> When software is open-source, it is open-source, not necessarily free and open-source (FOSS), and even if it is FOSS, it might still have a restrictive licence. 2 The code being available in and of itself does not give you a right to take it, modify it, or redistribute it.
Hmmm, if the license doesn't allow modification and redistribution, then it's not open source — it's "source-available." Open source, by definition, includes those rights. Just having access to the code isn't enough.
I agree. Similarly, if the license doesn't allow modification and redistribution, it does not grant the user fundamental freedoms, and is not free software, so it cannot possibly be free-and-open-source.
Can we not consider free as in beer here? I think if we can make a distinction between modification and redistribution then open source can be less proscriptive. If the source is open to view, and you are able to change that code in any way, just not redistribute your modifications, what would you call that?
I mean, <edit: previous poster is> saying it's not free software if you don't have rights to modify and redistribute
Free as in beer, as in the software is free if you are giving it away.
> I think new definitions should have new terms.
The point of the article is that we already have terms that sound like they fit. You're saying you don't agree with that as new definition and the article is saying they don't agree with your more strict definition.
Isn't that the way with everything when there's a difference of opinion on the internet. The ones that have the most funding/largest platform end up winning the argument.
The article is deliberately ignoring the history of software, which in the 90's gave a very clear definition of "open source" by the OSI as a synonym of "free software", which also historically means the four freedoms defined by Richard Stallman.
You may want to ignore what words have come to mean and try to derive a literal meaning from their constituent words, but then you need to fight established common use.
Maybe it's not deliberate, maybe they are just unaware, the 90's is a long time ago now.
It feels like OSI and Open Source advocates spend a lot of time and effort trying to fight people from using "open source" and "free software" in their literal meanings rather than the other way round. The established common use is only really within the communities that use them.
Which raises the fascinating question of whether it’s legal to sell forks of GPL software under a separate rider agreement that your customers won’t distribute it. I.e. you license it to them under the GPL (which lets them redistribute it), but sign a separate agreement with them in which they promise not to exercise that right.
I think that’s probably illegal, but I have definitely heard of companies doing it.
That’s not what I’m talking about. I’m talking about software that people are obligated to distribute under a GPL-compatible license. For example, some random company’s private fork of the Linux kernel.
> When software is open-source, it is open-source, not necessarily free and open-source (FOSS), and even if it is FOSS, it might still have a restrictive licence.
I strongly dislike this obviously controversial bit being snuck into an otherwise reasonable argument. The term FOSS is meant to unify "free software" and "open source software" and there is absolutely no reason to accept GNU's definition of "free software" and not OSI's arguably more successful definition of "open source" software. After all, "free" software, to this day, just sounds like freeware to the layperson, and meanwhile, there is other industry jargon you can use for when the source code is available but the copyright license doesn't meet these criteria.
> and there is absolutely no reason to accept GNU's definition of "free software" and not OSI's arguably more successful definition of "open source" software
The reason was to not dilute the politcal agenda of the original movement, and this dilution is exactly why "open-source" was created.
> The term FOSS is meant to unify "free software" and "open source software"
I have always assumed it to mean the intersection of the two contexts. As in, necessarily copyleft.
A layperson doesn't know what copyleft is anyway; they hardly seem worth tailoring communicating for if you're not also going to explain the difference the terms encapsulate and why people might care. Just use "open source" at that point.
As far as I'm concerned the OSI doesn't have any weight and there's little reason to think we share values sufficiently for me to start caring. Do I look like a corporation trying to slap some sense of community on a product?
> I have always assumed it to mean the intersection of the two contexts. As in, necessarily copyleft.
The definitions are very, very similar. Neither imply copyleft; almost all open source licenses are free software licenses and vice versa.
> As far as I'm concerned the OSI doesn't have any weight and there's little reason to think we share values sufficiently for me to start caring. Do I look like a corporation trying to slap some sense of community on a product?
All I'm saying is that if we accept "free" as implying the GNU free software definition, then we should accept "open source" as implying the OSI definition. There is no logical reason for why GNU's jargon is somehow better than OSI's.
> I have always assumed it to mean the intersection of the two contexts. As in, necessarily copyleft.
It's the union.
Actually it's harder than that: it's the union, but I challenge you to find a license that is open source and not free, or that is free and not open source :-).
I think there is a difference in philosophy between two parts of the movement, but in practice, what we call "open source software" is the same as "free software" (in terms of licences).
There is no Open Source software that is not free, if we are using the OSI definition of Open Source. FOSS just adds a certain emphasis for reasons of the historic tension between the FSF/GNU "free software" advocates and the "open source" advocates. FOSS doesn't mean that OSS is not free without the F.
There are a few twits out there who believe (or at least troll) that proprietary software whose source code is available is "open source". So FOSS provides the minor and largely unnecessary benefit of making it clear that it's not that.
No, "Open Source" is not "just that". This word combination has a specific meaning, not equal to a simple sum of these words. Similarly, "sparkling water", "hot dog", or "heavy metal" have meanings distinct form the simple sum of the meanings of constituent words.
The notion of "Open Source" has an agreed-upon definition, understood widely within the community of software development practice: https://opensource.org/osd
If you want a different definition, all you have to do is invent a new term. The term "open source" was invented by the people who made the definition at the URL above. (It had existing uses, but the people who made the definition above were the first to apply the term to software or software licenses.)
A bunch of for-profit companies proped OSI up to exclude "open source" projects that are not free to use for businesses so they can profit more from OSS community and you eat it up.
Did for-profit companies prop up Stallman, too? According to you they must have because his definition of Free Software likewise excludes any license (e.g., no commercial use) that restricts how the software can be used.
As does the Debian Free Software Guidelines.
(Also, I think you mean trademark case, not copyright case.)
They absolutely were NOT the first to apply the term to software in general. See https://dieter.plaetinck.be/posts/open-source-undefined-part... for many proven uses of "open source" being used to describe source-available software prior to the OSI's existence.
