@osiralonI think he’s saying that as a company your argument is an “erm akshually [insert insane loophole here]” as a moral argument, it’s not really moral
@osiralon That's not what I said. I said things can't be morally and "technically legally" right at the same time. And by "technically legal" something like abusing loopholes, which is immoral, especially for misdirection and misinformation of consumers. TLDR, I refer to things that are legal only by technicality, not all legal things.
OSI with all their corporate sponsors have developed the definition of open source with a lot of room for exploitation of the developers, unknowingly or not. No wonder why a lot of companies use Open Source projects and never donate back or support their infrastructure. Because they are not required to do that. I will be more than happy to call my project open source despite being just a "source available". In my definition, Open Source just means that the code is available to be viewed by the public. Nothing more. If you don't agree with OSI regarding the term Open Source, it's fine. You are not legally required to follow the definition. It's the OSI with all their corporate sponsors that is the problem, not the ""fake"" open source projects that claims to be open source because they disagree with the OSI.
@@erich_ika Just be mindful when commenting on videos that attract programmers. I too was misled by your statement at first and had to read it three times over before I understood what you meant, and as-is without quotation marks (if everyone would interpret that as indicating foul play or exploitation) does not make a logical argument.
llama AI model is still Open Source. Same as you can open source images, photos, fonts. You are not required to open source tools that you used to create them.
@@dyto2287 they release the dataset and training code to make it 1:1 reproducible? If not then It's open weights which is like praising a company for releasing compiled binaries.
@ayesaac the 4 freedoms only talk about the software itself, not derived works. While the fsf encourages the use of restrictive licenses like the GPL, it recognizes permissive licenses as free software licenses as well, including the MIT license. In particular, MIT is one of the GPL compatible licenses. basically, while the MIT license grants you the 4 freedoms itself, it also allows you to create derived works that don't grant those freedoms. GPL requires derived works to grant the freedoms as well. (and LGPL & co allow for non-free programs to link to free software, essentially sitting between the permissive MIT and restrictive GPL)
@ayesaac distributors are required to make the source code available to anyone who receives a copy of the software. This means that while the software itself can be sold or distributed under various business models, the source code must be provided upon request. The source code needs to be available in its "as-is" state, meaning it should be provided without modifications.
@ayesaac as a user of MIT code, you have all the freedoms. Only if you're a contributor and you want to make sure that everyone that ever gets your code has those freedoms too, you can license the code with GPL. When you contribute code to a GPL project, you license your own as GPL too. No difference. The MIT code just allows you to license it differently as well. As in, "not allow your code to end up in a closed source software" is not your user freedom, it's your creator/programmer freedom that you ensure by licensing your own work accordingly.
Google with android is a even clearer example the software itself is open source, but google removed as much as possible to make it nearly mandatory to use their bloatware to get any use out of it
Yea the core it self is open source you can download it online and it has no google things what so ever but when you sell it is replace a lot of the open source features with closed source features so if you want real open source android you should go with 3rd party roms.
Also while it took them a really long time, they eventually kicked out Chromium builds that used official API keys of Google Chrome to sync with Google services.
Just a little tip: if someone is avoiding saying the common term for something, then it's likely intentional and not actually that thing. E.g. "Open Code" instead of "Open Source". This is also done in the world of selling food where for example if something isn't cheese, they can't write cheese on the packaging in a way that implies it's cheese. So they find clever workarounds, for example just saying "grated" rather than "grated cheese". So people not looking carefully enough into it, for example only looking for the word "grated", but wanting "grated cheese", they may be tricked into thinking they are buying "grated cheese". (I have actually seen products labeled "grated" because they don't actually contain cheese, or not enough cheese to call it cheese)
@@dimedriver we have records showing that almond milk was referred to as milk in 1390 AD (The Forme of Cury). Nut milks have been referred to since over 200 years before the microscope was invented. You're just spreading dairy lobby propaganda, nonsense made by a trillion dollar industry that sees it as a legitimate business strategy to force the government to ban their competitors by spreading the lie that consumers only stopped buying their product because they were mistaken.
I often look for the license, and what it's implications are, rather than reading their marketing. Since if I see a MIT, GPL, BSD, Apache, zlib, etc, I know that it's actually open source. If I see some other weird license, I either quickly look it over, or look it up online, that way I know if it actually is open source, or just source available.
This is the fundamental problem with OSI's definition of open source. The term is far too simple to describe anything in such detail, particularly for people who aren't already well-acquainted with the more restrictive definition.
@@trajectoryunown No, the problem isn't with the OSI definition. It's with companies that want the marketing shine of "open source", but not the downsides when it comes to commercial exploitation.
If I plan on just using it myself, I don't care about the license, but if I just have plans to potentially use the source, then I absolutely check the license. Just to understand if it's a permissive one, a copyleft type, etc.
Which, waaayyy back then, may have been the case since they were made in ASM and designed to be simple, but in an age of compilers, JIT, and dealing with OS stuff too, that's not nearly as easy.
@rakshith-ravi Only if it was originally written in assembly. "Source" means the origin form that the software was written in, not an intermediate, compiled or transformed form.
@@k225yeah, and we can only really decompile old and antequated software you can decompile just about anything written in C, but it'll give you very compressed and truncated code
I can read assembly and even decode quite a few instructions easily by hand just reading the hex printout for a few arches, but it doesn't mean I can understand a whole program from binary. 🙄
Crazy how Elasticsearch annouced they are going open source the same day this video came out (by adding AGPL as a license option, but they still have the other licenses)
Wait... WHAT!? They actually did just make that announcement about 8 hours after the video went public, didn't they? Can't wait for the deluge of "um actually" comments once they follow through with that. 😆 It looks like the AGPL will not apply to the binary downloads of Elasticsearch, so there still a few blurred lines; but it's a step towards clarity for sure and they'll legitimately be able to use the phrase "open source" again.
I'm curous why AGPL won't apply to the binaries. I'm assuming that includes future binary releases. Would it be possible for a third party to compile their own binaries and release them under the AGPL license? I have so many questions. Can you share a link to where you got that info? Maybe I missed it but I didn't see that being mentioned anywhere on the announcement.
They've added it to this page. The current binaries aren't covered by the SSPL, either. I assume you can compile them yourself (or wait for someone else to, like Opensearch's predecessor, the Open Distro). www.elastic.co/pricing/faq/licensing
@@ProTechShowProbably to lure you into paying for commercial support, and this is a fine business model. You can choose AGPL if you compile it yourself. Also, AGPL is still FLOSS. It requires every user, even if over a network, is free, but doesn't require that everyone is a user, and doesn't even require the rest of your app to be FLOSS.
Foss is collective effort to benefit humanity as a whole. Corporations who publish their "source available" or similar non foss source codes are just fishing for free labor for commercial product.
I heartily disagree. Making the source code accessible to the public makes businesses accountable to the general public. Developers aren't able to hide security vulnerabilities or pretend like additional functions unrelated to the user's intention for the software aren't there.
I don't have the figures to back this up, but I suspect that in most cases the free labour they get in public contributions is not worth the effort that goes into maintaining the repos. If they're fishing for something I would guess it's more likely credibility/reputational.
Actually, a lot of times it's to allow their customers to be able yo look at the code, so they can better debug issues with their own code. It's not about soliciting fixes from 3rd parties.
@@ProTechShowMost certainly it helps them in finding bugs and other stability issues that users share in they GitHub issues page thus making their software more robust over time
Same reason why some wine is called "champagne" and some is just "sparkling wine" (at least here in the EU). Only code written in Söurke region of Finland should be allowed to be called "Open Source" and everything else is merely "sparkling code"
I expect the companies in question will have had their legal teams check this and would point at the software licence as the actual agreement. The language is clear in the licence itself, but their marketing doesn't match the licence. Of course they will know that most people don't read the licence, but they'll say that's their own fault.
@@ProTechShow having tos that goes completely against all your advertising is false/misleading/manipulative advertising and a bad strategy if you're not trying to scam people "your honour, they used/bought our product because of our intentionally false advertising and thus broke our tos (that's often designed to be so complicated that the layman can't fully understand it)" is not an argument that would win a fair trial
I hate to be the one to tell you this, but the law doesn't require plain language it only requires interpretation if a lawyer can make a strong enough argument to a judge that black is actually green, and that judge on a federal level identifies with that lawyers interpretation being fact in accordance with the jury standing by as witness then that becomes a precedent to be used latter on to make cases for that interpretation being the standard. In other words it doesn't matter what you, or I think cause the law is a tangled web of arguments built on a backbone of lies and deceit just to get precedent... if you don't like that, then your free, and open to get a law degree, and fight your own interpretation, and hopefully that will become the new standard, I truly do wish you the best, but that's all I can wish for you, because the law is a heated world almost no one wins because interpretation often is born from the financial, and political incentive to stay in line, and not always to make the world better for the everyday individual. it sucks, but that's life. we all have to play the hands we're dealt, and if that means folding cause we can't always get a full house in life, then that's better off then busting, and loosing it all on a gamble we have no hope of winning.
Privacy is also a problem. The brave browser is a good example. You can't turn off the leo ai and it will read and look at everything you do. Same as all the other browsers.
I remember being disappointed when David Murray claimed he was going to "open source" his game Attack of the PETSCII Robots, only to then put the code under a source available proprietary licence. That would be fine if he used the right terminology, but he didn't.
I feel like his case isn’t him being intentionally misleading, rather, I think he probably doesn’t understand the difference between open source and source available. He is on the older side and is mostly interested in older technology so I’d cut him some slack
An example that really toasts me is Gab, which took its fork of an AGPL codebase almost completely private (and deleted all the Git history) for "security reasons". Last i checked, they only published a questionably-current objectively-incomplete source tarball once every 2 months.
Oh wonderful, companies have realized they can use the label “open source” the same way food companies realized they can put “Keto” on anything that they put an absurd amount of fibre in since a lot of people just subtract the fibre count from the carbs.
@@tolkienfan1972 not all fibres. Lignin is a non carb component of dietary fibre as an example. Using some of these specific kinds of fibres that don’t add to the carb total can allow them to release a product that has 10g of fibre and 10g of carbs, but none of those 10g of carbs come from the fibre, making it so you can’t perform the basic subtraction to get the remaining carb content.
vscode is open source... untill you press microsofts download button. Then, telemetry will be added, and some features like using microsofts extensions will be enabled.
People come to open source for a variety of reasons. Admittedly my original reason to prefer open source was to "stick it to the man," and still is. I'm a free software advocate in the rms tradition. Whether or not rms himself, is problematic, I'm with him 100% on the philosophical questions. Even the Open Source Definition is some openwashing, and even that isn't enough for commercial operators, who of course game that system, as they do all systems. For me, ever since I learned what free software was, back in the 20th century, have operated with the assumption that the opposite of "open source" is "commercial."
