the word "hacker" meant one (generally positive) thing, then it meant computer criminal, and now it has come back to be somewhere in the middle.
Of course it is your right to fight for whichever definition you prefer, but the idea that anyone has a monopoly on the definition and can claim a term "doesn't mean that" is a bit silly.
Essentially this. As talking heads said,
Never for money Always for love
My girlfriend wanted to turn her life-long artistic craft (ceramics) into proper work so we opened an art studio together.
It's not hard, problem's with you, and problem is certainly with anyone arguing that you can't make money with open source. Just because you don't know how doesn't mean it's not possible.
Living off your hobbies makes for a good life.
I don’t think this is entirely accurate in practice. I talk to businesses, individuals, and enterprises every single day and they all refer to my SAS as open source software. So I don’t think everybody sees OSS like you and OSI wills them to. I’m stepping on toes by saying that, but from my sample size, that’s what I see and hear.
The only places I’ve seen this dogma is here on HN, in GitHub issues about gaslighting [^0], and on r/opensource.
[^0]: https://github.com/n8n-io/n8n/issues/40 (and other linked issues therein)
I know it gets tedious, but its useful to politely correct this terminology when it's used incorrectly.
I get that managers are not techies, but allowing them to use this term is harmful to them.
A) they heard the term from one of their programmers, which means the programmer doesn't understand the difference, which can lead to legal troubles or
B) they know the term, but use it incorrectly here, which they could then say to an actual programmer, which he then treats as open source, and can cause legal troubles.
All I'm saying is that if the words you use have a reasonable interpretation that is different from what is intended, it is better to just adopt new phrasing.
> which can lead to legal troubles
I very much doubt this. LeCun has been calling MAIR's work Open Source for quite awhile and people have been arguing about it being Source Available and not Open Source™. IANAL but I have a feeling that if OSI took Meta to court that they would lose, just like I expect Taco John to lose Taco Tuesday and how Bayer lost aspirin. I'd say that the case would probably even be clearer than those, as all Meta needs is a dictionary. I'm sure there's something similar to Roger's test for this.
Fwiw, I don't know a single person, in the flesh, that thinks Open Source is different from Source Available. My bubble is mostly nerds and techy folks. I was only introduced to the OSI definition here on HN, and I was already starting graduate school by that time, even having work experience developing software. I'm not sure what bubble I'm in or the "OSI's definition is well known" is in, but clearly we have to recognize that at least people like me exist. It seems like we're not uncommon either.
Many of them can use, modify, and distribute my SAS free of charge, like every other OSS license they know, so it actually won't get them into legal trouble. My SAS is open source to them, because the ELv2 only has a single usage restriction and that's not their use-case. For most, it's more permissive than GPL/AGPL!
I'm not going to correct them.
Part of the problem is the "Open Software Foundation" likes to take a bunch of corporate money and then pen definitions of what "really is" and "really isn't" open source that people pick up on.
I do not know how one can say such a thing with a straight face. The confusion does not require bad faith actors, but just the ability to speak English. LITERACY is what causes the confusion. It is right there in the words: _open_ source.
> Open: adjective
> 1: having no enclosing or confining barrier : accessible on all or nearly all sides
> 2a(1): being in a position or adjustment to permit passage : not shut or locked
> 3a: completely free from concealment : exposed to general view or knowledge
> Synonyms: clear, free, unobstructed, ==> _AVAILABLE_ <==
Why is there confusion? Because the source is right there, in clear view, unobstructed, and not hidden behind any doors. Yes, OSI's definition also fits the definition. It is also a reasonable interpretation of those combined words. But to say that confusing "available source" and "open source" is only caused because of bad faith actors is ludicrous when they mean the same thing. Talk about calling the kettle black.
> Guess what, most people can't make money from their hobbies... You have no entitlement to be successful in your attempts to do so.
Guess what, most people can't make money from their businesses. You're not entitled to have a successful business. So what? This is irrelevant. We're talking about the definition of a word that we can look up in the dictionary.
> The only place I've ever seen anyone pretend that "open source" can legitimately mean something other than the OSI's definition is on this forum.
Being terminally online is not an excuse for not being able to open a dictionary. It isn't an excuse for understanding that the rest of the world does not think identically to you. Go touch some grass and talk with people offline. I guarantee you that they will confuse these words. Even if we assumed "bad faith actors" caused this, well if Taco John's can lose "Taco Tuesday" and if Bayer can lose "aspirin" then OSI can lose "Open Source," and I honestly don't care. Words mean what people use them to mean. Having access to the source is a valid and extremely reasonable interpretation of the words "Open source".
I am not calling OSI a bad faith actor, but I am calling you one. Being unwilling to recognize that it is possible to confuse these meanings simply through plain literacy isn't only inane, but delusional. The reason people like me get so upset/frustrated about this is because it is like trying to prove to someone that the sky exists. We can see it, it is right there. Being obstinate isn't doing you any good. Saying I'm tricking you or there's a witch hunt isn't helping you either. Even if there was a grand conspiracy, as you have suggested, it would still be reasonable to call it the sky. The conspiracy wouldn't fucking work if it wasn't. So even that premise doesn't make a lick of sense.
Open a dictionary and talk to some people.
[0] (in fairness, you have to click the thesaurus for available, but it is under "as in public" which is exactly what we're talking about) whhttps://www.merriam-webster.com/dictionary/open
What a terrible fucking take.
First of all yes it matters. OSS developers not being able to pay their bills despite maintaining software used by trillion dollar corps has been in the news several times. It’s an ongoing issue and talked about a lot.
Second, “go get a job” is essentially “stop doing open source”. As in “I would rather have you write closed source software than source available/open core”.
This isn’t your decision to take. This isn’t your life to live. And this isn’t your culture to dictate.
Also, you live in a bubble if you think the majority of people give a crap about the OSI definition. I say this as someone who does give a crap (though not that big of one).
It probably just won't make a ridiculous amount of money for open source developers. It's the tradeoff you have to make when deciding to write open source or proprietary software.
I don't think anyone is arguing this. We're only arguing that "Source Available" is an extremely reasonable interpretation of the words "Open Source."
And I'll be honest, I don't see much of a difference between a Common Clause License and MIT. For the average person, they are functionally equivalent. I don't see this as a big deviation from the original ethics and goals but rather than it is an update because bad actors abused the openness.
It's just pragmatic. People are into open source for different reasons. I've been in it for almost two decades and it was always for the the benefits for learning, hackability, transparency, future-proofing, etc etc. All aspects which are perfectly preserved by open core.
The issue is that open source is a clear technical term that has a certain usage within the community. This caused confusion, especially since that is a specific definition that I grew up with, and existed before I was even born.
Calling it 'source available' caused none of the confusion.
A. If you mean that there is some authority that can do that, what authority other than a dictionary?(not rhetorical question)
Alternatively, B. if you mean that there shouldn't be any authority that can define the meaning of the word. In this way you can't stop people to understand it as it's literal meaning and blame them.
And sure, while ELv2 is not an OSI open license, it's not exactly a proprietary license either. So the line here is very blurred.
Your position in this case makes sense.
Open software can be distributed freely. If a programmer is told it is Open, there is a chance they will unwittingly use it in a way that is not permissable under the actual license terms.
Even as an honest mistake, this can lead to serious copyright violations that some companies will not hesitate to demand payment for.
Then educate them. Open source has had a stable definition since before I was born. It is not up for debate. All of a sudden in the last couple of years shitty people have tried to confuse people by labelling non-free licenses 'open source'. That is fraudulent.