As for being the first to apply the term to software licenses, yes you're correct there. But that's the inherent source of this endless confusion: they took an existing term which didn't relate to licenses, and redefined it to be entirely about licenses.
It's quite ironic to say "If you want a different definition, all you have to do is invent a new term" since that's literally what the OSI failed to do!
Edit to add: I would appreciate it if downvoters would cite what specific part of the linked article you disagree with, considering that it has numerous instances of irrefutable proof of "open source" being used to describe software prior to 1998, including by famous folks like Bill Joy.
They made up the term specifically from looking at word combinations that were not in common use and could reasonably be trademarked for this specific purpose. One of the problems with the at the time popular term Free Software was that it could not be trademarked. They did a thorough job but in the end it wasn't enough.
The other points make sense to me, but I have to disagree with this one.
> It does not mean it is free and open-source (FOSS).
When the campaign to describe software as "open source" kicked off in 1998, it consciously choose the Debian Free Software Guidelines (DFSG) as the basis for the Open Source Definition (OSD), so that "open source" would refer to the same thing as "free software".
Clearly the connotations of the two terms have diverged quite a lot, but their denotations weren't meant to.
Definitions do change, but this one hasn’t. 99% of people I see using the term “open source” in the present day are using it to mean software that you can freely modify and redistribute. For proprietary software with publicly readable source code, the term “source available” is used instead.
Personal experience? I have never seen any notable case of a software project describing itself as “open source” that wasn’t also free software.
Companies that want to publish their source code but not give you the right to use it however you want, e.g. Cockroach or (my former employer) Materialize, are described as “source-available” instead.
My personal experience is 100%-0%. I've yet to see a project that uses "open source" as a synonym for FOSS or OSI. There are a lot of projects out there that use "open source" meaning "open source". HN comment section is always in outrage.
Prove it! Can you give any example of a notable software project that describes itself as “open source” despite only being source-available, and not free software? Or can you find any article in a notable publication that uses the term this way?
I rather retract it. I'm all for unambiguous communication, and if the wiki doesn't even say that the meaning is disputed[0] then I'll accept it.
> Or can you find any article in a notable publication
Well the very original meaning of open source as various parties agreed on the term in April, 1998. It did not grant the right to distribute or redistribute (possibly modified) open source libraries as part of your program. Then the meaning changed shortly after it, and they switched to the Debian social contract. (As far as I could gather.)
1) OSI says that public domain and open source are not the same thing ("Here’s why it’s a mistake to treat the two terms as synonyms"), not that public domain software cannot be open source.
2) It is simply not true that the SQLite distribution terms "contain[] a prohibition on using it for evil". That is not in the text you linked.
The OSI post concludes that "an open source user or developer cannot safely include public domain source code in a project". Has SQLite done something that makes it an exception to this?
I will concede that the exhortation against use for evil in the license text is probably not legally binding.
An advisory blog post warning people not to assume that "public domain" code is actually unencumbered is not the same as saying that actually public domain code is not open source.
> It is an attempt to dedicate a work to the public domain (which, taken alone, would not be approved as an open source license) but it also has wording commonly used for license grants.
It's clear OSI considers this extra wording, above and beyond the public domain declaration, to be what qualifies it as truly unencumbered. SQLite's license does not contain similar language, and has not been similarly qualified by the OSI.
Regardless of their dumb nonsensical rejection of public domain, you did the opposite of what was asked. I was asking for examples of open-source software that is not free in the Free Software sense, and you gave an example of something that is too free for the OSI.
> the LICENSE.md contains a prohibition on using it for evil
No it doesn’t and now I feel like you’re trying to waste my time on purpose. It contains a “blessing” exhorting (i.e., requesting) people not to do evil, not a “prohibition” of any kind.
Both of these sources contradict your somewhat bizarre thesis that sqlite, one of the most commonly used free software in the world, somehow is not open source.
It doesn't fail to meet the OSI's definition of open source. As you elsewhere conceded, the "blessing" in the SQLite source doesn't have legal weight and doesn't violate the OSD. Public domain software has always been considered open source. For instance, the Debian project, famous for their exacting standards for free software, accepts public domain software:
As they mention here, it is theoretically possible that code dedicated to the public domain might still be encumbered in a way that makes it not open source: "we are unaware of a case where a jurisdiction has upheld a copyright claim to a work which has been dedicated to the public domain everywhere".
what does "Free" mean? What do you think the author meant in this context?
The problem is that "Free" means two things in English, which is why some like to use the French/Spanish "Libre" instead, to separate the "free-as-in-speech" from "free-as-in-beer".
I haven't seen a person in this thread use free in the sense of payment yet, it seems pretty disambiguous that everyone is referring to free in the sense of liberty.
This includes the point the person you are responding to is trying to make, in which case they are noting that open source and source available are not the same things. Their point seems to be that open source, by the very definition to most who matter/care about the concept, implies the "free/libre/disentangled" portion. Whether that means derivative continuation (GPL and its kin) or not (MIT, BSD, etc).
This is a complaint as old as the internet, which probably bears repeating from time to time.
One part I think is unnecessarily controversial:
> It does not mean it is free and open-source (FOSS).
This is another old argument that it's pointless to re-litigate, but one should at least note that this way of using the term ‘Open Source’ contradicts the OSI definition and risks causing avoidable misunderstandings.
Trademarks are the legal framework we use to protect phrases like this. In 1999, OSI applied for, and was denied, a trademark on the phrase "Open Source".[1] Perhaps there is a moral argument to be made to this effect, but there is not a legal one.
The entire dictionary is nothing but consensus on terms with no trademarks (and from which government?) anywhere.
OSI merely presents a definition as a service, for others to have something already thought-out to refer to rather than have to write a 10 page definition every time someone wants to refer to the concept.
