Hacker News Reader

Comment by pgeorgi | Hacker News Reader

Note, their idea[0] of "open source" doesn't match the popular Open Source Definition [1]

[0] https://github.com/anyproto/anytype-kotlin/blob/main/LICENSE...

A lot of the comments to this are angry and I can appreciate that. There is obviously some specific nuance to "open source" that individuals in this thread want to maintain and feel that the OSI has enshrined in stone as a definition that no one shall breach.

However, I strongly support the kind of license (in principle) that this software is released under. Source code is available for anyone to inspect, modify for their own use, contribute to, run locally for their own benefit. The main restriction is so clearly obvious: you can't create a commercial competitor. You can't take their code and with minimal effort or minor changes create a competing app and sell it to others.

This to me seems like a completely sane license. So common in fact that creative commons asks two basic questions when they recommend a license: "Allow adaptations of your work to be shared?" and "Allow commercial uses of your work?". In fact, they differentiate this difference with the moniker "Free cultural works" [1] (those that allow commercial use are termed "free").

I'd like to see the same nuance in software licenses. A difference between "open" and "free". That way, we can avoid this bickering in the comments where those who really want completely free software (free from all restrictions including those against commercial use) won't jump down the throats of those who want to open up their source while protecting themselves from competing commercial use.

1. https://creativecommons.org/share-your-work/public-domain/fr...

This isn’t bickering. These definitions have existed for literal decades. There are multiple models for making source code available, and people can choose what license they want. But this isn’t open source. This is source available, free (as in beer, kinda) license.

You don’t have to pay, and you can see the source code. But, in my quick reading, I don’t think you can make modifications, distribute modifications, distribute unmodified versions, and there is a restriction on how you use the software (non-commercial only).

This is the same type of license Microsoft gave certain large (TLA) customers for Windows, IIRC. I believe they called it “shared source,” as in they shared a copy of the source with you, but you couldn’t use the source for more than review. No one would claim that was open source.

There are differences between free, open, and available. This is only the later. No one cares about what license something is available through. Authors get to do whatever they want. People only care when you try to claim one thing, but it is really something else. In this case, the company is trying to use the term “open source” as a selling point of their software, when it isn’t. This license doesn’t even match the definitions they use on their own site!

I’m happy the authors want to make it possible to audit their software. That’s a laudable goal. If they want to restrict usage of the source code to non-commercial use, that’s fine and up to them. Just don’t call it “open source”.

Just because something is free doesn’t make it open. And just because something is open doesn’t make it free (as in freedom or beer). Similarly, just because something auditable and available, doesn’t make it free or open.

The fact that the authors don’t know the difference (or are potentially misrepresenting the difference) will only make the community mad - especially the part of the community that would care about seeing the source code in the first place. If they instead were marketing the project as “source available” for auditing or non-commercial use, this wouldn’t have been an issue.

Your first points: "can’t make modifications, distribute modifications, distribute unmodified versions" appears to contradict the language from their license file:

> Any Association grants you (“Licensee”) a license to use, modify, and redistribute the Software, but only (a) for Non-Commercial Use, or (b) for Commercial Use in Allowed Networks.

> These terms do not allow Licensee to sublicense or transfer any of Licensee’s rights to anyone else.

IANAL, but I thought transferring rights was required to redistribute a work. If someone downloads this software from me, I can’t give them a right to use that software. Only the original authors can do that. And even if an earlier clause says that I can redistribute, this cause suggests that I can’t.

As it is, this is a software license that I wouldn’t touch.

But even if I am wrong on this point, the rest of my argument stands. With the restrictions on use, this is neither an open nor free (as in freedom) license.

(Side note: this is why new authors shouldn’t roll their own licenses. Ambiguity is not what you want to see in a license agreement.)

The right you are granted is more specific, so you can redistribute, but the next hop cannot. (I'm only reading what is quoted here in the thread, but I think the important parts are included.)

I think the issue is that they say "open source!" with no further qualification. That's misleading and disingenuous.

I support the use of alternative business models like source available, Business Source License etc. That's fine. But you should accurately describe your licensing. They should have said "source available".