@n8chz But that is our (the-not-the-man) gaming. There's nothing in any version of the GPL that prohibits commercialization. I, too, use GNU licenses for ethical, moral, or philosophical reasons, and I tend towards an evolution away from capitalism, but Linux is the most popular software in the world, commercial and non-commercial alike.
I never thought open source meant free to distribute. I thought open source meant the source code is open to view, thus potentially modifying it before compiling it
The term "Open Source" was first coined to avoid the common confusion that the "free" in "Free Software" related to price, when it was explicitly about the freedom to "run, copy, distribute, study, change and improve the software". Businesses had a habit of hearing "free" and running in the opposite direction before people could explain it was about freedom, so a group within the free software movement rebranded it as open source to avoid the confusion (and create a bit of distance from FSF politics). The freedom to modify and distribute software is foundational; but some companies would prefer to have their cake and eat it.
@@ProTechShow maybe I am just too young to see it by what it was originally made to be, there is too much complexity in what open source should mean and it isn't legally binding to begin with.. Maybe open source needs an other rebranding, it sounds to me that open source should be replaced in favor of a more concise and legally binding license that specifies what can and cannot be done with the source code
That would be a lot simpler, but good luck getting people to agree on the terms! There are over a hundred software licences approved by the OSI alone, and presumably many more that meet the definition of open source out there as well. You'd think at least one of those would work for any given project, but people can't seem to help making more to incorporate their unique tweaks...
@@nxtvim2521 I know that there are various licenses, I am just saying FOS should be retired in favor of a name that denotes a license, because it will be specific and legally binding unlike FOS.
From reading your responses to comments, I think we have a semantic difficulty. "LIbre" is not synonymous with "open source". Pedantically and culturally. E.g, the difference between the acronyms "FOSS" and "FLOSS" is the libre, the _copyleft_. Libre software positively protects freedoms as detailed in the license. It's kinda the whole idea. Libre: Copyleft software which protects the freedoms of the creator and user. The prototype is the GNU Public License. Copyleft: Use of copyright law to affirm the right of an author to set licensing terms. Open Source: Software for which source code is available. You _could_ add that binaries should be possible to build, and that a good faith effort has been made to make it possible, but it's not at all required. We can use the OSI, although I don't 100% approve of everything they do, to define this term. BSD, MIT, GPL, MPL, Apache, etc. are all "open source". FOSS: "Free and Open Source Software". Roughly aligned with OSI "open source", but coined to differentiate between "source available" and "open source" as users and developers know it. FLOSS: "Free, Libre Open Source Software". Again coined to disambiguated, this time between copyleft (libre, GPL) and non-copyleft (MIT/BSD) licenses. You really have to start from the 1970s to have a handle on it. Start with "Letter to Hobbyists" and the story of Basic, and RMS's printer. And no, advocates did not coin "open source" because .
You're far from the only one. The trigger for this video was witnessing a couple of near-misses where people were deploying software they thought was open source, in a way that would violate its actually proprietary licence.
If you want to release your source code without actually being free as in freedom, go right ahead but be honest about it. Freedom means I can use it for any purpose, learn from and study it, modify it for any purpose and redistribute my modified version whether gratis or for a distribution fee. If you don't care about freedom and only about free labor from the community be honest about it. Likewise we shouldn't provide free labor for code whose creators don't care about freedom.
This is so weirdly lacking in nuance. Here's a reality; Corporations fund Foss projects by and large. Your talk about freedom basically means Amazon sized companies can abuse the free software within their closed system no problem. This is literally enriching the biggest while they don't have to do anything.
(Note: I did not look into the exact licenses for the data Meta uses and as far as I'm aware they at least partially use very open datasets like The Pile. I'm talking about the general case, sparked from this example) That's (afaik) true, but I feel like that's a bit of a problem with "AI" (in this context). "AI" is (imo) much more about open *data* than it is open source - at least in the current state of things. Many advancements in the field *are* published openly and often under very open licenses (which is great and I'm glad to see that) But the meat of what a model can do is in the training data. (It's a different discussion how this data should be treated etc. but I just think talking about the source code - especially in the case of base models - is by far the smallest piece of the puzzle)
They can... sort of. They can't revoke the GPL from existing, published code. What tends to happen is they will change the licence and everything after the change gets the new licence while everything before the change retains the GPL. Someone can take the old version, fork it, and create a parallel project that sticks with the GPL. They can only do this if they own the copyright, or the copyright owner has granted them the necessary rights. The copyright owner doesn't require a licence to use the code because it's their property; so they aren't limited by the terms of the licence themselves. The licence is what lets other people use it. This is why companies often require contributors to sign CLAs, and some consider it a red flag, because if means they can change the rules in the future. If every contributor retains the rights to their code, though; it would be almost impossible to change the licence, unless you can get all of them to agree.
@@ProTechShow PySimpleGUI did exactly this. It was available for years under an open source license (LGPL I believe), but never accepted code contributions. Earlier this year they switched to a proprietary license then proceeded to *nuke the github repo and PyPI packages*. As you'd expect, this broke literally every project that doesn't use the new proprietary version. The relicensing was annoying enough, but deleting the code and pip packages feels downright malicious. Fortunately, forks like FreeSimpleGUI popped up almost instantly
This reminds me of a video pointing out that at least one company recently redefined words such as sale and purchase to mean borrow until we say you cannot.
This is the most nuanced take i seen in a while about open source and adjacent licenses, this is the way to go to avoid being unclear stating licenses while being able to contribute to software without having to deceive the public
you touched on microcode and firmware, that was where i was first thinking. stuff like graphics drivers that are 'open' but full of references to private corporate bugs, workarounds and other internal crap. other drivers full of misc i/o and stuff linked to firmware that's cryptic. even the "we have to provide source/license" stuff like android or some one off libraries nintendo and whoever use can be super minimal and useless in some cases.
I immediately understood that "built as an open source project" meant it wasn't an open source project, but only because I've learned to be fluent in corporate-speak. Whenever words appear in strange-sounding combinations, it's because the more natural combination of words would be a lie. Something called a "cheese product" means that they couldn't legally call it cheese. Something called "____-style" doesn't fit the requirements of _____. "We don't sell your data to unaffiliated companies" means we have an affiliate program for selling data. "All donations will increase our contribution to charity X" means that they'll contribute some fixed percentage of each donation. "100% of donations go toward our relation with charity X." means that some of the donations are spent on tangentially-related marketing and salaries.
The worst part is there is actually a valuable distinction to be drawn between companies that work "in the open" (public documentation, API, sometimes even source available) vs the pricks that want an NDA as table stakes. They just need to make that distinction without squatting on real open source.
I could actually see a possible sane license here that is something like a GPL non commercial that "expires" to GPL after a while, and I would be more fine with this sort of behavior then but currently this is just too much
1:31 true open-source: - the license states it's open-source (not technically a prerequisite, only legally) - it's fairly easy to get the source (actually: a revision tree such as git repo) - it's actually possible to get it working (sources can be compiled and executed ; very often this is not the case) - it's possible to contribute (ie. it's possible to discuss with the devs) - all components (ie. client AND server) are open-source 8:15 yeah this is just called source code and project management. because open-source projects do it so well is only the reason why commercial stuff is using the open-source tools and workflow
The problem with Open Source is, that users of Open Source always thinks that the Open Source Coders can work like slaves, since it is open source. Most guys don't help or donate to Open Source project. And then they wonder why there are unfixed bugs in Open Source if the coders can't pay the flat or meal. The second problem with Open Source is, that most commercial guys also don't care about Open Source. They just take the source and use it, even it is copyleft. They don't care about it and the small open source projects have no money to care about their rights and/or have trouble to proof that the commercial guys have stolen the code, even they sometimes knows exactly (because they asked questions about the algorithms in the forum of the software or by email, the saved files are nearly the same (except of the changed software name, ...).
Definitely a challenge, and I don't have a problem with source available licences in principle trying to solve this. It's when they are misrepresented it becomes an issue for me as I've come across a few instances of people using them in good faith as what they thought was open source, but they were actually in violation of the less open terms buried in the licence that nobody bothers to read. Technically, their fault for not reading the fine print; but don't set them up for it by advertising something else in the first place. Beyond that, if a developer wants to go source available to protect their revenue and they are upfront about it, I don't see a problem. It's the user's choice to take it or leave it.
@@ProTechShow Well, for some developers it is very difficult. They can't just go commercial or closed source. For example the developers of xz. No user knows them and we saw what critical security problems happens because of that (and there are a lot of other similar critical projects). Nearly nobody knows how to code such imported and critical source. And large parts of Linux distributions depends on work of coders, that can only code sometimes in their spare time for free. That is in fact a no go out of the view of security. Linux and open source users need to know that good software is only possible, if users help. A Linux user can see how he could support large projects, since some of them ask for donations on their home pages. But how to support all the other highly needed coders? They also complain about missing donations and need to earn their money with other jobs. I am also 20 years open source coder now. The software is used by several 10.000 people, but donations are so low, that even the main author can't live from that, even he is working full time on the project. He can only pay bills, because his family support him and his country pays a small social welfare (not for coding, just the "normal" welfare for unemployed citizens).
There are business models though: * Distributing libraries under the GPL which forbids use in nonfree projects, then selling exceptions. * Similarly distributing server software under AGPL and selling exceptions. * Selling support services for the software. * Selling binaries while keeping source free. * Getting paid by companies (like how Mozilla does for making Google the default search engine).
The Affero license is a thing of beauty. And you can make money with it, if that's your thing. GPLv3 and AGPLv3 are my default licenses when creating a project, and I'm not the only one. There's a possibility that the GPLv2 vs GPLv3 will heat up again soon, as the consequences of hardware openness, or lack of, becomes ever more consequential.
Haven't seen the discussion, but from a quick Google I think they've made the right choice by calling it "source first" instead. Yet another term to add to the pile, but at least it's distinct and avoids confusion.
My favourite license is MIT. And I make and use open source software. But SAL (Source Available License) licenses are not bad. They allow to have source code open while enabling profits from software which is sometimes needed, for example Aseprite. We need more standard open source software, but if companies start using SALs we all would benefit. SAL is way better than closed source. FYI, I am not advocating SSPL, I am advocating this family of licenses. "Source-available" is best wording for these licenses.
Fake Open Source might be better than Open Source in some (or all) cases. For example, I would prefer using an Ethical Open Source lisense instead of Open Source lisense if the was one.
The SSPL needs to exist because behemoth companies thought they could get away with it. Is that good? Nope. Is that the only way? Alas, yes. Do I like it? Nope.
SSPL is complicated, I think it's a genuine attempt at stoping big tech companies from just stealing all the effort that goes into opensource by making these opensource projects into products on their platforms. Which seems quite unfair.