It is well established by now, and the only people trying to argue about it are simply uneducated or have some deliberate agenda where they somehow benefit from artificially clouding an issue that has already gone through a process of being hashed out and recorded long ago. There is no reason to give them any air.
I'm not trying to make either a moral or a legal argument, only to warn against using the term in a way that is liable to cause confusion or needless antagonism.
>> The code being available in and of itself does not give you a right to take it, modify it, or redistribute it.
True, in the sense that some (not Open Source) things are distributed as source code. (I do so myself.)
But the title of the article references Open Source. Capital O, capital S. If something ships under an Open Source (OSS not necessarily FOSS) then you fo have those rights.
The bulk of the article is around support though, and I completely agree with the sentiments there.
I agree with everything except the concept of software whose source has been released not being free.
If you put it, openly available, on the internet, that’s tantamount to giving it away for free, regardless of what bizarre license you release it under.
Relying on restrictions in law is fundamentally a nasty proposition.
Yet that law is, what normal people are held accountable to, whether it is nasty to rely on it or not. Of course, if you happen to be a tech giant and have almost limitless financial means, you can do whatever you want.
I understand the complaint about entitled users: because it is open source does not mean that there is a community, support, reviews or anything. Open source just says something about the kind of licence and that's it.
This said, the author doesn't seem to have a good understanding of the definition of "open source" in the context of software. It is about the licence, not about the fact that one can read the source. It's a common mistake, but it's a mistake nonetheless.
To people who would say "well, if I want to understand "open source" as "source that is open", sure: you could also consider that a "hot dog" is a kind of dog. But don't be surprised when people tell you that they eat dogs ;-).
Ignoring the open-source vs free software discussions that are bound to come about from this well said. Large companies exploiting developers and abuse towards the maintainers is probably my biggest bugbear when it comes to this.
In fact I have a similar post https://boyter.org/posts/the-three-f-s-of-open-source/ which I redirect people towards if they become aggressive towards me when I am trying to help them. Thankfully I have only had to use it a handful of times.
I think that the examples are not something most users fight with maintainers about.
- Does open source mean that the maintainers are free to ignore single line security fixes PR and refuse to handover it to well trusted and willing contributors.
- Does open source mean that maintainers could plant crypto miner or malware in the project or sell to someone who might do that?
- Does open source mean that companies could change the license and keep the product name same or remove a core functionality and migrate it to paid version?
>>- Does open source mean that the maintainers are free to ignore single line security fixes PR
Yes.
>> Does open source mean that maintainers could plant crypto miner or malware in the project or sell to someone who might do that?
Yes
>> Does open source mean that companies could change the license and keep the product name same or remove a core functionality and migrate it to paid version?
Yes
None of the things you mentioned are restricted (or required) by the license.
You may not agree with these points, or even consider them ethical, but OSS licenses allow for any of the above.
I know license allows it in the same way it allows users to harass burned out maintainers. Both are bad and we should have some ethical guidelines.
A popular open source has lot of community contributions like blog post, answers in stack overflow, issue reporting, PRs etc. And if the maintainer changes the behaviour abruptly with no way for community to fix it, even if they are ready to pay for the development cost like terraform seems like abusing the open source ethos for their advantage.
I've always found the fixation on source code rather odd. Many people know how to fix or modify anything else they own without the original design documents. Why should software be any different?
I think it's due to the impracticality of reverse-compiling and then modifying executable code.
> A program that you can run consists of a series of numbers whose meaning is obscure. No one, not even a good programmer, can easily change the numbers to make the program do something different.
I've had many experiences where the source code was available, yet finding what to change and figuring out how to rebuild the whole thing --- but without changing anything else --- was far more difficult than patching a few bytes in the binary.
To quote the title of this article, "open-source is just that".
This is the whole reason why we should be using the term "free software" instead of "open source" [2].
Free software is about user freedoms; the creator is unable to stop the user from redistributing or modifying the software. By using the term "open source", these freedoms end up getting ignored, until it simply gets diluted down to source availability, as seen in this article [1].
Actually, this article goes one step further and even says that free software is not free software; that is, it can have a license that restricts fundamental user freedoms. This is obviously wrong.
[1]
From the article:
> When software is open-source, it is open-source, not necessarily free and open-source (FOSS), and even if it is FOSS, it might still have a restrictive licence. The code being available in and of itself does not give you a right to take it, modify it, or redistribute it.
I don't agree that this is evidence of why we should use one term over the other. If you asked an uninformed person what it meant for software to be "free" they would undoubtedly and reasonably assume it meant free of charge. This isn't really any better, imo, than people thinking that "open" source just means you can access it.
First, the name sound cool, who doesn't want to be in the flow, putting its stream of consciousness into some hikigai endeavor?
It makes clear that there is something about liberty, rather than a focus on gratious in the monetary sense.
It also bring on the table that there is some work at play. So human time and attention. The polysemie helps in the sense that it also stand for the result of the work, be it software, music or some graphic design.
Open-source and FOSS mean the same thing. Any software that’s not Free Software is not open source. The only difference between the two terms is in the philosophy/ideology of the people who use them, not in their actual denotative meaning.
If you describe source-available proprietary software as “open source” you are using the term differently from practically everyone else, including the people who coined it.
open source is very simple. It means you can freely read the source code. Thats it. Its in the name. If you want to add all these addendums you need a phrase that doesnt have intuitive meaning.
The vast vast majority of people believe Open Source includes free.
This is nonsense & actively harmful posting. There are some good paragraphs about what you don't get, but up at top saying it doesn't have to be free is not correct.
Yet another article purporting to "defend" open source while simultaneously undermining its communal aspects by implying maintainers owe NOTHING to others.
Where's the vigorous criticism of CLAs? The criticism of monetization that leaves out unpaid volunteers? The acknowledgment of all the free labor provided by users, bug reporters, documenters, evangelists, etc.?
This attitude is as toxic as the users' attitudes it's railing against, just on the opposite side of the fence.