>The main restriction is so clearly obvious: you can't create a commercial competitor.

I don't see it. Commercial use there is defined as "where the Software facilitates any transaction of economic value other than on Allowed Networks". As I understand, it means financial application.

When will VC companies stop conflating source available with open source? It's even worse than just being closed source as it's trying to pander to a group for money, and without any respect to the actual values of what they are saying.

Our philosophy surrounding open-source is uncomplicated and clear. All essential protocols and data formats are subject to the MIT license. However, considering our clients' needs and with the MongoDB situation as a reference, we must maintain some degree of defensibility.

Our objective is to foster a collaborative atmosphere, co-creating with the community. The size of the community matters to us; when it reaches a significant scale, we want to make collective decisions regarding licensing, reflecting a democratic approach.

Antype, a creation of our non-profit organization, is aimed at sustainability rather than becoming another digital ghost. Our mission isn't simply to exist, but to thrive and make meaningful contributions to the open-source landscape. By intertwining our growth with that of our community, we're setting the stage for a sustainable future.

These are all noble goals, and I don't think many commenters here are arguing that they're not justified. Just that the specific use of the term "open source" is incorrectly applied (to your own benefit).

If a subset of things you do are open source, that's great. Say that.

Yep, you right, we'll correct it.

Thank you! Terminology matters. I really appreciate that you’re willing to make that change.

Your goals sound laudable, but they do not alter the fact that your use of the term “open source” is, deliberately or otherwise, misleading.

> Our philosophy surrounding open-source is uncomplicated and clear.

It does not appear to be the case that your philosophy surrounding open-source is clear. You state very clearly, without caveats, that the product is open-source, which strongly implies that the product is, well, open-source without caveats – and this is not the case. That feels rather disingenuous, if not deliberately dishonest.

It is not unlike printer manufactures loudly proclaiming a page-per-minute value without any note about that rate is only attainable feeding entirely blank A6 sheets out of the device.

Most of our repositories are MIT licensed, which holds significant value. Some of our repositories are source-available. We believe we are building an open-source product, and we reserve the right to define what 'open-source' means for us. Our only prohibition is against competitors making minor modifications and then selling our product. Our position on this is clear. We recognize that our definition may not align with others' perspectives here, and we are open to understanding that. However, labeling our approach as dishonest isn't something we consider accurate. To avoid any confusion for those who uphold the traditional notion of 'open source', we will change the term 'open source' to 'open code' on our website.

It's unfortunate it's not actually going to be Free Software, but being able to at least audit and do non-commercial is... better than Notion, I guess. But I do retract a fair portion of my excitement over the project.

Your "reserved right" to define what a word means to you... puts the work to somehow figure out what you mean onto the reader. This isn't quite nice. The speakers are the ones that should strive to make themselves understood in the first place.

I don't have any love for "open source" since it is just "the part of Free Software that appeases people in suits", but please, use the thing as it is.

Source available, of client and server, is definitely better than no source

I do think open source business model, especially consumer, is difficult to tackle, while protecting the creator

So congratulations and good luck

While language is by nature subjective and ever changing, be careful defending the right to redefine commonly understood terms so significantly. By the same measure people could choose to interpret anything you say, including your other license agreements, however they like!

Are you seriously trying to suggest that source code being publicly accessible is worse than nobody being able to access it? What a ridiculous statement.

You are intentionally misrepresenting OP.

> When will VC companies stop *conflating source available with open source*?

*Misrepresenting* "open source" is the problem. "source available" is the correct phrase

It was worse in the case of early BIOS, which IBM made public in an instruction manual while keeping all relevant rights on distribution. I wonder why they did not publish it with a GPL or AGPL license, where they would keep commercial use to themselves.

IBM PC was released in 1981, GPL 2.0 (which was the first version in wide use) in 1991.

You should be always skeptic but when someone write "Pure transparency — trust our code, not our words" in an EXTRA BIG font, you should be extra skeptic.