I kinda disagree with the common "open source" definition, I think it should mean "the source is available" and that's it. After all, the words "open source" only say that the source is open, available to be viewed and modified. Not necessarily distributed. FOSS, or "free and open source", actually mean "the source is available and you're free to do what you want with it, and I think that people should start using it to mean what it's usually implied with just "open source". Maybe it's an unpopular opinion, but it's my opinion :)
That's better referred to as "source available" so it doesn't get mixed up. There is a difference between "free software" as defined by the FSF and "open source" as defined by the OSI, although they are closely related. I didn't cover the FOSS term directly in the video because it would have meant applying both the FSF and OSI definitions. None of the examples used claimed to be FOSS, so it seems unfair to apply the slightly stricter FSF definition.
I'll ask the wife... 😬 I'd be a terrible RMS, anyway. I don't care what licences other people choose. I just don't like people being misled. Be open source, or do not be open source, but don't claim to be something you're not - my take.
@@ProTechShow In order to transcend into a pure RMS being, one must first act weirdly around everyone who isn't an adult male and carry around absurd amounts of cash. And use Emacs as their interface to everything. Edit: One of my late friends worked on XEmacs and wrote the (first?) book on open source licenses back in the day.
Sorry, Barry. I couldn't read your comment because it requires JavaScript. Please come to my follow-up video which will actually be a stage production to avoid the use of any non-free audio/video codecs. Am I getting there? 😁
Where I think the biggest issue lies, is that people assume Open Source = FOSS/Libre. Open Source just means that the source is available ffor you to look at the code. FOSS means that you can actually participate in the development and can fork the project if you don’t agree with the maintainers opinions. For example, SQLite is widely considered open source, but it’s not actually FOSS, as they don’t allow outside/external contributons, making it non-FOSS, albeit open source.
Open source does (or should) not just mean the source is available. That's "source available". Open source came from the Free Software Foundation's definition of "Free Software". The term "Open Source" was first coined to avoid the common confusion that the "free" in "Free Software" related to price, when it was explicitly about the freedom to "run, copy, distribute, study, change and improve the software". Businesses had a habit of hearing "free" and running in the opposite direction before people could explain it was about freedom, so a group within the free software movement rebranded it as open source to avoid the confusion (and create a bit of distance from FSF politics). The freedom to modify and distribute software is foundational to open source. I'm not sure where you get your definition of FOSS from? Usually, it is used as an inclusive term to mean software that meets the definition of either Open Source (OSI) or Free Software (FSF). I've seen people use it to mean software that meets both definitions, although that is redundant given that Free Software is also Open Source. The FSF considers SQLite to be Free Software. directory.fsf.org/wiki/SQLite
There is another type of fake open source software: the software that lacks 30% of its code because that part is not open source. Unfortunately, without these parts you cannot even build the software.
And I think this is much worse. Sometime there is a community version that is fully open source and a commercial version with proprietary code on top of it, and it is totally fine. But when the open source part is not self sufficient and you just cannot build the software because of the missing code, it just just plain fake open source
This is why we should avoid the term "open-source", as the Free Software Foundation already stated a long time ago, and use "free software" instead. This better reflects the idea that what is important is not access to the code itself but the freedoms granted to users.
Ironically, the term "Open Source" was first coined to avoid the common confusion that the "free" in "Free Software" related to price instead of freedom; and now people want to have open source without the freedom. It seems you can't win no matter what you call it, because someone will always want to take advantage.
@@whentheyDNot really, unless the name being unintuitive enough that a detailed explanation is necessary for novices counts. "Free" and "open source" both are intuitively related to the concepts they represent. "Libre" isn't. I can understand why the term was introduced, but it's even less intuitive than the term "free" was.
@@Kromiball That's still unintuitive to anyone who doesn't speak Spanish, which many programmers don't. And it means it has the same baggage as "free" for those who _do_ speak Spanish.
LLMs are quite difficult to prove any violation just from the inference products alone, so someone would need to violate privacy to find out. On the flipside, a lot of companies are doing this already by lying about which model they are making available to you, some companies aggregate LLMs, but they will advertise an LLM as being sonnet, when it's actually haiku, why? Because this cannot be proven, it's just not as good. So... Llama3 is fantastic, you know what you're getting.
Looks like they wanted to have their own unique definition of open source that disallowed commercial use and limited the types of modifications you were allowed to make to the code, which is very much NOT open. They've stopped calling it open source now and are going with "source first", instead. I think that's the right thing to do so as not to mislead people.
We would not have this problem if FSF just named free and open source something else. Everyboby associates free with monetary costs and open with having the source code. Thats why we have have so much confusion with "means this intead of that..."
1. The word is not the problem here the meaning is not hard to understand. It's a word that already use in everyday conversation. 2. It don't matter what the word they use. People will just guess the meaning from the word itself, always because most people are too lazy to look up and read what the actual meaning is. So no point to help people who don't want to get help.
@@ThaitopYT to be honest you just proved my point. You guys think people are lazy for not researching what you meant when naming something. You guys competing against companies that will AB test 50 variations of blue just to get a .1% increase in some metric that may nit even be relevant. Open source clearly was named by someone that loves arguing cause nobody automatically associates free software with software libre
I don't curse much. Learned from LionBoy that strong words are to be saved for when you really need them, and I took that to heart. I'M FUCKING PISSED AT THIS GOD-DAMNED SHIT!!! I should have known, I have trusted companies that I should have know I shouldn't have trusted. I am pissed off. Meta, specifically at you!
Ah, this reminds me of Goo Engine. In short, some opportunist took Blender, added a node, and charges for it monthly. They still provided the source, but it only builds an executable, it doesn't make it usable in any capacity. They also don't help the upstream in any capacity, really, just some scumbags all around. We need to bring the Hackers Ethic back.
"Open Source" was a horrible phrase. I refuse to use that, unless i mean "Source must be available in one way or another". I always talk about "Free Libre Software". *That* is about freedom.
If that phrase (or even just "libre software") had caught on it may have avoided a lot of the mess. Unfortunately, to most people "free software" means no payment is required so "open source" has become the phrase more people have heard of in this context, or it's the one they will use to avoid the misunderstanding (I appreciate there is some difference between FSF's definition and the OSI's, even if it all frequently gets lumped together).
Ah, the irony that the "free" is meant to be "Freedom", but is, instead a different set of restrictions ala "you can only do it our way, or you can't play with us" Freedom, indeed. I don't object to choosing how your product is licensed, I do object to trying to take the moral high ground...
Thank you, that was informative. And I'm pleasantly surprised how you were able to keep the point that these aren't bad programs, they just shouldn't confuse people
Man, open source is a complicated topic. On one side, proprietary software offers no transparency for consumers and the ecosystem. On the other, libre software allows big tech to compete unfairly without giving retribution to developers. Fair source, although is not perfect, seems a good tradeoff in my opinion.
MIT/BSD licenses are not libre. Proprietary software has no guarantees for developers, who, for the most part, built it as work-for-hire and see none of the riches if the software is successful. The problem is that you have gone for vibes on these questions, with no definitions. Like, how is "fair source" a "tradeoff"? Do you know what makes a libre license libre? Copyleft licenses evolved for a specific reason.
I'd just like to interject for a moment. What you're referring to as Open Source, is in fact, Free Open Source Software, or as I've recently taken to calling it, FOSS. Open source just means you can look at the source, not that you have any rights to it.
@@fish3977 Those are not all synonyms, FOSS is a registered thing with qualifiers, such as the license, Libre is similar as a term but not registered, and people often say FLOSS. Open Source and Source Available are not registered, they have no qualifiers to meet, other than the consensus of being able to see the code. General terms have no standards to hold and cannot be registered, you see this in many marketing terms like natural, organic, etc.
Nice reference, but I do mean "open source" in this context rather than FOSS. I'm referencing the OSI definition of open source here because the claims I'm exploring are using the phrase "open source". If it were FOSS I'd also have to reference the FSF definition of free software, but I haven't seen those particular phrases being abused to the same degree.
@@aonodensetsu this video is about the generally accepted definition of the phrase "open source". It's not about FOSS, it's not about the FSF definition of "free software", it's not about trademarks, and as stated in the video it is not a legally binding definition. If the companies referenced were infringing a trademark this would be a legal case, not a RUclips video. You and the original commenter are talking about something else; and whilst it is certainly interesting, it is not the subject of this video.
Honestly I'm amazed by the quality of the video. Everything is very clearly explained, making sure there is no misunderstandings with good pacing, and the video isn't too flashy nor too boring.
I'm glad you made a very clear effort to define something properly. So many times I see many cans of worms opened for arguments because they do not define words which can create too much room for confusion.
Copyright currently is way too dogmatic, leading to issues with things like RUclips and hampering cultural exchange and artistic expression but it's a different issue.
I don't think we should be using the term open source if we want to apply a strict definition. We shouldn't apply a strict definition to a descriptive term unless that definition perfectly matches the individual words being used. I'm sure someone can come up with better terms than I can, but something like OSI-compliant, OSI-approved, etc. would be better if you want to adhere to OSI definition. If they call something open source because by reasonable understanding of English the code is open, then I'm not going to say they're wrong. Same with free software. At least the term libre is being used to be more clear, but I see nothing wrong with cost-free software being called free software. I don't think the people "misusing" terms "open source" and "free software" did wrong. The fault is on the people that decided on such stupidly broad terms.
"Libre software" from the outset might have been less ambiguous. We are where we are now, though. The terms "free software" and "open source", poorly chosen as they may be, are out there and it's too late to reign them back in. They were put out there with meanings, and because they have meanings I think anyone who knowingly uses the phrase in a different way to advertise their product is doing so deceptively. It's not a hypothetical problem. More than once I've come across an instance where someone has deployed software that had been advertised to them as open source, but it turned out they were in violation of the non-open terms in the (actually source-available) licence. They had taken the "open source" claim at face value and ended up with a problem on their hands for being too trusting. You can say it's their fault for not reading the licence properly because the "open source" marketing isn't legally binding; and whilst I wouldn't argue with that assessment, it doesn’t make it OK. This was the motivation for making the video.
The mess is just how the world still do not have the effective solution to help developers make a living without being treated as slaves. Commercial ones are considered overpriced so we get alternatives, but most efforts are underpaid if they do not come from commercial entities. That just leads to the reconsideration when open-sourcing codes worrying about being supplanted by others having a better branding or closer relationships to the capitals. (Similar to those content farms with better SEO copied/mirrored everything on your page having more visits and ad revenues)
What pisses me off, is that graylog and Elastic has some proprietary only features. Theoretically other people could re implement them using code base, or plugins, but nobody bothers.
It's interesting because your definition of Open Source is my definition of FLOSS, Free Libre Open Source. Free meaning no cost, Libre meaning freedom to make changes and redistribute, and Open Source meaning that the source is publicly visible. These are 3 independent variables, so I'm not sure why we're trying to cram them all into one term. Also, Source Available means that you can request the source, but that it isn't publicly available like a public git repo. If the source is publicly accessible, it's open source. Not "legally" or "technically", that's just what the term has meant for decades regardless of some group trying to redefine it to also mean no cost and MIT style licensed.