I agree partly with the article about entitlement, but you need to define what "open" means, and "access to source code" doesn't cut it due to IP laws. It can't be "open" if it's a legal landmine.
And yes, Open-Source does mean "FOSS", always. The definitions of Open-Source and Free Software being very much equivalent.
Yes, agree with the rest of the article but "open source" most certainly does _not_ include "source-available". This is true regardless of which side of the permissive-vs-copyleft issue you fall on. Being obtuse about things like this dilutes the term, which I suspect is the desired outcome. To be clear, there's nothing particularly wrong with source-available programs; they simply do not rise to the level of benefits offered by open source programs.
I'm personally not even super bent out of shape about some licensing schemes that are technically not OSD-compliant being called "open source" (i.e. ones that have targeted field-of-endeavor restrictions). There's at least a debate to be had. But what this article includes in its definition -- source code tossed over the wall that you're not allowed to do anything with -- is not anywhere close to what anybody has been calling "open source" for decades.
Sorry, this is not true. Free software simply means that the software grants the user 4 fundamental freedoms: the freedom to run, study, redistribute, and modify. It does not require virality, though it is encouraged.
From what I understand, Free Software and Open Source are largely similar, except that the former is political and moral, while the latter is mostly concerned about practical effects.
No, open source and free software are synonyms according to proponents of both terms, including Richard Stallman, the most prominent “free software” crusader. All (or at least virtually all) Free Software Foundation-approved licenses are also OSI-approved, and vice versa.
Your misconception is very common, but you, and other people with the same misconception, are conflating a few different things. Most of the people who prefer the term “free software” also prefer copyleft licenses to permissive licenses. Which is where the misunderstanding that “free software” means copyleft licenses and “open source” means permissive licenses comes from.
Yes, I'm viewing the concept of Free Software and Open Source through the political not legal lens. Sorry for that - I should have been more careful in a thread focusing on the legal parts.
> When software is open-source, it is open-source, not necessarily free and open-source (FOSS), and even if it is FOSS, it might still have a restrictive licence. 2 The code being available in and of itself does not give you a right to take it, modify it, or redistribute it.
Hmmm, if the license doesn't allow modification and redistribution, then it's not open source — it's "source-available." Open source, by definition, includes those rights. Just having access to the code isn't enough.
I agree. Similarly, if the license doesn't allow modification and redistribution, it does not grant the user fundamental freedoms, and is not free software, so it cannot possibly be free-and-open-source.
Can we not consider free as in beer here? I think if we can make a distinction between modification and redistribution then open source can be less proscriptive. If the source is open to view, and you are able to change that code in any way, just not redistribute your modifications, what would you call that?
> Can we not consider free as in beer here?
The rest of your comment didn't. How did you want to consider it?
> I think if we can make a distinction between modification and redistribution then open source can be less proscriptive.
I think new definitions should have new terms.
> If the source is open to view, and you are able to change that code in any way, just not redistribute your modifications, what would you call that?
Source available.
I mean, <edit: previous poster is> saying it's not free software if you don't have rights to modify and redistribute
Free as in beer, as in the software is free if you are giving it away.
> I think new definitions should have new terms.
The point of the article is that we already have terms that sound like they fit. You're saying you don't agree with that as new definition and the article is saying they don't agree with your more strict definition.
Isn't that the way with everything when there's a difference of opinion on the internet. The ones that have the most funding/largest platform end up winning the argument.
The article is deliberately ignoring the history of software, which in the 90's gave a very clear definition of "open source" by the OSI as a synonym of "free software", which also historically means the four freedoms defined by Richard Stallman.
You may want to ignore what words have come to mean and try to derive a literal meaning from their constituent words, but then you need to fight established common use.
Maybe it's not deliberate, maybe they are just unaware, the 90's is a long time ago now.
It feels like OSI and Open Source advocates spend a lot of time and effort trying to fight people from using "open source" and "free software" in their literal meanings rather than the other way round. The established common use is only really within the communities that use them.
The words put, make, set, and run have literally hundreds of senses. Open source is very tightly defined with only three or four.
Funny thing, “free and open-source” can be paid. You can charge all your customers, it’s just that they are free to redistribute. https://www.gnu.org/licenses/gpl-faq.en.html#DoesTheGPLAllow...
Which raises the fascinating question of whether it’s legal to sell forks of GPL software under a separate rider agreement that your customers won’t distribute it. I.e. you license it to them under the GPL (which lets them redistribute it), but sign a separate agreement with them in which they promise not to exercise that right.
I think that’s probably illegal, but I have definitely heard of companies doing it.
The GPL’s FAQ says no, you can’t give GPL software with an additional NDA. But as usual with licenses: I suppose it was never tested in court.
I think it would depend on the agreement and what you would lose when exercising your GPL rights.
For example, with RedHat your subscription for security updates/etc can be terminated.
if it's your software you can simply sell them a proprietary license separately
That’s not what I’m talking about. I’m talking about software that people are obligated to distribute under a GPL-compatible license. For example, some random company’s private fork of the Linux kernel.
Interesting thought experiment and at the same time a terrible real world experiment.
> When software is open-source, it is open-source, not necessarily free and open-source (FOSS), and even if it is FOSS, it might still have a restrictive licence.
I strongly dislike this obviously controversial bit being snuck into an otherwise reasonable argument. The term FOSS is meant to unify "free software" and "open source software" and there is absolutely no reason to accept GNU's definition of "free software" and not OSI's arguably more successful definition of "open source" software. After all, "free" software, to this day, just sounds like freeware to the layperson, and meanwhile, there is other industry jargon you can use for when the source code is available but the copyright license doesn't meet these criteria.
> and there is absolutely no reason to accept GNU's definition of "free software" and not OSI's arguably more successful definition of "open source" software
The reason was to not dilute the politcal agenda of the original movement, and this dilution is exactly why "open-source" was created.
> The term FOSS is meant to unify "free software" and "open source software"
I have always assumed it to mean the intersection of the two contexts. As in, necessarily copyleft.