(co-founder of anytype) our main promises are privacy, end-to-end encryption, user controlled keys, self-hosting, p2p sync - all of which should add up to what we can user autonomy from the software provider which we believe to be important. To prove these claims the best way is open the source code. As promises of encryption and ownership stay promises unless you can be sure of it. That was one main motivation and why we think it's worth highlighting.

So I see the networking portions of the code are Free Software - that's great. How are people expected to use it if their use does not fall under "non-commercial use" as defined in the client license? Do you expect people to write their own clients for commercial use, or do you offer commercial licenses?

For non-commercial use, you're prohibited from selling it, but using it within your organization is permitted

You do realize this also means that a for-profit company can’t use the software at all, right?

Unmodified, modified, whatever.

I don’t think that was your intent, but that’s exactly how it reads. Or maybe that is your intent?

The idea if you want to sell clients, you need to get a permission form our association.

Why not use the GPL?

Because we want to provide other organizations with the opportunity to offer paid sync services, we needed to incorporate the concept of a network into the license. We crafted the license with that consideration in mind

You may employed a wrong strategy.

To prevent paid sync services, you should license your protocols and data formats in AGPL, which requires derivative work (third-party sync services) to be open sourced.

The client app in contrast, should be fine even in permissive licenses.

> Because we want to provide other organizations with the opportunity to offer paid sync services

Why do you think the GPL is not compatible with this?

Exactly…

Plus, they can always offer other licenses in addition to the GPL (or really AGPL would be a better fit for their concerns). It’s their software, so they can license it to anyone with whatever terms they want. (Assuming there aren’t outside contributions, but even that can be dealt with)

> Plus, they can always offer other licenses in addition to the GPL

Yep, that's exactly what I was thinking. They're clearly offering alternative licenses to the one in the public repository. There is no reason the same tactic couldn't be applied with the GPL.

Is there a way to make sure I am using the code you published and not a different version?

We do releases in the Github Actions CI. So you can inspect the CI logs and published artefacts(desktop/android). Then you can compare the binaries checksums. I would appreciate ideas on how we can make it more transparent

The conventional open source model makes everything free labor for billion dollar companies and hustlers who just take it and slap it up behind a paywall. It's used to underpin SaaS models that are significantly less open or free than closed-source local commercial software.

It'd be nice if the OSI or the open source community addressed this issue head-on, but so far they refuse and insist nothing is wrong. This refusal leads to a proliferation of almost-open-source licenses that just muddy the waters.

If they continue to refuse I think we'll see more and more of this until the definition of open source becomes hopelessly muddy and the whole community starts to wither.

Hence why the "free" in "Free and Open Source Software" really matters. :)

Not really, someone could just use that misleadingly too, just as with "open source".

"Open Source" already doesn't cover these special licenses, instead, the "source available" is used. Any also acknowledges this actually - the license they are using is called the "Any Source Available License 1.0".

After a bit of brushing up on my acronyms, one could indeed use FOSS or FLOSS to denote that a piece of software is either free (or libre) or just open source [0].

The term "open source" is not referring to software with a free license, but to software whose source code is available to the public irrespective of license [1].

[0] https://www.gnu.org/philosophy/floss-and-foss.en.html

[1] https://www.gnu.org/philosophy/open-source-misses-the-point....

That article you post as [1] refers to https://opensource.org/osd/ for "The official definition of open source software (which is published by the Open Source Initiative and is too long to include here)"

The FSF article directly follow with this, which contradicts your claim: "However, the obvious meaning for the expression “open source software” is “You can look at the source code.” Indeed, most people seem to misunderstand “open source software” that way. (The clear term for that meaning is “source available.”) That criterion is much weaker than the free software definition, much weaker also than the official definition of open source. It includes many programs that are neither free nor open source."

And the OSD also disagrees: "The license must allow modifications and derived works, and must allow them to be distributed under the same terms as the license of the original software."

Terms that can be summarized as "whose source code is available to the public" are called "source available", even by the FSF.