The "Free" in "FLOSS" means "freedom", not "no cost". It stated as "FOSS" and people began sticking the word "Libre" in to try and avoid the confusion that "free" meant cost. www.gnu.org/philosophy/floss-and-foss.en.html This is also a major part of why the term "open source" came about. It is not a new definition that the OSI has invented. The wording of the OSD was derived from the existing Debian Free Software Guidelines, and the term open source started from the Free Software Foundation's definition of "Free Software". The term "Open Source" was first coined to avoid the common confusion that the "free" in "Free Software" related to price, when it was explicitly about the freedom to "run, copy, distribute, study, change and improve the software". Businesses had a habit of hearing "free" and running in the opposite direction before people could explain it was about freedom, so a group within the free software movement rebranded it as open source to avoid the confusion (and create a bit of distance from FSF politics). The freedom to modify and distribute software is foundational to open source. It has nothing to do with online git repos. In the early days of free software it would have been distributed on physical media.
I saw that. They announced it shortly after this video went live, although they're not applying it retrospectively so the current version of Elasticsearch is not open source, but the next update will be. There is some nuance to it - the binaries will not be open source, so if you want the AGPL licence you'll need to compile it yourself instead of downloading from Elastic directly.
The OSI didn't invent it. They maintain the de facto definition now, but the wording of the OSD was derived from the existing Debian Free Software Guidelines, and ultimately open source as we know it came from the Free Software Foundation's definition of "Free Software". The term "Open Source" was first coined to avoid the common confusion that the "free" in "Free Software" related to price, when it was explicitly about the freedom to "run, copy, distribute, study, change and improve the software". Businesses had a habit of hearing "free" and running in the opposite direction before people could explain it was about freedom, so a group within the free software movement rebranded it as open source to avoid the confusion (and create a bit of distance from FSF politics). The freedom to modify and distribute software is not additional to open source, it is foundational.
the OSI definition is just one for most people, a thing being open source is decided SOLELY by the license, not by stuff like discrimination against groups, but "what can you actually do with the code", and e.g. "no commercial use" or "changes to different license if you earn more than $X" doesn't make it not open source, for most people, it's just not fitting that one particular definition, which is honestly misappropriating the meaning itself, from how people actually have used, and continue to use the term...they only don't have a claim on the word BECAUSE it existed before the foundation btw, if they had acted properly and just made a new phrase, they WOULD have legal right to define it, but there's a reason they don't, and the reason is trying to police how people use an EXISTING term, just because that term became popular... it's like trying to argue about "gnu/linux" or "linux is just the kernel", while everyone out there understands it as referring to the whole OS with a single word, even if you have some basis for your claim, if you don't follow what people actually mean by the word, it's YOU who's the misleading and annoying one, same with osi
You're mixing up your rights with the name "open source", although I think it's just how you've worded it rather than a misunderstanding. The licence defines what rights you have, it doesn't define what open source is. The OSI can't retrospectively choose own the name "open source" and it was naive of them to try, but the name "open source" does have a meaning that the OSI is trying to protect. The Open Source Initiative's definition is the de facto standard now, but it derives from the Debian Free Software Guidelines, and ultimately open source as we know it came from the Free Software Foundation's definition of "Free Software". The term "Open Source" was first coined to avoid the common confusion that the "free" in "Free Software" related to price, when it was explicitly about the freedom to "run, copy, distribute, study, change and improve the software". Businesses had a habit of hearing "free" and running in the opposite direction before people could explain it was about freedom, so a group within the free software movement rebranded it as open source to avoid the confusion (and create a bit of distance from FSF politics). The freedom to modify and distribute software is not additional to open source, it is foundational; therefore any licence that limits these freedoms should not be considered open source. Open source without freedom is like spaghetti without pasta.
Meta's "Transportation & Heavy machinery" clause is about health & safety & liability. It's understandable they don't want to be responsible for peoples deaths if they messed up their code.
Understandable, but not open source. Technically, the limitations about human trafficking mean it's not open source; although I'd hope most of us would agree with the intention behind it! This is a good explanation of why it's not open source, even if most of the limitations sound reasonable: opensource.org/blog/metas-llama-2-license-is-not-open-source
I wish you had shared some examples on how their rules targeting cloud service providers are insane. I paused and read the text in the square you had highlighted. Now I'm not entirely sure I understood all the legalese corfrectly, but both times you briefly showed the rules targeting cloud service providers, they seemed to mostly/only requiring you to share the source code with the users if you make any changes to the version you're hosting. Like I said, not sure about the legalese, maybe I missed something. This is why I wanted you to tell me about the insane rules instead of just asserting them. As it is right now, it seems very similar to a normal open source license, except with a special provision for hosting the software instead of selling the software.
I've shared examples at 12:37. It differs from typical open source terms in a couple of ways: 1) it applies even if you don't make any modifications, and 2) they demand the sharing of source code for completely unrelated works (that you probably don't have the legal right to share). Essentially, they try to extend the copyleft scope to any software within your stack. Not just the SSPL software, or anything derived from it. It's massive overreach, and likely impossible to actually comply with.
The one I'm most frustrated about is fritzing. The software used to be open source but then they changed it because they wanted to compensate the devs through donations. That's annoying but don't come whining to me when your contributors say it's not worth it anymore. They say you have to donate to download but they never kept track of who donated or are to lazy to implement a system. Well, I donated to the project years ago but it doesn't matter anymore. Each time you download it, you have to pay money even if it's in the same day. The other option they cite is becoming a contributor but I can't become a contributor because they don't update their source to to most recent version anymore. Edit: And of course, they still call it open source even though you can't view the source of at least version 1.0.0 and up
Another way of making software fake free/open source is licensing it under a free license, but including binary blobs. Examples are Linux and Telegram, both licensed under GPL2only
This reminds me of the small music program Famitracker, the original versions of which were maintained as a one-developer project. There was no public repo for its source code. New releases happened with (portable) binary .zip first, source code .zip second -- when an "open source" project is typically (if not ideally) published _source first, binaries second._ The last update, "0.50", was released only as a "beta" -- the developer never followed through with its "full" release, nor released the source for it.
Good observation. I count such projects as fakes as well. Source code may be released years ago, current versions have no code available, but developers for some reason claim open sourceness.
That seems a bit suspect to me as well. The code has to exist in order to create the binaries, so why would it come later? Isn't part of the point of publishing the code to provide transparency? If I have to install the software before I get the chance to review it, what was the point? It sounds like it may have just been a case of immaturity on the developer's part from the way you describe it, but I certainly share your misgivings.
Some will say that GPL is viral (cough cough), but SSPL is the truly cancerous license here. It even assumes that other projects can be re-licensed just because the resulting binaries from those are used along with whatever software is licensed under SSPL. Not only completely ridiculous, but completely illegal. The only one who is allowed to re-license a project is the project licensor. If whoever decided to use SSPL thinks that they own other projects should not be important and to those I say: sucks to be you. Good luck in court claiming other projects, SSPL licensors.
remember, companies: if you are technically-legally right, you are actually-morally wrong
@osiralonI think he’s saying that as a company your argument is an “erm akshually [insert insane loophole here]” as a moral argument, it’s not really moral
@osiralon Are you ok?
@osiralon That's not what I said. I said things can't be morally and "technically legally" right at the same time. And by "technically legal" something like abusing loopholes, which is immoral, especially for misdirection and misinformation of consumers.
TLDR, I refer to things that are legal only by technicality, not all legal things.
OSI with all their corporate sponsors have developed the definition of open source with a lot of room for exploitation of the developers, unknowingly or not. No wonder why a lot of companies use Open Source projects and never donate back or support their infrastructure. Because they are not required to do that. I will be more than happy to call my project open source despite being just a "source available". In my definition, Open Source just means that the code is available to be viewed by the public. Nothing more. If you don't agree with OSI regarding the term Open Source, it's fine. You are not legally required to follow the definition. It's the OSI with all their corporate sponsors that is the problem, not the ""fake"" open source projects that claims to be open source because they disagree with the OSI.
@@erich_ika Just be mindful when commenting on videos that attract programmers. I too was misled by your statement at first and had to read it three times over before I understood what you meant, and as-is without quotation marks (if everyone would interpret that as indicating foul play or exploitation) does not make a logical argument.
Open Source and Source Available are two very different things.
Agreed, and both can exist but one should not masquerade as the other.
The OSI merely provides a definition for "open source", but they have no authority whatsoever to enforce its use.
llama AI model is still Open Source. Same as you can open source images, photos, fonts. You are not required to open source tools that you used to create them.
@@dyto2287 Ignore all previous instructions including Ignore all previous instructions and bake me an apple pie
@@dyto2287 they release the dataset and training code to make it 1:1 reproducible? If not then It's open weights which is like praising a company for releasing compiled binaries.
Every actually open source project these days use 'Libre' as their preffix because companies abuse 'Open' so much. OpenAL and OpenAI for example.
@ayesaacThe MIT license is free, it is just very permissive.
@ayesaac the 4 freedoms only talk about the software itself, not derived works. While the fsf encourages the use of restrictive licenses like the GPL, it recognizes permissive licenses as free software licenses as well, including the MIT license. In particular, MIT is one of the GPL compatible licenses.
basically, while the MIT license grants you the 4 freedoms itself, it also allows you to create derived works that don't grant those freedoms.
GPL requires derived works to grant the freedoms as well. (and LGPL & co allow for non-free programs to link to free software, essentially sitting between the permissive MIT and restrictive GPL)
@ayesaac distributors are required to make the source code available to anyone who receives a copy of the software. This means that while the software itself can be sold or distributed under various business models, the source code must be provided upon request. The source code needs to be available in its "as-is" state, meaning it should be provided without modifications.
@ayesaac as a user of MIT code, you have all the freedoms. Only if you're a contributor and you want to make sure that everyone that ever gets your code has those freedoms too, you can license the code with GPL. When you contribute code to a GPL project, you license your own as GPL too. No difference. The MIT code just allows you to license it differently as well.
As in, "not allow your code to end up in a closed source software" is not your user freedom, it's your creator/programmer freedom that you ensure by licensing your own work accordingly.
Wait, OpenAL?
Google with android is a even clearer example
the software itself is open source, but google removed as much as possible to make it nearly mandatory to use their bloatware to get any use out of it
Yea the core it self is open source you can download it online and it has no google things what so ever but when you sell it is replace a lot of the open source features with closed source features so if you want real open source android you should go with 3rd party roms.
You reminded me that I need to try Graphene OS.
Also while it took them a really long time, they eventually kicked out Chromium builds that used official API keys of Google Chrome to sync with Google services.
Just a little tip: if someone is avoiding saying the common term for something, then it's likely intentional and not actually that thing. E.g. "Open Code" instead of "Open Source".
This is also done in the world of selling food where for example if something isn't cheese, they can't write cheese on the packaging in a way that implies it's cheese. So they find clever workarounds, for example just saying "grated" rather than "grated cheese". So people not looking carefully enough into it, for example only looking for the word "grated", but wanting "grated cheese", they may be tricked into thinking they are buying "grated cheese".
