Yes. If we want others to respect open source licenses, we need to honor them ourselves. We can't assume that source-available is "available for all uses".
I find it difficult to invest time in reading this when the author demonstrates such a basic misunderstanding of copyright:
>...software, by default, probably cannot be used, modified, or redistributed without a license from the copyright holder.
Of course it can be used and modified. It's the redistribution that is covered by copyright law.
>...you can’t really use software without someone giving you permission.
This is completely false, except where a license is required to obtain a copy to begin with. It is the license that creates the restriction, not the copyright.
I have a hard time believing this, are you suggesting that it is perfectly legal to run pirated software so long as you never entered a license agreement with the creator? Or that the existence of a license you've never interacted with and potentially have no knowledge of has an impact on the legality of you running pirated software? What if a company's internal software is leaked, surely they have the right to prevent other companies from using it?
I am not. Of course it's not legal to use pirated software. By definition pirated software is illegally obtained. But a legally obtained copy can be run without a license, if none was required to obtain the software.
As for "leaked" software: again, if copies are illegally obtained then of course it's illegal to use them.
> Of course it can be used and modified. It's the redistribution that is covered by copyright law.
By what law or principle? IIRC most software is licensed and not sold. And the copyright dictates who defines the terms of the license for software which can be copied. Otherwise how would 'freeware' and trial software be practical if folks could just modify them to bypass any limiters therein?
The same law or principle that allows me to use the pages of a book as wallpaper. Why can't I use my legally obtained copy in anyway I choose so long as it doesn't violate any law or contract? My point is that copyright only protects from the making of copies, and licenses are something else entirely.
“Software, by default,” is created with all rights reserved for the author.
I understand where you’re trying to go with your argument but if the author doesn’t have a means of legal distribution then there’s no way to have a legal copy on your computer to modify and run.
—edit—
And a license doesn’t restrict the rights to software but grants exceptions to the default “all rights reserved” copyright grant.
What? Are you really suggesting that software can’t be legally distributed by its author? And that a license is required before I can make copies for my own use? This is completely incorrect. What about fair use?
You can’t possibly have a legal copy of the software without expressed permission of the author. They may not care who does what with it (IDK, implied permission?) but they could change their mind at any time and chose to enforce their copyright. Which also means first-sale doctrine doesn’t apply as there was no “first sale”.
I also “make copies for my own use” of TV shows, written things, &etc and don’t think anyone would claim that isn’t violating copyright law, especially the producers of said content. In fact, I know this to be true as they have sent me nastygrams via email before. Also, by your reasoning, any company, like a cloud service, could take any code they want as long as they weren’t committing acts of distribution and use it however they want. I think we all know this isn’t the way it works because the anti-cloud licenses which have become popular lately would mean absolutely nothing.
And fair use is not really applicable to the current discussion because statically linking a library without permission isn’t fair use.
MAI Systems v. Peak Computer[0] says otherwise, at least in the 9th Circuit and at least covering use of the program. Other courts and Congress have bent over backwards to try and not disturb the ruling.
Even if it was overturned, modification has been a core part of copyright law since Folsom v. Marsh[1], where the Supreme Court interpreted the entire concept of derivative rights into the law. In fact, this ruling was so expansive they had to invent fair use at the same time just to avoid accidentally overturning the 1st Amendment.
Copyright creates the restriction, specifically to legally mandate a negotiation for the license that absolves you of the restriction. If you do not have a license, you have nothing but a pending lawsuit.
This case establishes that the copy in question was in fact a "copy" as defined by the copyright act, but the infringement was due to the disposition of the parties' rights via a license - not inherent to the copyright. This ruling is highly specific to the case.
Problem is, if you do not license your software, then under existing 9th Circuit precedent it is illegal to use your software - for any purpose. You are creating a trap to be sprung on people who don't understand the law.
Is the US legal system anti-justice, at least in the sense that the adversarial model of courts breaks down when applied to power disparities[0]? Absolutely.
Is worrying about copyright at best a detour, in the sense that copyright has nothing to do with justice except to protect a certain few from a specific kind of exploitation? Yes.