A layperson doesn't know what copyleft is anyway; they hardly seem worth tailoring communicating for if you're not also going to explain the difference the terms encapsulate and why people might care. Just use "open source" at that point.
As far as I'm concerned the OSI doesn't have any weight and there's little reason to think we share values sufficiently for me to start caring. Do I look like a corporation trying to slap some sense of community on a product?
> I have always assumed it to mean the intersection of the two contexts. As in, necessarily copyleft.
The definitions are very, very similar. Neither imply copyleft; almost all open source licenses are free software licenses and vice versa.
> As far as I'm concerned the OSI doesn't have any weight and there's little reason to think we share values sufficiently for me to start caring. Do I look like a corporation trying to slap some sense of community on a product?
All I'm saying is that if we accept "free" as implying the GNU free software definition, then we should accept "open source" as implying the OSI definition. There is no logical reason for why GNU's jargon is somehow better than OSI's.
> Neither imply copyleft
The "free software" part necessarily implies copyleft.
> There is no logical reason for why GNU's jargon is somehow better than OSI's.
There's no logical reason to conflate them, either. I'm free to mock the term as I see fit.
> I have always assumed it to mean the intersection of the two contexts. As in, necessarily copyleft.
It's the union.
Actually it's harder than that: it's the union, but I challenge you to find a license that is open source and not free, or that is free and not open source :-).
I think there is a difference in philosophy between two parts of the movement, but in practice, what we call "open source software" is the same as "free software" (in terms of licences).
“Free software” does not imply copyleft, not even to the FSF. The MIT license is a free software license even by their definition.
There is no Open Source software that is not free, if we are using the OSI definition of Open Source. FOSS just adds a certain emphasis for reasons of the historic tension between the FSF/GNU "free software" advocates and the "open source" advocates. FOSS doesn't mean that OSS is not free without the F.
There are a few twits out there who believe (or at least troll) that proprietary software whose source code is available is "open source". So FOSS provides the minor and largely unnecessary benefit of making it clear that it's not that.
No, "Open Source" is not "just that". This word combination has a specific meaning, not equal to a simple sum of these words. Similarly, "sparkling water", "hot dog", or "heavy metal" have meanings distinct form the simple sum of the meanings of constituent words.
The notion of "Open Source" has an agreed-upon definition, understood widely within the community of software development practice: https://opensource.org/osd
By the same token, "Free Software" has a specific, narrow meaning: https://www.gnu.org/philosophy/free-sw.en.html
What would you tell a person who would peddle you deep-frozen apple juice as "hard cider", because it's technically cider, and is undeniably hard?
(While at it: I must praise the choice of the font on the page, super clean and legible. It's Lexend, freely available from Google Fonts.)
Who made opensource.org the king?
The definition at https://opensource.org/osd has existed with only very minor changes since 1999.
If you want a different definition, all you have to do is invent a new term. The term "open source" was invented by the people who made the definition at the URL above. (It had existing uses, but the people who made the definition above were the first to apply the term to software or software licenses.)
Is that why they lost the copyright case?
A bunch of for-profit companies proped OSI up to exclude "open source" projects that are not free to use for businesses so they can profit more from OSS community and you eat it up.
Did for-profit companies prop up Stallman, too? According to you they must have because his definition of Free Software likewise excludes any license (e.g., no commercial use) that restricts how the software can be used.
As does the Debian Free Software Guidelines.
(Also, I think you mean trademark case, not copyright case.)
The words open and source have been around far longer than that
They absolutely were NOT the first to apply the term to software in general. See https://dieter.plaetinck.be/posts/open-source-undefined-part... for many proven uses of "open source" being used to describe source-available software prior to the OSI's existence.
As for being the first to apply the term to software licenses, yes you're correct there. But that's the inherent source of this endless confusion: they took an existing term which didn't relate to licenses, and redefined it to be entirely about licenses.
It's quite ironic to say "If you want a different definition, all you have to do is invent a new term" since that's literally what the OSI failed to do!
Edit to add: I would appreciate it if downvoters would cite what specific part of the linked article you disagree with, considering that it has numerous instances of irrefutable proof of "open source" being used to describe software prior to 1998, including by famous folks like Bill Joy.
They made up the term specifically from looking at word combinations that were not in common use and could reasonably be trademarked for this specific purpose. One of the problems with the at the time popular term Free Software was that it could not be trademarked. They did a thorough job but in the end it wasn't enough.
It's called a consensus.
Playing devil's advocate, I claim that the consensus is that open source just means public code, not free licenses. Prove me wrong.
I claim that a hot dog is a dog, not some kind of sandwich. Prove me wrong.
Is that just by consensus?
So what would you call projects with restrictive licenses but with their source code made public?
Since I started on the internet in 1991, people have been calling that source-code-available software.
The other points make sense to me, but I have to disagree with this one.
> It does not mean it is free and open-source (FOSS).
When the campaign to describe software as "open source" kicked off in 1998, it consciously choose the Debian Free Software Guidelines (DFSG) as the basis for the Open Source Definition (OSD), so that "open source" would refer to the same thing as "free software".
Clearly the connotations of the two terms have diverged quite a lot, but their denotations weren't meant to.
I don't think the author knows what Free Software is, sadly.
That's exactly what happens when you muddy the waters by using an ambiguous term like "open source". People not understanding the original motive.
1998 was 27 years ago. Definitions change.
Definitions do change, but this one hasn’t. 99% of people I see using the term “open source” in the present day are using it to mean software that you can freely modify and redistribute. For proprietary software with publicly readable source code, the term “source available” is used instead.
Where are you getting that 99% from?
Personal experience? I have never seen any notable case of a software project describing itself as “open source” that wasn’t also free software.
Companies that want to publish their source code but not give you the right to use it however you want, e.g. Cockroach or (my former employer) Materialize, are described as “source-available” instead.