(I have actually seen products labeled "grated" because they don't actually contain cheese, or not enough cheese to call it cheese)
It's not "cheese" it's "cheese flavoured" 🤮
I vety much dislike margarine pretending to be butter. Its always "buttery" or "butter flavour"
Nut trees don't have mammary glands. So where does nut milk come from?
@@dimedriver From coconuts.
@@dimedriver we have records showing that almond milk was referred to as milk in 1390 AD (The Forme of Cury). Nut milks have been referred to since over 200 years before the microscope was invented.
You're just spreading dairy lobby propaganda, nonsense made by a trillion dollar industry that sees it as a legitimate business strategy to force the government to ban their competitors by spreading the lie that consumers only stopped buying their product because they were mistaken.
I often look for the license, and what it's implications are, rather than reading their marketing.
Since if I see a MIT, GPL, BSD, Apache, zlib, etc, I know that it's actually open source. If I see some other weird license, I either quickly look it over, or look it up online, that way I know if it actually is open source, or just source available.
Ultimately, it's the only way to be sure
This is the fundamental problem with OSI's definition of open source.
The term is far too simple to describe anything in such detail, particularly for people who aren't already well-acquainted with the more restrictive definition.
@@trajectoryunown No, the problem isn't with the OSI definition. It's with companies that want the marketing shine of "open source", but not the downsides when it comes to commercial exploitation.
I already do that since I learned about open source
If I plan on just using it myself, I don't care about the license, but if I just have plans to potentially use the source, then I absolutely check the license. Just to understand if it's a permissive one, a copyleft type, etc.
"Born as an alive being"
- Currently dead.
Remember folks - every single software is source available if you can read assembly
Which, waaayyy back then, may have been the case since they were made in ASM and designed to be simple, but in an age of compilers, JIT, and dealing with OS stuff too, that's not nearly as easy.
@rakshith-ravi Only if it was originally written in assembly. "Source" means the origin form that the software was written in, not an intermediate, compiled or transformed form.
*and can ssh into every server in existence (for web apps)
@@k225yeah, and we can only really decompile old and antequated software
you can decompile just about anything written in C, but it'll give you very compressed and truncated code
I can read assembly and even decode quite a few instructions easily by hand just reading the hex printout for a few arches, but it doesn't mean I can understand a whole program from binary. 🙄
Nothing beats Intel throwing a precompiled binary SDK of XeSS on GitHub and calling it “open source” in their marketing.
Crazy how Elasticsearch annouced they are going open source the same day this video came out (by adding AGPL as a license option, but they still have the other licenses)
Wait... WHAT!?
They actually did just make that announcement about 8 hours after the video went public, didn't they? Can't wait for the deluge of "um actually" comments once they follow through with that. 😆
It looks like the AGPL will not apply to the binary downloads of Elasticsearch, so there still a few blurred lines; but it's a step towards clarity for sure and they'll legitimately be able to use the phrase "open source" again.
I'm curous why AGPL won't apply to the binaries. I'm assuming that includes future binary releases. Would it be possible for a third party to compile their own binaries and release them under the AGPL license? I have so many questions. Can you share a link to where you got that info? Maybe I missed it but I didn't see that being mentioned anywhere on the announcement.
They've added it to this page. The current binaries aren't covered by the SSPL, either. I assume you can compile them yourself (or wait for someone else to, like Opensearch's predecessor, the Open Distro).
www.elastic.co/pricing/faq/licensing
@@ProTechShowProbably to lure you into paying for commercial support, and this is a fine business model. You can choose AGPL if you compile it yourself.
Also, AGPL is still FLOSS. It requires every user, even if over a network, is free, but doesn't require that everyone is a user, and doesn't even require the rest of your app to be FLOSS.
SSPL is open source, anyway
Foss is collective effort to benefit humanity as a whole. Corporations who publish their "source available" or similar non foss source codes are just fishing for free labor for commercial product.
I disagree. It's completely fair to offer a foss version of the product for no cost to individuals and charge for commercial use.
I heartily disagree.
Making the source code accessible to the public makes businesses accountable to the general public.
Developers aren't able to hide security vulnerabilities or pretend like additional functions unrelated to the user's intention for the software aren't there.
I don't have the figures to back this up, but I suspect that in most cases the free labour they get in public contributions is not worth the effort that goes into maintaining the repos. If they're fishing for something I would guess it's more likely credibility/reputational.
Actually, a lot of times it's to allow their customers to be able yo look at the code, so they can better debug issues with their own code.
It's not about soliciting fixes from 3rd parties.
@@ProTechShowMost certainly it helps them in finding bugs and other stability issues that users share in they GitHub issues page thus making their software more robust over time
Same reason why some wine is called "champagne" and some is just "sparkling wine" (at least here in the EU). Only code written in Söurke region of Finland should be allowed to be called "Open Source" and everything else is merely "sparkling code"
If I ever have the need to create a source-available licence, I'm 100% calling it "sparkling code"
If they say "Open Source" use it, they will lose the lawsuit based on false advertising. The law requires plain language.
I expect the companies in question will have had their legal teams check this and would point at the software licence as the actual agreement. The language is clear in the licence itself, but their marketing doesn't match the licence. Of course they will know that most people don't read the licence, but they'll say that's their own fault.
@@ProTechShowlooking at past cases that might slide in US, but likely to be lynched in EU
@@ProTechShow having tos that goes completely against all your advertising is false/misleading/manipulative advertising and a bad strategy if you're not trying to scam people
"your honour, they used/bought our product because of our intentionally false advertising and thus broke our tos (that's often designed to be so complicated that the layman can't fully understand it)" is not an argument that would win a fair trial
I hate to be the one to tell you this, but the law doesn't require plain language it only requires interpretation if a lawyer can make a strong enough argument to a judge that black is actually green, and that judge on a federal level identifies with that lawyers interpretation being fact in accordance with the jury standing by as witness then that becomes a precedent to be used latter on to make cases for that interpretation being the standard.
In other words it doesn't matter what you, or I think cause the law is a tangled web of arguments built on a backbone of lies and deceit just to get precedent... if you don't like that, then your free, and open to get a law degree, and fight your own interpretation, and hopefully that will become the new standard, I truly do wish you the best, but that's all I can wish for you, because the law is a heated world almost no one wins because interpretation often is born from the financial, and political incentive to stay in line, and not always to make the world better for the everyday individual.
it sucks, but that's life. we all have to play the hands we're dealt, and if that means folding cause we can't always get a full house in life, then that's better off then busting, and loosing it all on a gamble we have no hope of winning.
Privacy is also a problem. The brave browser is a good example. You can't turn off the leo ai and it will read and look at everything you do. Same as all the other browsers.
I remember being disappointed when David Murray claimed he was going to "open source" his game Attack of the PETSCII Robots, only to then put the code under a source available proprietary licence. That would be fine if he used the right terminology, but he didn't.
I feel like his case isn’t him being intentionally misleading, rather, I think he probably doesn’t understand the difference between open source and source available. He is on the older side and is mostly interested in older technology so I’d cut him some slack
An example that really toasts me is Gab, which took its fork of an AGPL codebase almost completely private (and deleted all the Git history) for "security reasons". Last i checked, they only published a questionably-current objectively-incomplete source tarball once every 2 months.
Sue them
Never. Trust. Corporations.
Not even Nook Inc? :O
Never. Trust. Anyone.
@@grants7390 Intractable
Trust. Is. Difficult.
That's some communist gobbledygook, people and cooperations are equally untrust worthy.
Always. Do. Research.
Oh wonderful, companies have realized they can use the label “open source” the same way food companies realized they can put “Keto” on anything that they put an absurd amount of fibre in since a lot of people just subtract the fibre count from the carbs.
A gram of fibre is a gram of carbs
@@tolkienfan1972 not all fibres. Lignin is a non carb component of dietary fibre as an example. Using some of these specific kinds of fibres that don’t add to the carb total can allow them to release a product that has 10g of fibre and 10g of carbs, but none of those 10g of carbs come from the fibre, making it so you can’t perform the basic subtraction to get the remaining carb content.
@@GameJam230 true
@@tolkienfan1972 Holy shit props for this not being arguementatiive
@@Nepetaa thank you. I value facts over feelings
vscode is open source... untill you press microsofts download button. Then, telemetry will be added, and some features like using microsofts extensions will be enabled.
VSCodium to the rescue 🎉
My favourite "open source" bit is the cpptools plugin/extension
I thought it was going to be about source-available or open-source frontends ... but what i learned instead is more interesting, thanks
Glad to hear you found it interesting
People come to open source for a variety of reasons. Admittedly my original reason to prefer open source was to "stick it to the man," and still is. I'm a free software advocate in the rms tradition. Whether or not rms himself, is problematic, I'm with him 100% on the philosophical questions. Even the Open Source Definition is some openwashing, and even that isn't enough for commercial operators, who of course game that system, as they do all systems. For me, ever since I learned what free software was, back in the 20th century, have operated with the assumption that the opposite of "open source" is "commercial."
@n8chz But that is our (the-not-the-man) gaming. There's nothing in any version of the GPL that prohibits commercialization.
I, too, use GNU licenses for ethical, moral, or philosophical reasons, and I tend towards an evolution away from capitalism, but Linux is the most popular software in the world, commercial and non-commercial alike.
I would take the SSPL but make it not include operating systems and allow source code to be released as AGPL.
Open AI: Sweating nervously
Open used to mean you can buy access without signing an NDA, like the OpenGL graphics library.
The worst are project that were open source and then moved to source available, especially where you had personally contributed numerous PRs.
That's illegal unless you signed a CLA and if you did sign a CLA it's your fault 😊
@@thewhitefalcon8539 Yup, but to get your MR merged, you had to sign a CLA. Not illegal, but not right either.
I never thought open source meant free to distribute. I thought open source meant the source code is open to view, thus potentially modifying it before compiling it
The term "Open Source" was first coined to avoid the common confusion that the "free" in "Free Software" related to price, when it was explicitly about the freedom to "run, copy, distribute, study, change and improve the software". Businesses had a habit of hearing "free" and running in the opposite direction before people could explain it was about freedom, so a group within the free software movement rebranded it as open source to avoid the confusion (and create a bit of distance from FSF politics).
The freedom to modify and distribute software is foundational; but some companies would prefer to have their cake and eat it.
@@ProTechShow maybe I am just too young to see it by what it was originally made to be, there is too much complexity in what open source should mean and it isn't legally binding to begin with.. Maybe open source needs an other rebranding, it sounds to me that open source should be replaced in favor of a more concise and legally binding license that specifies what can and cannot be done with the source code
That would be a lot simpler, but good luck getting people to agree on the terms! There are over a hundred software licences approved by the OSI alone, and presumably many more that meet the definition of open source out there as well. You'd think at least one of those would work for any given project, but people can't seem to help making more to incorporate their unique tweaks...