But we still need to at least pay lip service to it, if only to protect your users from being sued by you. "You", in this case, is maximally extended to not only every "you" that might exist in the future, but even people who may wind up inheriting your estate in the far future - which isn't far enough to cover the expiration of the copyright. You - as in, not you, but You - can stomp in at any time and sue individual users for large sums of money they can't pay, even if they merely ran the program.
This isn't a hypothetical. The way that people share (freely-licensed and otherwise) software puts a target on them, and the system occasionally takes the shot[1]. For as long as copyright applies to the air we breathe, we have to worry about it.
[0] Copyright trolling is a good example of why adversarial court systems don't work when parties are not on a level playing field
I don't agree with that interpretation of the MIT license. I'm aware that it's popular, perhaps majority, but I think it wouldn't hold up in court, though I hope we never find out, it would be a churlish point to sue over imho.
The clause (the "Software") is defined, and when we unroll the relevant line using that definition we get this:
> The above copyright notice and this permission notice shall be included in all copies or substantial portions of this software and associated documentation files
First note: it says this software, and source code is routinely distributed without the binary. It also says "and associated documentation files", and here we have a dog which doesn't bark: the clear opportunity to say "as well as compiled artifacts and object" was not taken, leading a reasonable person to be able to conclude this omission was intentional.
I happen to think the point was to include copyright notice on source and documentation, but not on binaries, which is awkward to do. If you want to be safe, ok, sure, find a way, and IANAL.
Is the binary no longer the software? All precedent would suggest otherwise. (Though I can see it may not be a copy of the software, as compilation is usually more than that.)
That's the subtlety, the clause is introduced thus:
> Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the "Software")
And I maintain that the use of this in reference to the software is dispositive that it doesn't refer to anything which doesn't exist in the repo/tarball/whatever, particularly not object code, and that the reference to associated documentation files is evidence that the omission was deliberate.
This is why lawyers should write licenses though. The ambiguity sucks.
> …the risk-reward ratio of theoretical copyright litigation is not good, especially when litigating against open source projects tends to paint a target on your back.
And the example they give (without the link being preserved) is a patent case which is completely different.
75 years from now the heirs of the developer who wrote the library being discussed can go on a copyright lawsuit spree and there’s no amount “FLOSS community pressure” which could stop them.
People keep giving advice on this issue of “it probably doesn’t matter, they put it on GitHub because they want people to use the code” which might be true today but it also might matter a whole lot tomorrow for whatever reason.
The article is very opinionated ad I don't agree with some of it, but the problem is real and we should think of how to deal with it for the future.
As I see it, currently the only sure-fire way to deal with unlicensed material is to treat it as "all rights reserved", until it hits the public domain.
There are other reasonable approaches, but none universal and they sure imply a fair amount of due diligence and risk assessment.
A good solution could be a universal, retroactive, default-license. Any license would do. Because it's the absence which creates the uncertainty, the traps and the variable interpretations all over the world. Yet, while this would be very useful, I don't see it happening anytime soon...
So, what we can do in the meantime? We could gently ask authors to choose a license for their unlicensed artifacts whenever we stumble upon one. — Which I already do, with a few gratifying successes, if I must say myself. :)
Also, many code-repositories are already helping the problem by pushing for the choice of a license at the creation of new projects, but we all need to push more. Lobbying the code-repos to progressively give stronger nudges to the existing unlicensed stuff, which is indeed the more difficult to address otherwise...
19 comments
[ 5.9 ms ] story [ 58.8 ms ] threadYes. If we want others to respect open source licenses, we need to honor them ourselves. We can't assume that source-available is "available for all uses".
>...software, by default, probably cannot be used, modified, or redistributed without a license from the copyright holder.
Of course it can be used and modified. It's the redistribution that is covered by copyright law.
>...you can’t really use software without someone giving you permission.
This is completely false, except where a license is required to obtain a copy to begin with. It is the license that creates the restriction, not the copyright.
As for "leaked" software: again, if copies are illegally obtained then of course it's illegal to use them.
By what law or principle? IIRC most software is licensed and not sold. And the copyright dictates who defines the terms of the license for software which can be copied. Otherwise how would 'freeware' and trial software be practical if folks could just modify them to bypass any limiters therein?