My personal experience is 100%-0%. I've yet to see a project that uses "open source" as a synonym for FOSS or OSI. There are a lot of projects out there that use "open source" meaning "open source". HN comment section is always in outrage.
Share some examples?
[flagged]
Prove it! Can you give any example of a notable software project that describes itself as “open source” despite only being source-available, and not free software? Or can you find any article in a notable publication that uses the term this way?
I rather retract it. I'm all for unambiguous communication, and if the wiki doesn't even say that the meaning is disputed[0] then I'll accept it.
> Or can you find any article in a notable publication
Well the very original meaning of open source as various parties agreed on the term in April, 1998. It did not grant the right to distribute or redistribute (possibly modified) open source libraries as part of your program. Then the meaning changed shortly after it, and they switched to the Debian social contract. (As far as I could gather.)
[0] : https://en.wikipedia.org/wiki/Open-source_software
SQLite
https://www.sqlite.org/copyright.html https://opensource.org/blog/public-domain-is-not-open-source
No. SQLite is public domain. Public domain is free software.[1]
[1] https://www.gnu.org/licenses/license-list.html#PublicDomain
Troll answer. You can obviously freely modify and redistribute SQLite.
It's not obvious to me. OSI (as per link in parent comment) says it's not "open source", and the LICENSE.md contains a prohibition on using it for evil https://github.com/sqlite/sqlite/blob/master/LICENSE.md
1) OSI says that public domain and open source are not the same thing ("Here’s why it’s a mistake to treat the two terms as synonyms"), not that public domain software cannot be open source.
2) It is simply not true that the SQLite distribution terms "contain[] a prohibition on using it for evil". That is not in the text you linked.
The OSI post concludes that "an open source user or developer cannot safely include public domain source code in a project". Has SQLite done something that makes it an exception to this?
I will concede that the exhortation against use for evil in the license text is probably not legally binding.
The OSI has approved a public domain dedication, so clearly they accept public domain software as OSD-compliant:
https://opensource.org/license/unlicense
An advisory blog post warning people not to assume that "public domain" code is actually unencumbered is not the same as saying that actually public domain code is not open source.
The OSI approval of the Unlicense notes:
> It is an attempt to dedicate a work to the public domain (which, taken alone, would not be approved as an open source license) but it also has wording commonly used for license grants.
https://lists.opensource.org/pipermail/license-review_lists....
It's clear OSI considers this extra wording, above and beyond the public domain declaration, to be what qualifies it as truly unencumbered. SQLite's license does not contain similar language, and has not been similarly qualified by the OSI.
Regardless of their dumb nonsensical rejection of public domain, you did the opposite of what was asked. I was asking for examples of open-source software that is not free in the Free Software sense, and you gave an example of something that is too free for the OSI.
> the LICENSE.md contains a prohibition on using it for evil
No it doesn’t and now I feel like you’re trying to waste my time on purpose. It contains a “blessing” exhorting (i.e., requesting) people not to do evil, not a “prohibition” of any kind.
What popular source-available, non-OSD software is commonly referred to as "open source"?
SQLite
Which elements of the Open Source Definition does SQLite fail to meet?
The prohibition on using it for evil[1] violates guidelines #5 and #6. And more generally, OSI argues[2] that "public domain" != "open source".
[1]: https://github.com/sqlite/sqlite/blob/master/LICENSE.md [2]: https://opensource.org/blog/public-domain-is-not-open-source
Both of these sources contradict your somewhat bizarre thesis that sqlite, one of the most commonly used free software in the world, somehow is not open source.
I think you misunderstand my thesis. SQLite is open source, despite the fact that it fails to meet OSI's somewhat bizarre definition of "open source".
It doesn't fail to meet the OSI's definition of open source. As you elsewhere conceded, the "blessing" in the SQLite source doesn't have legal weight and doesn't violate the OSD. Public domain software has always been considered open source. For instance, the Debian project, famous for their exacting standards for free software, accepts public domain software:
https://wiki.debian.org/DFSGLicenses#Public_Domain
As they mention here, it is theoretically possible that code dedicated to the public domain might still be encumbered in a way that makes it not open source: "we are unaware of a case where a jurisdiction has upheld a copyright claim to a work which has been dedicated to the public domain everywhere".
You are using your own definitions of the term and you will find that no one else agrees with that.
This is a silly game and there is nothing to be gained from playing it, for anyone.
what does "Free" mean? What do you think the author meant in this context?
The problem is that "Free" means two things in English, which is why some like to use the French/Spanish "Libre" instead, to separate the "free-as-in-speech" from "free-as-in-beer".
I haven't seen a person in this thread use free in the sense of payment yet, it seems pretty disambiguous that everyone is referring to free in the sense of liberty.
This includes the point the person you are responding to is trying to make, in which case they are noting that open source and source available are not the same things. Their point seems to be that open source, by the very definition to most who matter/care about the concept, implies the "free/libre/disentangled" portion. Whether that means derivative continuation (GPL and its kin) or not (MIT, BSD, etc).
This is why the phrase "source available" exists.
This is a complaint as old as the internet, which probably bears repeating from time to time.
One part I think is unnecessarily controversial:
> It does not mean it is free and open-source (FOSS).
This is another old argument that it's pointless to re-litigate, but one should at least note that this way of using the term ‘Open Source’ contradicts the OSI definition and risks causing avoidable misunderstandings.
https://opensource.org/osd
Trademarks are the legal framework we use to protect phrases like this. In 1999, OSI applied for, and was denied, a trademark on the phrase "Open Source".[1] Perhaps there is a moral argument to be made to this effect, but there is not a legal one.
[1]: https://opensource.org/pressreleases/certified-open-source.p...
Ridiculous.
The entire dictionary is nothing but consensus on terms with no trademarks (and from which government?) anywhere.
OSI merely presents a definition as a service, for others to have something already thought-out to refer to rather than have to write a 10 page definition every time someone wants to refer to the concept.