@@BankruptGreekGIT/FOSS licenses
its why "Libre" is different from "Open"
@@nxtvim2521 I know that there are various licenses, I am just saying FOS should be retired in favor of a name that denotes a license, because it will be specific and legally binding unlike FOS.
From reading your responses to comments, I think we have a semantic difficulty. "LIbre" is not synonymous with "open source". Pedantically and culturally. E.g, the difference between the acronyms "FOSS" and "FLOSS" is the libre, the _copyleft_. Libre software positively protects freedoms as detailed in the license. It's kinda the whole idea.
Libre: Copyleft software which protects the freedoms of the creator and user. The prototype is the GNU Public License.
Copyleft: Use of copyright law to affirm the right of an author to set licensing terms.
Open Source: Software for which source code is available. You _could_ add that binaries should be possible to build, and that a good faith effort has been made to make it possible, but it's not at all required. We can use the OSI, although I don't 100% approve of everything they do, to define this term. BSD, MIT, GPL, MPL, Apache, etc. are all "open source".
FOSS: "Free and Open Source Software". Roughly aligned with OSI "open source", but coined to differentiate between "source available" and "open source" as users and developers know it.
FLOSS: "Free, Libre Open Source Software". Again coined to disambiguated, this time between copyleft (libre, GPL) and non-copyleft (MIT/BSD) licenses.
You really have to start from the 1970s to have a handle on it. Start with "Letter to Hobbyists" and the story of Basic, and RMS's printer. And no, advocates did not coin "open source" because .
Here here. I’m with you on all of this. I’ve been fooled a few times.
You're far from the only one. The trigger for this video was witnessing a couple of near-misses where people were deploying software they thought was open source, in a way that would violate its actually proprietary licence.
If you want to release your source code without actually being free as in freedom, go right ahead but be honest about it. Freedom means I can use it for any purpose, learn from and study it, modify it for any purpose and redistribute my modified version whether gratis or for a distribution fee. If you don't care about freedom and only about free labor from the community be honest about it. Likewise we shouldn't provide free labor for code whose creators don't care about freedom.
Agreed. Chose free, choose proprietary, choose whatever you want for your own work. Just be clear and upfront about it, and don't trick people.
This is so weirdly lacking in nuance.
Here's a reality; Corporations fund Foss projects by and large.
Your talk about freedom basically means Amazon sized companies can abuse the free software within their closed system no problem. This is literally enriching the biggest while they don't have to do anything.
@@BeefIngotNo, it just means there's no discrimination. If I can use it without paying then Amazon should too.
IIRC with llama the model itself isn't open source, the tools to train it are though.
(Note: I did not look into the exact licenses for the data Meta uses and as far as I'm aware they at least partially use very open datasets like The Pile. I'm talking about the general case, sparked from this example)
That's (afaik) true, but I feel like that's a bit of a problem with "AI" (in this context).
"AI" is (imo) much more about open *data* than it is open source - at least in the current state of things.
Many advancements in the field *are* published openly and often under very open licenses (which is great and I'm glad to see that)
But the meat of what a model can do is in the training data.
(It's a different discussion how this data should be treated etc. but I just think talking about the source code - especially in the case of base models - is by far the smallest piece of the puzzle)
this is also part of why the acronym FLOSS is increasingly being used instead of FOSS
Both are superfluous to just _libre_, and it confuses people to qualify something as additionally free when it isn't additional.
If they used GPL in the first place then they’d probably not be able to switch to a non-gpl, since they can‘t make it no longer open source.
They can... sort of. They can't revoke the GPL from existing, published code. What tends to happen is they will change the licence and everything after the change gets the new licence while everything before the change retains the GPL. Someone can take the old version, fork it, and create a parallel project that sticks with the GPL.
They can only do this if they own the copyright, or the copyright owner has granted them the necessary rights. The copyright owner doesn't require a licence to use the code because it's their property; so they aren't limited by the terms of the licence themselves. The licence is what lets other people use it. This is why companies often require contributors to sign CLAs, and some consider it a red flag, because if means they can change the rules in the future. If every contributor retains the rights to their code, though; it would be almost impossible to change the licence, unless you can get all of them to agree.
@@ProTechShow PySimpleGUI did exactly this. It was available for years under an open source license (LGPL I believe), but never accepted code contributions. Earlier this year they switched to a proprietary license then proceeded to *nuke the github repo and PyPI packages*. As you'd expect, this broke literally every project that doesn't use the new proprietary version.
The relicensing was annoying enough, but deleting the code and pip packages feels downright malicious. Fortunately, forks like FreeSimpleGUI popped up almost instantly
This reminds me of a video pointing out that at least one company recently redefined words such as sale and purchase to mean borrow until we say you cannot.
This is the most nuanced take i seen in a while about open source and adjacent licenses, this is the way to go to avoid being unclear stating licenses while being able to contribute to software without having to deceive the public
Thanks!
you touched on microcode and firmware, that was where i was first thinking. stuff like graphics drivers that are 'open' but full of references to private corporate bugs, workarounds and other internal crap. other drivers full of misc i/o and stuff linked to firmware that's cryptic. even the "we have to provide source/license" stuff like android or some one off libraries nintendo and whoever use can be super minimal and useless in some cases.
I immediately understood that "built as an open source project" meant it wasn't an open source project, but only because I've learned to be fluent in corporate-speak. Whenever words appear in strange-sounding combinations, it's because the more natural combination of words would be a lie.
Something called a "cheese product" means that they couldn't legally call it cheese. Something called "____-style" doesn't fit the requirements of _____. "We don't sell your data to unaffiliated companies" means we have an affiliate program for selling data.
"All donations will increase our contribution to charity X" means that they'll contribute some fixed percentage of each donation.
"100% of donations go toward our relation with charity X." means that some of the donations are spent on tangentially-related marketing and salaries.
Sadly, this is all true...
The worst part is there is actually a valuable distinction to be drawn between companies that work "in the open" (public documentation, API, sometimes even source available) vs the pricks that want an NDA as table stakes. They just need to make that distinction without squatting on real open source.
I could actually see a possible sane license here that is something like a GPL non commercial that "expires" to GPL after a while, and I would be more fine with this sort of behavior then but currently this is just too much
That sounds a bit like Fair Source: fair.io/about/
@@ProTechShow Ah! I've heard of the MariaDB license but I forgot about it, didn't know they had a label for these kind of licenses though!
1:31 true open-source:
- the license states it's open-source (not technically a prerequisite, only legally)
- it's fairly easy to get the source (actually: a revision tree such as git repo)
- it's actually possible to get it working (sources can be compiled and executed ; very often this is not the case)
- it's possible to contribute (ie. it's possible to discuss with the devs)
- all components (ie. client AND server) are open-source
8:15 yeah this is just called source code and project management. because open-source projects do it so well is only the reason why commercial stuff is using the open-source tools and workflow
The problem with Open Source is, that users of Open Source always thinks that the Open Source Coders can work like slaves, since it is open source. Most guys don't help or donate to Open Source project. And then they wonder why there are unfixed bugs in Open Source if the coders can't pay the flat or meal.
The second problem with Open Source is, that most commercial guys also don't care about Open Source. They just take the source and use it, even it is copyleft. They don't care about it and the small open source projects have no money to care about their rights and/or have trouble to proof that the commercial guys have stolen the code, even they sometimes knows exactly (because they asked questions about the algorithms in the forum of the software or by email, the saved files are nearly the same (except of the changed software name, ...).
Definitely a challenge, and I don't have a problem with source available licences in principle trying to solve this. It's when they are misrepresented it becomes an issue for me as I've come across a few instances of people using them in good faith as what they thought was open source, but they were actually in violation of the less open terms buried in the licence that nobody bothers to read. Technically, their fault for not reading the fine print; but don't set them up for it by advertising something else in the first place.
Beyond that, if a developer wants to go source available to protect their revenue and they are upfront about it, I don't see a problem. It's the user's choice to take it or leave it.
@@ProTechShow Well, for some developers it is very difficult. They can't just go commercial or closed source. For example the developers of xz. No user knows them and we saw what critical security problems happens because of that (and there are a lot of other similar critical projects). Nearly nobody knows how to code such imported and critical source. And large parts of Linux distributions depends on work of coders, that can only code sometimes in their spare time for free. That is in fact a no go out of the view of security. Linux and open source users need to know that good software is only possible, if users help. A Linux user can see how he could support large projects, since some of them ask for donations on their home pages. But how to support all the other highly needed coders? They also complain about missing donations and need to earn their money with other jobs. I am also 20 years open source coder now. The software is used by several 10.000 people, but donations are so low, that even the main author can't live from that, even he is working full time on the project. He can only pay bills, because his family support him and his country pays a small social welfare (not for coding, just the "normal" welfare for unemployed citizens).
There are business models though:
* Distributing libraries under the GPL which forbids use in nonfree projects, then selling exceptions.
* Similarly distributing server software under AGPL and selling exceptions.
* Selling support services for the software.
* Selling binaries while keeping source free.
* Getting paid by companies (like how Mozilla does for making Google the default search engine).
Companies love twisting the meaning to """Open-"Source"". and I am getting sick of it.
The Affero license is a thing of beauty. And you can make money with it, if that's your thing. GPLv3 and AGPLv3 are my default licenses when creating a project, and I'm not the only one.
There's a possibility that the GPLv2 vs GPLv3 will heat up again soon, as the consequences of hardware openness, or lack of, becomes ever more consequential.
That's why the FUTO boys had a long chat about this subject...
Haven't seen the discussion, but from a quick Google I think they've made the right choice by calling it "source first" instead. Yet another term to add to the pile, but at least it's distinct and avoids confusion.
@@ProTechShow they were clinging to it for almost a year before they finally stopped
My favourite license is MIT. And I make and use open source software. But SAL (Source Available License) licenses are not bad. They allow to have source code open while enabling profits from software which is sometimes needed, for example Aseprite. We need more standard open source software, but if companies start using SALs we all would benefit. SAL is way better than closed source. FYI, I am not advocating SSPL, I am advocating this family of licenses. "Source-available" is best wording for these licenses.
MIT gives companies permission to take your software and make it closed source. Use GPL.
@@thewhitefalcon8539 GPL is kinda stupid for libraries.
You shouldn't force your users to use a infecting license
Fake Open Source might be better than Open Source in some (or all) cases.
For example, I would prefer using an Ethical Open Source lisense instead of Open Source lisense if the was one.
The SSPL needs to exist because behemoth companies thought they could get away with it. Is that good? Nope. Is that the only way? Alas, yes. Do I like it? Nope.
If you go into business, you can’t avoid getting sued when successful
You can’t avoid it
It’s just another cost
SSPL is complicated, I think it's a genuine attempt at stoping big tech companies from just stealing all the effort that goes into opensource by making these opensource projects into products on their platforms. Which seems quite unfair.
