Comment by knocte
7 years ago
I could have written the same last sentences as you, but using a GPL license. No difference. Licensing is not related to this. (PS: When you choose "GPL", if someone modifies your software, they don't "owe" you those modifications; they only owe those modifications to their users, and only in case they have users. Again, no community-management here involved, just distributing the sources.)
I would argue that the condition that you provide sources to your users is something you owe the author. But perhaps I'm over-philosophical.
> I would argue that the condition that you provide sources to your users is something you owe the author. But perhaps I'm over-philosophical.
No, I think you're right.
But you only owe it to the author if you're a "distributor". If you're a "user" (=a human who interacts with an application), you don't owe her anything.
The GPL keeps the distinction between the "users" and the "distributors".
This is important, because they have different interests ; They can't both simultaneously have complete freedom to do what they want.
If the "distributors" have the freedom not to redistribute the source code, then the "users" lose their freedom to study and modify the program (or to have it studied/modified by anyone competent - even software companies sometimes hire consultants to modify free software they don't feel like modifying themselves).
Basically:
- the GPL says to the users "do what you want", and says to the distributors "let the users do what they want (which implies: let them have the source code)".
- The MIT says to everyone "do what you want" (I'm omitting the copyright notice stuff here for brevity)
Thus, the MIT blurs this distinction "user"/"distributor" - which is fine, but creates lots of confusion when people try to understand the rationale behind the copyleft (leading to nonsensical reasonning like "there are things I can't do so GPL is less free!").
In that case, with MIT you owe the author some credits. (By experience, a lot of people won’t even honor that, even going to the trouble of erasing your copyright notice - literally the only thing they shouldn’t do...)
And only if their users receive a copy of the software in the process of using it. Source: Myself, who slings lots of technically-GPLed proprietary code in the form of network-accessible software like websites so never sees the light of day outside the customer's private repos.
Users must be able to readily obtain a copy of the source code. They don't have to receive it.
GPLv3 Section 6 d: "Convey the object code by offering access from a designated place (gratis or for a charge), and offer equivalent access to the Corresponding Source in the same way through the same place at no further charge. You need not require recipients to copy the Corresponding Source along with the object code."
https://www.gnu.org/licenses/gpl.html
No, I mean, if the "users" are on the other end of a socket and never receive any copy of the sofware (in any form) then they don't get the entitlement to the software in code form either. It's the problem AGPL was designed for.
1 reply →
From section 0: To “convey” a work means any kind of propagation that enables other parties to make or receive copies. Mere interaction with a user through a computer network, with no transfer of a copy, is not conveying.