[dllthomas]

Uncontroversially, dual licensing is not open source when neither of the licenses are open source.

More controversially, a "non-commercial" license is not open source. This is clearly the position of the OSI, though that isn't necessarily dispositive. But note that this is not the same thing as a license with terms that happen to incidentally make some commercial use impractical (like the AGPL), which may be what the parent meant by "non-commercial", in which case we just have people talking past each other.

[Abishek_Muthian]

If I publish the source of a core application under AGPL and then use that application in my web service under EULA(Say general T&C) would I myself be in violation of AGPL since I didn't release the source for my web service (or) Would I be able to claim it as dual licensed?

Note: AGPL in the above example is specifically to deter commercialization of the application by competitors while allowing the consumers to read the source and use the application without paying to my web service if they wish so.

[liminvorous]

If you own the copyright to the application then you do not need a license to use it. The case where you would be violating a license would be if someone else owns the copyright to parts of the application, for example because you accepted code from external contributers under the AGPL, or reused other AGPL code.

[Abishek_Muthian]

Thanks for the clarification.

So, It's best to use permissible license for the open-source version of the application in case of dual-license if we want to take back the contributions to our non open-source version.

I see this is what dual-licensed projects seem to do, I use open-source version of Aseprite[1] under MIT but I paid for the license on their website under EULA.

If anyone has seen a dual-licensed project where the open-source version is under AGPL or similar restrictive license then please mention.

[1] https://github.com/aseprite/aseprite/

[dllthomas]

> So, It's best to use permissible license for the open-source version of the application in case of dual-license if we want to take back the contributions to our non open-source version.

That's one option; the other is to require that contributors assign you copyright of their contributions (perhaps with compensation) so you can release it in the same manner as your own code.

Narrowly focused on the effectiveness of the dual license scheme itself, permissive licenses "don't work as well" because there is less incentive for anyone to pay you for the other license; typically all they'd be avoiding is the attribution requirements.

[Abishek_Muthian]

> That's one option; the other is to require that contributors assign you copyright of their contributions (perhaps with compensation)

I thought about it before I made my previous comment, Good to know this is a thing.

> permissive licenses "don't work as well" because there is less incentive for anyone to pay you for the other license; typically all they'd be avoiding is the attribution requirements.

As far consumers are concerned, I think it largely depends upon the project; If it's complex to setup and run then I guess people would use the hosted version. I use open-source Aseprite because it's available on AUR, But I did pay for the license on their website which I presume most wouldn't.

Perhaps the real concern with permissible license are the competitors, Who get to use your code to reduce their barrier for entry. Then again, Software code at current times aren't that big of a barrier to entry.

One thing which all these discussions prove to me is that open-source application funding is very complex, Depending upon it for livelihood with just good will of the consumers is very risky as consumers are on average poor and corporations are on average greedy.

[dllthomas]

As you say, it depends a lot on the project, but you're thinking too narrowly when it comes to types of projects. IME, dual license is most applicable to libraries that might wind up in native applications. With, say, "GPL or pay us for proprietary" a company trying to ship a native app is faced with the choice of not using the library, paying for the library, or releasing their product under the GPL to use the library while avoiding paying for it.

[TAForObvReasons]

OSI muddies the waters by accepting AGPL as an open source license. As argued many times, it is really a EULA, not a proper license in the sense that GPL or BSD or MIT are software licenses.

[monocasa]

Since when are EULAs not software licenses?

[TAForObvReasons]

https://news.ycombinator.com/item?id=31190064 explains it well:

> The GPL is a copyright license. It applies when copyright law would otherwise prevent distributing some program. If I were to violate the GPL, for example by sending someone a compiled version and refusing to provide the source, then there's a clear legal theory by which copyright law could be applied.

> That "one added requirement" in the AGPL turns it into a EULA, because it's an attempt to regulate for which purposes a user may run the program. If I were to run a modified AGPL'd SSH server on my own hardware, the question of whether I'm violating the AGPL depends on whether I'm allowing others to access it remotely. If the AGPL can be violated without copyright infringement, it's clearly in a different legal category than the GPL.

[teraflop]

That explanation is, as far as I can tell, completely wrong. I think it's telling that it doesn't actually cite the sections of the AGPL's text that would point out the error.

Copyright law -- at least, in the US -- protects both the right to distribute something and the right to modify it (technically, to "prepare derivative works"; see 17 USC §106 for the exact wording). Both the GPL and the AGPL grant you the right to modify and/or redistribute a covered work, it's just that the conditions that they impose are different.

In particular, section 9 of the AGPL says quite clearly:

> You are not required to accept this License in order to receive or run a copy of the Program. Ancillary propagation of a covered work occurring solely as a consequence of using peer-to-peer transmission to receive a copy likewise does not require acceptance. However, nothing other than this License grants you permission to propagate or modify any covered work.

And all of the other clauses of the AGPL are carefully phrased as conditions under which you may modify or distribute the work, not conditions under which you may use it.

So if you simply receive a modified copy of an AGPL-licensed program and run it, then no matter who accesses it you aren't in violation of the AGPL, because you have no need to accept it. But if you are the one who does the modification, and the modification doesn't comply with the AGPL (e.g. because the modified software doesn't provide remote users with the source) then you are in violation, and it's because you did something that wasn't permitted by copyright law.