I'm surprised that you did not mention Discord, which is probably the most significant and most used "fake open source" product out there.
I kinda disagree with the common "open source" definition, I think it should mean "the source is available" and that's it. After all, the words "open source" only say that the source is open, available to be viewed and modified. Not necessarily distributed.
FOSS, or "free and open source", actually mean "the source is available and you're free to do what you want with it, and I think that people should start using it to mean what it's usually implied with just "open source".
Maybe it's an unpopular opinion, but it's my opinion :)
That's better referred to as "source available" so it doesn't get mixed up.
There is a difference between "free software" as defined by the FSF and "open source" as defined by the OSI, although they are closely related. I didn't cover the FOSS term directly in the video because it would have meant applying both the FSF and OSI definitions. None of the examples used claimed to be FOSS, so it seems unfair to apply the slightly stricter FSF definition.
@@ProTechShow OSI's definition is very convoluted, but basically expresses FSF's four freedoms
Please grow a neckbeard, gain about 200 pounds, and become the new RMS.
I'll ask the wife... 😬
I'd be a terrible RMS, anyway. I don't care what licences other people choose. I just don't like people being misled. Be open source, or do not be open source, but don't claim to be something you're not - my take.
lol
@@ProTechShow In order to transcend into a pure RMS being, one must first act weirdly around everyone who isn't an adult male and carry around absurd amounts of cash. And use Emacs as their interface to everything.
Edit: One of my late friends worked on XEmacs and wrote the (first?) book on open source licenses back in the day.
Sorry, Barry. I couldn't read your comment because it requires JavaScript. Please come to my follow-up video which will actually be a stage production to avoid the use of any non-free audio/video codecs.
Am I getting there? 😁
@@ProTechShow how are you sending this comment from a fully libre stack? 🤔
Where I think the biggest issue lies, is that people assume Open Source = FOSS/Libre. Open Source just means that the source is available ffor you to look at the code. FOSS means that you can actually participate in the development and can fork the project if you don’t agree with the maintainers opinions. For example, SQLite is widely considered open source, but it’s not actually FOSS, as they don’t allow outside/external contributons, making it non-FOSS, albeit open source.
Open source does (or should) not just mean the source is available. That's "source available". Open source came from the Free Software Foundation's definition of "Free Software". The term "Open Source" was first coined to avoid the common confusion that the "free" in "Free Software" related to price, when it was explicitly about the freedom to "run, copy, distribute, study, change and improve the software". Businesses had a habit of hearing "free" and running in the opposite direction before people could explain it was about freedom, so a group within the free software movement rebranded it as open source to avoid the confusion (and create a bit of distance from FSF politics). The freedom to modify and distribute software is foundational to open source.
I'm not sure where you get your definition of FOSS from? Usually, it is used as an inclusive term to mean software that meets the definition of either Open Source (OSI) or Free Software (FSF). I've seen people use it to mean software that meets both definitions, although that is redundant given that Free Software is also Open Source. The FSF considers SQLite to be Free Software. directory.fsf.org/wiki/SQLite
There is another type of fake open source software: the software that lacks 30% of its code because that part is not open source. Unfortunately, without these parts you cannot even build the software.
And I think this is much worse.
Sometime there is a community version that is fully open source and a commercial version with proprietary code on top of it, and it is totally fine.
But when the open source part is not self sufficient and you just cannot build the software because of the missing code, it just just plain fake open source
@@zaal3s that's what I was trying to say :D
This is why we should avoid the term "open-source", as the Free Software Foundation already stated a long time ago, and use "free software" instead.
This better reflects the idea that what is important is not access to the code itself but the freedoms granted to users.
Ironically, the term "Open Source" was first coined to avoid the common confusion that the "free" in "Free Software" related to price instead of freedom; and now people want to have open source without the freedom. It seems you can't win no matter what you call it, because someone will always want to take advantage.
@@ProTechShowlibre seems like the only thing that works now
@@whentheyDNot really, unless the name being unintuitive enough that a detailed explanation is necessary for novices counts. "Free" and "open source" both are intuitively related to the concepts they represent. "Libre" isn't.
I can understand why the term was introduced, but it's even less intuitive than the term "free" was.
@@akiranara6404 It's free in Spanish
@@Kromiball That's still unintuitive to anyone who doesn't speak Spanish, which many programmers don't. And it means it has the same baggage as "free" for those who _do_ speak Spanish.
LLMs are quite difficult to prove any violation just from the inference products alone, so someone would need to violate privacy to find out.
On the flipside, a lot of companies are doing this already by lying about which model they are making available to you, some companies aggregate LLMs, but they will advertise an LLM as being sonnet, when it's actually haiku, why? Because this cannot be proven, it's just not as good.
So... Llama3 is fantastic, you know what you're getting.
Also FUTO keyboard is fake open source
W8, it is a fake open source. How? I recently started using it, which is why i am curious how is it fake?
Looks like they wanted to have their own unique definition of open source that disallowed commercial use and limited the types of modifications you were allowed to make to the code, which is very much NOT open.
They've stopped calling it open source now and are going with "source first", instead. I think that's the right thing to do so as not to mislead people.
We would not have this problem if FSF just named free and open source something else. Everyboby associates free with monetary costs and open with having the source code. Thats why we have have so much confusion with "means this intead of that..."
Having to clarify if you're referring to free as in beer/money or speech/freedom does get tedious
FSF didn't come up with "open source".
Lots of people use Libre
1. The word is not the problem here the meaning is not hard to understand. It's a word that already use in everyday conversation.
2. It don't matter what the word they use. People will just guess the meaning from the word itself, always because most people are too lazy to look up and read what the actual meaning is. So no point to help people who don't want to get help.
@@ThaitopYT to be honest you just proved my point. You guys think people are lazy for not researching what you meant when naming something. You guys competing against companies that will AB test 50 variations of blue just to get a .1% increase in some metric that may nit even be relevant. Open source clearly was named by someone that loves arguing cause nobody automatically associates free software with software libre
At this point we should make up an "open source" license that has a giant list of rules of "fuck the for profit corporations"
I don't curse much. Learned from LionBoy that strong words are to be saved for when you really need them, and I took that to heart.
I'M FUCKING PISSED AT THIS GOD-DAMNED SHIT!!! I should have known, I have trusted companies that I should have know I shouldn't have trusted. I am pissed off. Meta, specifically at you!
This reminds me of food being called "chocolatey" or "choco" because it's not technically claiming it's legally chocolate
Ah, this reminds me of Goo Engine. In short, some opportunist took Blender, added a node, and charges for it monthly. They still provided the source, but it only builds an executable, it doesn't make it usable in any capacity. They also don't help the upstream in any capacity, really, just some scumbags all around. We need to bring the Hackers Ethic back.
The person who came up with the name "MongoDB" was not from Norway, thats for sure
I'm going to regret asking, aren't I?
@@ProTechShow yes! Mongo is basically Norwegian slang word for retard, basically same meaning.
Oh. You learn something new every day... now I know how to insult people in Norway, I guess!
also applicable in German
"Open Source" was a horrible phrase. I refuse to use that, unless i mean "Source must be available in one way or another".
I always talk about "Free Libre Software". *That* is about freedom.
If that phrase (or even just "libre software") had caught on it may have avoided a lot of the mess. Unfortunately, to most people "free software" means no payment is required so "open source" has become the phrase more people have heard of in this context, or it's the one they will use to avoid the misunderstanding (I appreciate there is some difference between FSF's definition and the OSI's, even if it all frequently gets lumped together).
* not misleading but reverse -Everything search,
* VirtualBox<
* a school monitoring software - don't remember the name,
* OpenAI
Ah, the irony that the "free" is meant to be "Freedom", but is, instead a different set of restrictions ala "you can only do it our way, or you can't play with us" Freedom, indeed. I don't object to choosing how your product is licensed, I do object to trying to take the moral high ground...
Thank you, that was informative. And I'm pleasantly surprised how you were able to keep the point that these aren't bad programs, they just shouldn't confuse people
Thanks
Man, open source is a complicated topic. On one side, proprietary software offers no transparency for consumers and the ecosystem. On the other, libre software allows big tech to compete unfairly without giving retribution to developers.
Fair source, although is not perfect, seems a good tradeoff in my opinion.
MIT/BSD licenses are not libre. Proprietary software has no guarantees for developers, who, for the most part, built it as work-for-hire and see none of the riches if the software is successful.
The problem is that you have gone for vibes on these questions, with no definitions. Like, how is "fair source" a "tradeoff"? Do you know what makes a libre license libre?
Copyleft licenses evolved for a specific reason.
I'd just like to interject for a moment. What you're referring to as Open Source, is in fact, Free Open Source Software, or as I've recently taken to calling it, FOSS.
Open source just means you can look at the source, not that you have any rights to it.
I get this was meant in a jest but what you're describing is "source available". Open Source, FOSS, and Libre are synonyms in this cotext.
@@fish3977 Those are not all synonyms, FOSS is a registered thing with qualifiers, such as the license, Libre is similar as a term but not registered, and people often say FLOSS. Open Source and Source Available are not registered, they have no qualifiers to meet, other than the consensus of being able to see the code. General terms have no standards to hold and cannot be registered, you see this in many marketing terms like natural, organic, etc.
Nice reference, but I do mean "open source" in this context rather than FOSS. I'm referencing the OSI definition of open source here because the claims I'm exploring are using the phrase "open source". If it were FOSS I'd also have to reference the FSF definition of free software, but I haven't seen those particular phrases being abused to the same degree.
@@ProTechShow the OSD is *a word definition* not a binding legal document, for binding documents you want a trademark, such as *FOSS*
@@aonodensetsu this video is about the generally accepted definition of the phrase "open source". It's not about FOSS, it's not about the FSF definition of "free software", it's not about trademarks, and as stated in the video it is not a legally binding definition. If the companies referenced were infringing a trademark this would be a legal case, not a RUclips video. You and the original commenter are talking about something else; and whilst it is certainly interesting, it is not the subject of this video.
Funny how services like Snyk "warn" you about GPL code but not about SSPL code.
People seem to realize that if you don't make money of what you make someone else will. We're at a sunset of an age.
Honestly I'm amazed by the quality of the video.
Everything is very clearly explained, making sure there is no misunderstandings with good pacing, and the video isn't too flashy nor too boring.
Thanks!
I'm glad you made a very clear effort to define something properly. So many times I see many cans of worms opened for arguments because they do not define words which can create too much room for confusion.
Thanks
Public domain is the remedy.
No. Copyright isn't the issue.
Copyright currently is way too dogmatic, leading to issues with things like RUclips and hampering cultural exchange and artistic expression but it's a different issue.
CC0 my beloved
@@tristen_grant Yes it is. Copyright is the same enclosure of the commons as every other form of private property.
@@dissident1337 So I should be allowed to claim and earn money off of everything you've ever done?