You are indeed correct, should have dug a little deeper (or maybe not doubt the OSD's definition in the first place). Thank you for the clarification :)

GNU also misses the point a bit. With open source, the source is open, but some other general rights are included too, like to restriction to the type usage. Lately, people and corporations made a lot of money on the backs of open source developers, so a new type of license emerged, and this would be the one that really is just about the "open" "source", but to make it distinct from the already widely known term, people call these "source available". Getting back to the topic, Any knows these distinctions too - or at least their lawyer did, because they call their license a "Source Available License"[0]. Source-available however doesn't carry the coolness of what "open source" brings - so on the marketing page, they refer to the project as "open source", which kind of can be argued, since the majority of it is indeed proper open source.

[0] https://github.com/anyproto/anytype-kotlin/blob/main/LICENSE...

Damn, didn't even think about that.

So what is the appropriate term then for "source code freely available, but cannot be used for commercial use?"

I guess probably an unpopular opinion here, but I don't see why "open source" must imply that anyone should be allowed to fork the repo and sell it.

Source available does sadly not really mean that much. Just means the source code is available. MIT licensed code and code you can see but are not allowed to use can be described as source available. That is what https://faircode.io/ got created for to solve.

Background: https://github.com/n8n-io/n8n/issues/40

> I guess probably an unpopular opinion here, but I don't see why "open source" must imply that anyone should be allowed to fork the repo and sell it.

Because that is the definition of "Open Source"[0]. As was already said, "source available" is the correct term here.

[0] https://en.wikipedia.org/wiki/The_Open_Source_Definition

Sorry but this is not the definition of "open source", I would argue there is only a conceptual and cultural definition of "open source". What you are linking to is The Open Source Initiative (OSI) Foundation's declaration called the "Open Source Definition".

They are a single organization, that have done a tremendous job at trying to come up with a global and shared legal framework to which people can license code under. They have gone so far to come up with a pretty good definition of "open source", but not the definition.

This would be equivalent to saying that "Freedom" is defined by the US Constitution or the Canadian Charter of Rights & Freedoms. It is not, those are both examples of a legal definition of freedom, but neither are the sole authority for the global and cultural concept of "Freedom"

> This would be equivalent to saying that "Freedom" is defined by the US Constitution or the Canadian Charter of Rights & Freedoms.

The idea of freedom existed before both documents. The idea of Open Source was proposed in 1998 [0], and the OSI was created to define it in the same year [1]. This is not at all equivalent.

[0] https://en.wikipedia.org/wiki/History_of_free_and_open-sourc...

[1] https://en.wikipedia.org/wiki/Open_Source_Initiative

Source Available.

> I don't see why "open source" must imply that anyone should be allowed to fork the repo

Apart from 20+ years of historical use in that way?

You're completely misrepresenting my quote by cutting off the "and sell it" at the end, that's incredibly disingenuous.

MIT license restrict commercialization. Is the MIT license not open source then?

The MIT License does not restrict commercialization aside from requiring attribution. In fact, the text of the license includes the phrase "without restriction":

> Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the "Software"), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so, subject to the following conditions:

> The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software.

https://en.wikipedia.org/wiki/MIT_License

Open source software licenses have always allowed the licensed software to be resold by others. The very first criterion in the Open Source Definition is:

> 1. Free Redistribution

> The license shall not restrict any party from selling or giving away the software as a component of an aggregate software distribution containing programs from several different sources. The license shall not require a royalty or other fee for such sale.

https://opensource.org/osd/

The "Any Source Available License 1.0" is not an open source software license because it restricts commercial use, but it happens to be correctly named because it is a source-available software license. Source-available software (such as Anytype) is still preferable to proprietary software with little to no source code published (such as Notion), since it is easier to audit software when the source code is available.

Ok I stand corrected on the MIT license, thanks for the explanation. I wonder why so many on HN criticize it then, seems reasonable to me.

It's not freely available, it's available for a limited set of uses. So it's not open source.

Co-founder of anytype is here. The most of are repos are fully open-source. We have our philosophy regarding open source here https://blog.anytype.io/our-open-philosophy/ Happy to discuss concerns regarding our approach.

Using the term Open Source for a product when most of the code is under a license that isn’t Open Source feels dishonest to me. The product you build is certainly yours to release however you see fit, but if I’m looking for open source software and I find this I’m going to be extremely skeptical of everything you say.