It is well established by now, and the only people trying to argue about it are simply uneducated or have some deliberate agenda where they somehow benefit from artificially clouding an issue that has already gone through a process of being hashed out and recorded long ago. There is no reason to give them any air.
I'm not trying to make either a moral or a legal argument, only to warn against using the term in a way that is liable to cause confusion or needless antagonism.
> Perhaps there is a moral argument to be made to this effect, but there is not a legal one
Nobody was even talking about law until you chimed in.
Believe it or not words can mean things in English independent of whether there is a court case establishing that meaning.
>> The code being available in and of itself does not give you a right to take it, modify it, or redistribute it.
True, in the sense that some (not Open Source) things are distributed as source code. (I do so myself.)
But the title of the article references Open Source. Capital O, capital S. If something ships under an Open Source (OSS not necessarily FOSS) then you fo have those rights.
The bulk of the article is around support though, and I completely agree with the sentiments there.
I agree with everything except the concept of software whose source has been released not being free.
If you put it, openly available, on the internet, that’s tantamount to giving it away for free, regardless of what bizarre license you release it under.
Relying on restrictions in law is fundamentally a nasty proposition.
Yet that law is, what normal people are held accountable to, whether it is nasty to rely on it or not. Of course, if you happen to be a tech giant and have almost limitless financial means, you can do whatever you want.
As seen in the comments here, better to use a term like “source available”. It’s also more clear because the word “open” is way too abstract.
I understand the complaint about entitled users: because it is open source does not mean that there is a community, support, reviews or anything. Open source just says something about the kind of licence and that's it.
This said, the author doesn't seem to have a good understanding of the definition of "open source" in the context of software. It is about the licence, not about the fact that one can read the source. It's a common mistake, but it's a mistake nonetheless.
To people who would say "well, if I want to understand "open source" as "source that is open", sure: you could also consider that a "hot dog" is a kind of dog. But don't be surprised when people tell you that they eat dogs ;-).
Ignoring the open-source vs free software discussions that are bound to come about from this well said. Large companies exploiting developers and abuse towards the maintainers is probably my biggest bugbear when it comes to this.
In fact I have a similar post https://boyter.org/posts/the-three-f-s-of-open-source/ which I redirect people towards if they become aggressive towards me when I am trying to help them. Thankfully I have only had to use it a handful of times.
I think that the examples are not something most users fight with maintainers about.
- Does open source mean that the maintainers are free to ignore single line security fixes PR and refuse to handover it to well trusted and willing contributors.
- Does open source mean that maintainers could plant crypto miner or malware in the project or sell to someone who might do that?
- Does open source mean that companies could change the license and keep the product name same or remove a core functionality and migrate it to paid version?
I have seen multiple instances of all of these.
>>- Does open source mean that the maintainers are free to ignore single line security fixes PR
Yes.
>> Does open source mean that maintainers could plant crypto miner or malware in the project or sell to someone who might do that?
Yes
>> Does open source mean that companies could change the license and keep the product name same or remove a core functionality and migrate it to paid version?
Yes
None of the things you mentioned are restricted (or required) by the license.
You may not agree with these points, or even consider them ethical, but OSS licenses allow for any of the above.
I know license allows it in the same way it allows users to harass burned out maintainers. Both are bad and we should have some ethical guidelines.
A popular open source has lot of community contributions like blog post, answers in stack overflow, issue reporting, PRs etc. And if the maintainer changes the behaviour abruptly with no way for community to fix it, even if they are ready to pay for the development cost like terraform seems like abusing the open source ethos for their advantage.
OSS is a license. Ethics are not legal, so don't form part of a license.
You are of course free to determine your own ethical standard and the use, or don't use, companies that apply that standard.
>> abusing the open source
You say "abusing" - they would say "using".
Tiny caveat: With a viral license and the absence of a CLA or such, a company might not be able to meaningfully change the license. (AIUI; IANAL)
Yes, if the copyright is held by multiple contributors then they would all need to agree before changing the license.
I am confused. Has the author (and most of the comments below) not heard of the "Open Source Definition", or is it not considered relevant anymore?
https://opensource.org/osd
I've always found the fixation on source code rather odd. Many people know how to fix or modify anything else they own without the original design documents. Why should software be any different?
I think it's due to the impracticality of reverse-compiling and then modifying executable code.
> A program that you can run consists of a series of numbers whose meaning is obscure. No one, not even a good programmer, can easily change the numbers to make the program do something different.
Source: https://www.gnu.org/philosophy/shouldbefree.en.html
Stallman clearly had zero exposure to the cracking scene, nor the PC/microcomputer community.
Back in the old days, we only needed a hex editor and disassembler, maybe sometimes a debugger, to "make the program do something different".
PC magazines would publish binary patches too.
IDA showed up sometime in the 90s but it was extremely expensive, and cracking it became a "rite of passage".
Now, Ghidra is free and there are tons of other tools available.
> Stallman clearly had zero exposure to the cracking scene, nor the PC/microcomputer community.
I would say it was clear you had zero experience writing for a general audience if I wished to make unfounded assumptions like you.
You overlooked he wrote easily maybe? Cracking IDA was a rite of passage because it demonstrated skill.
Some software is easier to hex edit that figure out how to compile ;)
Is this a serious question?
The answer is: because modifying software without the source code (i.e. reverse-engineering) is extremely difficult.
Depends what you mean by "extremely difficult".
I've had many experiences where the source code was available, yet finding what to change and figuring out how to rebuild the whole thing --- but without changing anything else --- was far more difficult than patching a few bytes in the binary.
To quote the title of this article, "open-source is just that".
This is the whole reason why we should be using the term "free software" instead of "open source" [2].
Free software is about user freedoms; the creator is unable to stop the user from redistributing or modifying the software. By using the term "open source", these freedoms end up getting ignored, until it simply gets diluted down to source availability, as seen in this article [1].