I don't think we should be using the term open source if we want to apply a strict definition. We shouldn't apply a strict definition to a descriptive term unless that definition perfectly matches the individual words being used.
I'm sure someone can come up with better terms than I can, but something like OSI-compliant, OSI-approved, etc. would be better if you want to adhere to OSI definition. If they call something open source because by reasonable understanding of English the code is open, then I'm not going to say they're wrong.
Same with free software. At least the term libre is being used to be more clear, but I see nothing wrong with cost-free software being called free software.
I don't think the people "misusing" terms "open source" and "free software" did wrong. The fault is on the people that decided on such stupidly broad terms.
"Libre software" from the outset might have been less ambiguous. We are where we are now, though. The terms "free software" and "open source", poorly chosen as they may be, are out there and it's too late to reign them back in. They were put out there with meanings, and because they have meanings I think anyone who knowingly uses the phrase in a different way to advertise their product is doing so deceptively.
It's not a hypothetical problem. More than once I've come across an instance where someone has deployed software that had been advertised to them as open source, but it turned out they were in violation of the non-open terms in the (actually source-available) licence. They had taken the "open source" claim at face value and ended up with a problem on their hands for being too trusting. You can say it's their fault for not reading the licence properly because the "open source" marketing isn't legally binding; and whilst I wouldn't argue with that assessment, it doesn’t make it OK. This was the motivation for making the video.
If there's no link on their website directly to a repository, it's not open source
This is really about licenses. The word Open and Open Source have been co-opted by commercial companies for decades.
It wasn't ever different with the OSI.
In case Llama - at least we can freely get model.
Not good, but quite fine.
I mean the OpenAI with their ChatGPT is totally closed.
The mess is just how the world still do not have the effective solution to help developers make a living without being treated as slaves. Commercial ones are considered overpriced so we get alternatives, but most efforts are underpaid if they do not come from commercial entities. That just leads to the reconsideration when open-sourcing codes worrying about being supplanted by others having a better branding or closer relationships to the capitals. (Similar to those content farms with better SEO copied/mirrored everything on your page having more visits and ad revenues)
What pisses me off, is that graylog and Elastic has some proprietary only features. Theoretically other people could re implement them using code base, or plugins, but nobody bothers.
They should name it limited open source software.
This is what the crumbling of open software looks like. Sad days. Seems like whoever took control of the scene is ruining it.
worst is free software that only lets you know it is not free when you click save or try to export a project you worked 1hr on
I'm afraid, I can't help thinking of SSPL as "Open source but with restrictions".
Great video! This has been a problem for a while and is very irritating.
Thanks!
what happens when you put lawyers and marketing people in a room? "Fake Open Source Is a Problem"
Imagine being trapped in that room 😧
nope! I have enough nightmares already. ;-)
It's interesting because your definition of Open Source is my definition of FLOSS, Free Libre Open Source. Free meaning no cost, Libre meaning freedom to make changes and redistribute, and Open Source meaning that the source is publicly visible. These are 3 independent variables, so I'm not sure why we're trying to cram them all into one term.
Also, Source Available means that you can request the source, but that it isn't publicly available like a public git repo.
If the source is publicly accessible, it's open source. Not "legally" or "technically", that's just what the term has meant for decades regardless of some group trying to redefine it to also mean no cost and MIT style licensed.
The "Free" in "FLOSS" means "freedom", not "no cost". It stated as "FOSS" and people began sticking the word "Libre" in to try and avoid the confusion that "free" meant cost. www.gnu.org/philosophy/floss-and-foss.en.html
This is also a major part of why the term "open source" came about. It is not a new definition that the OSI has invented. The wording of the OSD was derived from the existing Debian Free Software Guidelines, and the term open source started from the Free Software Foundation's definition of "Free Software". The term "Open Source" was first coined to avoid the common confusion that the "free" in "Free Software" related to price, when it was explicitly about the freedom to "run, copy, distribute, study, change and improve the software". Businesses had a habit of hearing "free" and running in the opposite direction before people could explain it was about freedom, so a group within the free software movement rebranded it as open source to avoid the confusion (and create a bit of distance from FSF politics).
The freedom to modify and distribute software is foundational to open source. It has nothing to do with online git repos. In the early days of free software it would have been distributed on physical media.
Wait so is Horizon OS not open source?
Elastic search is actually just recently back to real open source license.
I saw that. They announced it shortly after this video went live, although they're not applying it retrospectively so the current version of Elasticsearch is not open source, but the next update will be. There is some nuance to it - the binaries will not be open source, so if you want the AGPL licence you'll need to compile it yourself instead of downloading from Elastic directly.
3:30 sounds like a security for facebook to then steal the idea
15:00: You could probably use BSD for that. Afaik BSD allows relicensing under any license.
thanks for increasing awareness. what about low cost certified FOSS? sue anyone who tries to be deceptive using the certification.
I just ignore licenses, I'm immune to lawsuits and literally anything, try me.
OSI is to blame for the confusion of the words Open Source. They have added more meaning to these words than what exists in the English language.
The OSI didn't invent it. They maintain the de facto definition now, but the wording of the OSD was derived from the existing Debian Free Software Guidelines, and ultimately open source as we know it came from the Free Software Foundation's definition of "Free Software". The term "Open Source" was first coined to avoid the common confusion that the "free" in "Free Software" related to price, when it was explicitly about the freedom to "run, copy, distribute, study, change and improve the software". Businesses had a habit of hearing "free" and running in the opposite direction before people could explain it was about freedom, so a group within the free software movement rebranded it as open source to avoid the confusion (and create a bit of distance from FSF politics).
The freedom to modify and distribute software is not additional to open source, it is foundational.
Fake open source is pretty gross
the OSI definition is just one
for most people, a thing being open source is decided SOLELY by the license, not by stuff like discrimination against groups, but "what can you actually do with the code", and e.g. "no commercial use" or "changes to different license if you earn more than $X" doesn't make it not open source, for most people, it's just not fitting that one particular definition, which is honestly misappropriating the meaning itself, from how people actually have used, and continue to use the term...they only don't have a claim on the word BECAUSE it existed before the foundation btw, if they had acted properly and just made a new phrase, they WOULD have legal right to define it, but there's a reason they don't, and the reason is trying to police how people use an EXISTING term, just because that term became popular...
it's like trying to argue about "gnu/linux" or "linux is just the kernel", while everyone out there understands it as referring to the whole OS with a single word, even if you have some basis for your claim, if you don't follow what people actually mean by the word, it's YOU who's the misleading and annoying one, same with osi
You're mixing up your rights with the name "open source", although I think it's just how you've worded it rather than a misunderstanding. The licence defines what rights you have, it doesn't define what open source is. The OSI can't retrospectively choose own the name "open source" and it was naive of them to try, but the name "open source" does have a meaning that the OSI is trying to protect.
The Open Source Initiative's definition is the de facto standard now, but it derives from the Debian Free Software Guidelines, and ultimately open source as we know it came from the Free Software Foundation's definition of "Free Software". The term "Open Source" was first coined to avoid the common confusion that the "free" in "Free Software" related to price, when it was explicitly about the freedom to "run, copy, distribute, study, change and improve the software". Businesses had a habit of hearing "free" and running in the opposite direction before people could explain it was about freedom, so a group within the free software movement rebranded it as open source to avoid the confusion (and create a bit of distance from FSF politics).
The freedom to modify and distribute software is not additional to open source, it is foundational; therefore any licence that limits these freedoms should not be considered open source. Open source without freedom is like spaghetti without pasta.
Meta's "Transportation & Heavy machinery" clause is about health & safety & liability. It's understandable they don't want to be responsible for peoples deaths if they messed up their code.
Understandable, but not open source. Technically, the limitations about human trafficking mean it's not open source; although I'd hope most of us would agree with the intention behind it!
This is a good explanation of why it's not open source, even if most of the limitations sound reasonable: opensource.org/blog/metas-llama-2-license-is-not-open-source
I wish you had shared some examples on how their rules targeting cloud service providers are insane. I paused and read the text in the square you had highlighted. Now I'm not entirely sure I understood all the legalese corfrectly, but both times you briefly showed the rules targeting cloud service providers, they seemed to mostly/only requiring you to share the source code with the users if you make any changes to the version you're hosting.
Like I said, not sure about the legalese, maybe I missed something. This is why I wanted you to tell me about the insane rules instead of just asserting them.
As it is right now, it seems very similar to a normal open source license, except with a special provision for hosting the software instead of selling the software.
I've shared examples at 12:37.
It differs from typical open source terms in a couple of ways: 1) it applies even if you don't make any modifications, and 2) they demand the sharing of source code for completely unrelated works (that you probably don't have the legal right to share). Essentially, they try to extend the copyleft scope to any software within your stack. Not just the SSPL software, or anything derived from it. It's massive overreach, and likely impossible to actually comply with.
@@ProTechShow Ah, thank you!
The one I'm most frustrated about is fritzing. The software used to be open source but then they changed it because they wanted to compensate the devs through donations. That's annoying but don't come whining to me when your contributors say it's not worth it anymore. They say you have to donate to download but they never kept track of who donated or are to lazy to implement a system. Well, I donated to the project years ago but it doesn't matter anymore. Each time you download it, you have to pay money even if it's in the same day. The other option they cite is becoming a contributor but I can't become a contributor because they don't update their source to to most recent version anymore.
Edit: And of course, they still call it open source even though you can't view the source of at least version 1.0.0 and up
Another way of making software fake free/open source is licensing it under a free license, but including binary blobs. Examples are Linux and Telegram, both licensed under GPL2only
This reminds me of the small music program Famitracker, the original versions of which were maintained as a one-developer project. There was no public repo for its source code. New releases happened with (portable) binary .zip first, source code .zip second -- when an "open source" project is typically (if not ideally) published _source first, binaries second._
The last update, "0.50", was released only as a "beta" -- the developer never followed through with its "full" release, nor released the source for it.
Good observation. I count such projects as fakes as well.
Source code may be released years ago, current versions have no code available, but developers for some reason claim open sourceness.
@@sdjhgfkshfswdfhskljh3360 It was a quality program for its purpose, to be sure, it just wasn't "open source".
That seems a bit suspect to me as well. The code has to exist in order to create the binaries, so why would it come later?
Isn't part of the point of publishing the code to provide transparency? If I have to install the software before I get the chance to review it, what was the point?
It sounds like it may have just been a case of immaturity on the developer's part from the way you describe it, but I certainly share your misgivings.
Some will say that GPL is viral (cough cough), but SSPL is the truly cancerous license here. It even assumes that other projects can be re-licensed just because the resulting binaries from those are used along with whatever software is licensed under SSPL. Not only completely ridiculous, but completely illegal. The only one who is allowed to re-license a project is the project licensor. If whoever decided to use SSPL thinks that they own other projects should not be important and to those I say: sucks to be you. Good luck in court claiming other projects, SSPL licensors.
I had a case of closed-source software under MIT. That means I am legally allowed to modify it, but technically it is hard to do.