On top of that, the license itself is actually incredibly restrictive. I’m not a lawyer, but my read of the section on economic value seems very broad:

> does not include uses where the Software facilitates any transaction of economic value other than on Allowed Networks.

My read of “facilitates any transaction of economic value” means that I would be in violation if I used this to keep track of trading cards, made a grocery list, or tried to keep track of what I want to buy my friends for their birthdays. At least it would if I installed this on my home server and accessed it from the couch on mh phone.

Thank you the comment, it's clear that licence is not clear and we need to improve. Our idea is straightforward: if you want to use the software, you can do so for free, whether for personal use or within an organization. However, if you aim to sell it for profit, you need to contribute to its creation in some way; this is why permission is required. At least, that's the case at this early stage.

Serious question - did you have a lawyer write this? The license text and some of the comments here lead me to believe that it wasn’t done by a lawyer.

If not, you should really talk to a lawyer first about this. Preferably one with knowledge of open source licensing. New software licenses are tricky and should be done by a lawyer and not by random HN comments. (Even if this is an overly well informed set of users on software).

The problem is that proprietary licenses (such as Source Available) are viral: whatever they touch becomes proprietary.

As such, "most repos are open source" (from what I can see: MIT, some forked ones Apache 2.0) is nice, but the end product still isn't open source according to OSD.

There are people who value using "Open Source" for OSD-compliant licenses only (I tend to agree with that notion to keep things clear), but I didn't really want to discuss this: It's your project, after all, license as you wish.

I just wanted to provide a heads-up that the use of "open source" in the header here (and the front page on your site) doesn't match the expectations of a bunch of folks, so they know whether to look closer or not based on that.

I see how making the entire situation transparent muddies the message, but "Everything is Source Available, many parts are Open Source" would already clear things up a lot.

Yep it's clear, you are right!

I was going to come here and post about how this is exactly what I've been looking for, but then I read this.

So, my kneejerk reaction to this deceptive use of open source is to just say "no" and move on. However, I read through your philosphy, and I have a question.

> considering the substantial R&D resources required for the application layer, we believe that businesses and networks utilizing our software for commercial purposes should contribute towards its ongoing development, allowing maintainers to support and enhance the platform.

That seems to be the crux of the concern here. I can respect that. So, why do existing open source licenses not suit you? For example, you could release the software under the AGPL, and still dual-license it as you wish.

Rather than assume bad faith, I'm going to give you the chance to correct yourself. At the very least, calling yourself open source at the moment is deceptive, whether you realize it or not.

The most of are repost are open-sourced with MIT licence. It's clear now, that the way we put words together in the license for clients is not clear. Our idea is straightforward: if you want to use the software, you can do so for free, whether for personal use or within an organization. However, if you aim to sell it for profit(like change the logo and put price tag on it), you need to contribute to its creation in some way; this is why permission is required. At least, that's the case at this early stage.

I don't see the problem with using Anytype License instead of AGPL. AGPL allows commercial use as long as the source code is provided, while Anytype disallows commercial use. Other than that it is the same, and is "open source" regardless. As in, OSI doesn't trademark "open source".

You deny your users the most basic freedom there is, the freedom to use your software for any purpose without discrimination. This is wrong, and so is your attempt to misuse and redefine the term "open source software".

This is the second time today that someone mistakes open source and free software.

See: https://www.gnu.org/philosophy/open-source-misses-the-point....

I don't blame you, it's a bit schadenfreude on my part because "open source" companies try to dress as free software but aren't

You replied to "You deny your users the most basic freedom there is, the freedom to use your software for any purpose without discrimination." and caim that open source isn't about that.

Let's see https://opensource.org/definition-annotated/, _the_ definition for open source, specifically the sections titled "No Discrimination Against Persons or Groups" and "No Discrimination Against Fields of Endeavor":

"The license must not discriminate against any person or group of persons."

"The license must not restrict anyone from making use of the program in a specific field of endeavor. For example, it may not restrict the program from being used in a business, or from being used for genetic research."

So what GP claimed seems to be exactly Open Source's point, no?