Actually, this article goes one step further and even says that free software is not free software; that is, it can have a license that restricts fundamental user freedoms. This is obviously wrong.
[1]
From the article:
> When software is open-source, it is open-source, not necessarily free and open-source (FOSS), and even if it is FOSS, it might still have a restrictive licence. The code being available in and of itself does not give you a right to take it, modify it, or redistribute it.
[2]
https://www.gnu.org/philosophy/open-source-misses-the-point....
I don't agree that this is evidence of why we should use one term over the other. If you asked an uninformed person what it meant for software to be "free" they would undoubtedly and reasonably assume it meant free of charge. This isn't really any better, imo, than people thinking that "open" source just means you can access it.
> This is the whole reason why we should be using the term "free software" instead of "open source"
The definitions in my head go something like this:
Open Source -> Open (e.g. free and unencumbered) & Source available
Source Available -> Only source is available, restricted under some odious license presumably
Free Software -> You can download this (generally in binary form) and use it for free.
Well you're wrong. The latter is referred to as freeware.
Although the closeness of "freeware" and "Free Software" is a problem IMHO.
Free (as in freedom), open (like a book).
I like to use FLOW for free libre open work.
First, the name sound cool, who doesn't want to be in the flow, putting its stream of consciousness into some hikigai endeavor?
It makes clear that there is something about liberty, rather than a focus on gratious in the monetary sense.
It also bring on the table that there is some work at play. So human time and attention. The polysemie helps in the sense that it also stand for the result of the work, be it software, music or some graphic design.
The MIT license spells this out extremely clearly IN CAPITAL LETTERS. It's actually really refreshing to read.
Open-source and FOSS mean the same thing. Any software that’s not Free Software is not open source. The only difference between the two terms is in the philosophy/ideology of the people who use them, not in their actual denotative meaning.
If you describe source-available proprietary software as “open source” you are using the term differently from practically everyone else, including the people who coined it.
open source is very simple. It means you can freely read the source code. Thats it. Its in the name. If you want to add all these addendums you need a phrase that doesnt have intuitive meaning.
Regardless of how intuitive you think it is, that’s not what it means, and not what anybody has been using it to mean since the 90s at least.
The vast vast majority of people believe Open Source includes free.
This is nonsense & actively harmful posting. There are some good paragraphs about what you don't get, but up at top saying it doesn't have to be free is not correct.
Open Source Definition: https://opensource.org/osd
Yet another article purporting to "defend" open source while simultaneously undermining its communal aspects by implying maintainers owe NOTHING to others.
Where's the vigorous criticism of CLAs? The criticism of monetization that leaves out unpaid volunteers? The acknowledgment of all the free labor provided by users, bug reporters, documenters, evangelists, etc.?
This attitude is as toxic as the users' attitudes it's railing against, just on the opposite side of the fence.
Eh, no, Open-Source means Open-Source and I hate it when people try overloading words due to market-appeal or ignorance.
https://opensource.org/definition-annotated
I agree partly with the article about entitlement, but you need to define what "open" means, and "access to source code" doesn't cut it due to IP laws. It can't be "open" if it's a legal landmine.
And yes, Open-Source does mean "FOSS", always. The definitions of Open-Source and Free Software being very much equivalent.
Stop overloading terms for your own gain.
Yes, agree with the rest of the article but "open source" most certainly does _not_ include "source-available". This is true regardless of which side of the permissive-vs-copyleft issue you fall on. Being obtuse about things like this dilutes the term, which I suspect is the desired outcome. To be clear, there's nothing particularly wrong with source-available programs; they simply do not rise to the level of benefits offered by open source programs.
I'm personally not even super bent out of shape about some licensing schemes that are technically not OSD-compliant being called "open source" (i.e. ones that have targeted field-of-endeavor restrictions). There's at least a debate to be had. But what this article includes in its definition -- source code tossed over the wall that you're not allowed to do anything with -- is not anywhere close to what anybody has been calling "open source" for decades.
>And yes, Open-Source does mean "FOSS", always.
This does not align with reality, irregardless of what some private company says or how strongly you feel about software freedoms.
Yes it does. Here in reality, 99% of people use “open source” and “free software” to mean the same thing.
Do you have examples of any notable software describing itself as “open source” that isn’t FOSS?
Free Software is a viral license. Open Source is not. So there is a difference, and I'd argue that Free <> Open Source, although they're similar.
Other than that I 100% agree either you.
Sorry, this is not true. Free software simply means that the software grants the user 4 fundamental freedoms: the freedom to run, study, redistribute, and modify. It does not require virality, though it is encouraged.
From what I understand, Free Software and Open Source are largely similar, except that the former is political and moral, while the latter is mostly concerned about practical effects.
No, open source and free software are synonyms according to proponents of both terms, including Richard Stallman, the most prominent “free software” crusader. All (or at least virtually all) Free Software Foundation-approved licenses are also OSI-approved, and vice versa.
Your misconception is very common, but you, and other people with the same misconception, are conflating a few different things. Most of the people who prefer the term “free software” also prefer copyleft licenses to permissive licenses. Which is where the misunderstanding that “free software” means copyleft licenses and “open source” means permissive licenses comes from.
Yes, I'm viewing the concept of Free Software and Open Source through the political not legal lens. Sorry for that - I should have been more careful in a thread focusing on the legal parts.
It’s not a question of what lens you’re viewing it through — your initial comment is simply wrong.
Viral licenses e.g. GPL are _both_ free software and open source. Permissive licenses e.g. BSD are also _both_ free software and open source.
Legally yes. But in the political sense Free Software favors virality, whereas Open Source does not.
And again I apologize for viewing it through that lens when discussing the legal aspects, not the political aspects.
But you said:
“Free software IS a viral license”
Not “free software as a political movement tends to prefer viral licenses”.
Why is it so hard to just admit you were wrong?
Now i have an article i can refer people to who ask me for a deadline on OSS features they request. ;)
[dead]