The way we crafted it is not clear. Our idea is straightforward: if you want to use the software, you can do so for free, whether for personal use or within an organization. However, if you aim to sell it for profit, you need to contribute to its creation in some way; this is why permission is required. At least, that's the case at this early stage.

That means it's zero-cost and source available, not open source. As it's your software that's your choice, but please don't abuse the term "open source" to describe it. It's no more "open source" than, say, DaVinci Resolve is.

> It's no more "open source" than, say, DaVinci Resolve is.

DaVinci Resolve is not source available. Anytype does not restrict you from forking the code as long as it is non-commercial. And it also lets you fork the code for commercial use, if you take their permission.

So it is simply a non-commercial open source license, with permission required for commercial use. Sure, not OSI Approved License™, but certainly "open source". If this was not "open source", neither would be GNU GPL, because it isn't permissive enough.

This is no different from Qt's dual license except that GPL allows commercial use too. Or, like Creative Commons NC licenses, but for software.

Amen!

Open-Source is a generic term. It's the opposite of "closed-source".

The OSI is on the record on this: https://opensource.org/pressreleases/certified-open-source.p...

The only people who want to push the whole "The OSI's version of Free Software" defines "open-source" rather than "OSI Approved License™" are the Anti-Property GPL folks that never liked the term "open source" anyway, the trolls that want to force other people to work for free, and the people at the OSI

Yes you are right. It's not even open source...

More specifically, their license doesn't meet these criteria of open source software (which are the same as the Debian free software guidelines):

5. No Discrimination Against Persons or Groups

6. No Discrimination Against Fields of Endeavor

No. Open source is the same as free software. It is a marketing term for free software, that defines the same concept in more practical terms. If you actually read the article you linked, you would have known that.

Open source is not the same as free software.

There are broadly two camps: Camp 1 who advocate for free software & free software alone, and Camp 2 who advocate for "open source" being an all-encompassing umbrella term for a few things, including free software. Those in Camp 1 are typically not supportive of the goals of those in Camp 2. Those in Camp 2 do often try and equivocate the two terms.

No, it really is. Open source is equivalent to free software in everything but definition, and does not include anything that is not free software. There are minor disagreements between different people from the two camps which licenses to accept (e.g. Debian where this definition originated from, does not consider GFDL with invariant sections free, while FSF apparently does).

It really is the whole point of it, define more clearly what criteria must be fulfilled for software to be considered free software.

> Open source is equivalent to free software in everything but definition

Is this a typo?

No, it isn’t. Open source and free software define the same thing in different ways. FSD defines the user’s freedoms, OSD defines what the license cannot forbid. The end result is the same though.

It's not, because Free Software handles practical and ethical advantages as an indivisible unit, while open source focus only on promoting practical advantages.

You are missing the point. The definition of free software focuses on the freedoms of the user, but these freedoms are not easy to verify against a specific software license. The practical aspect of the open source definition (which really is the Debian free software guidelines with the word Debian removed) is that it gives you a toolkit, ten criteria that a licence must fulfil to be considered free software. Importantly, when this definition was created, the alternative, the free software definition, was incomplete and lacked the freedom zero [0]. Even more importantly, that text apparently wasn’t widely known back then, and even Richard Stallman himself liked the DFSG as a definition of free software [1].

[0]: https://www.gnu.org/bulletins/bull1.txt

[1]: https://news.slashdot.org/comments.pl?sid=1129863&cid=268758...

Which is the first time?

Any hacker news link?

> This is wrong

This is a highly subjective take - it might be better to stick to objective dictionary definitions.

This project clearly isn't open source, & shouldn't be advertised as such, but on the other hand the intent here is a common/popular one these days, & its not the first of its kind: I'm surprised no-one has yet coined a term for this relatively new breed of "faux-pen source" or whatever it is.

Our idea is straightforward: if you want to use the software, you can do so for free, whether for personal use or within an organization. However, if you aim to sell it for profit, you need to contribute to its creation in some way; this is why permission is required. At least, that's the case at this early stage.

You used the word "fully", which means part of them isn't fully open source. :(

Reply on news.ycombinator.com