First-sale doctrine?
Which, ironically, doesn’t apply in this case since the software wasn’t legally distributed.
I understand where you’re trying to go with your argument but if the author doesn’t have a means of legal distribution then there’s no way to have a legal copy on your computer to modify and run.
—edit—
And a license doesn’t restrict the rights to software but grants exceptions to the default “all rights reserved” copyright grant.
You can’t possibly have a legal copy of the software without expressed permission of the author. They may not care who does what with it (IDK, implied permission?) but they could change their mind at any time and chose to enforce their copyright. Which also means first-sale doctrine doesn’t apply as there was no “first sale”.
I also “make copies for my own use” of TV shows, written things, &etc and don’t think anyone would claim that isn’t violating copyright law, especially the producers of said content. In fact, I know this to be true as they have sent me nastygrams via email before. Also, by your reasoning, any company, like a cloud service, could take any code they want as long as they weren’t committing acts of distribution and use it however they want. I think we all know this isn’t the way it works because the anti-cloud licenses which have become popular lately would mean absolutely nothing.
And fair use is not really applicable to the current discussion because statically linking a library without permission isn’t fair use.
Even if it was overturned, modification has been a core part of copyright law since Folsom v. Marsh[1], where the Supreme Court interpreted the entire concept of derivative rights into the law. In fact, this ruling was so expansive they had to invent fair use at the same time just to avoid accidentally overturning the 1st Amendment.
Copyright creates the restriction, specifically to legally mandate a negotiation for the license that absolves you of the restriction. If you do not have a license, you have nothing but a pending lawsuit.
[0] https://en.wikipedia.org/wiki/MAI_Systems_Corp._v._Peak_Comp....
[1] https://en.wikipedia.org/wiki/Folsom_v._Marsh
Problem is, if you do not license your software, then under existing 9th Circuit precedent it is illegal to use your software - for any purpose. You are creating a trap to be sprung on people who don't understand the law.
Is the US legal system anti-justice, at least in the sense that the adversarial model of courts breaks down when applied to power disparities[0]? Absolutely.
Is worrying about copyright at best a detour, in the sense that copyright has nothing to do with justice except to protect a certain few from a specific kind of exploitation? Yes.
But we still need to at least pay lip service to it, if only to protect your users from being sued by you. "You", in this case, is maximally extended to not only every "you" that might exist in the future, but even people who may wind up inheriting your estate in the far future - which isn't far enough to cover the expiration of the copyright. You - as in, not you, but You - can stomp in at any time and sue individual users for large sums of money they can't pay, even if they merely ran the program.
This isn't a hypothetical. The way that people share (freely-licensed and otherwise) software puts a target on them, and the system occasionally takes the shot[1]. For as long as copyright applies to the air we breathe, we have to worry about it.
[0] Copyright trolling is a good example of why adversarial court systems don't work when parties are not on a level playing field
[1] https://torrentfreak.com/comcast-subscriber-receives-dmca-no...
https://torrentfreak.com/paramount-wipes-infringing-ubuntu-t...
The clause (the "Software") is defined, and when we unroll the relevant line using that definition we get this:
> The above copyright notice and this permission notice shall be included in all copies or substantial portions of this software and associated documentation files
First note: it says this software, and source code is routinely distributed without the binary. It also says "and associated documentation files", and here we have a dog which doesn't bark: the clear opportunity to say "as well as compiled artifacts and object" was not taken, leading a reasonable person to be able to conclude this omission was intentional.
I happen to think the point was to include copyright notice on source and documentation, but not on binaries, which is awkward to do. If you want to be safe, ok, sure, find a way, and IANAL.
> Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the "Software")
And I maintain that the use of this in reference to the software is dispositive that it doesn't refer to anything which doesn't exist in the repo/tarball/whatever, particularly not object code, and that the reference to associated documentation files is evidence that the omission was deliberate.
This is why lawyers should write licenses though. The ambiguity sucks.
And the example they give (without the link being preserved) is a patent case which is completely different.
75 years from now the heirs of the developer who wrote the library being discussed can go on a copyright lawsuit spree and there’s no amount “FLOSS community pressure” which could stop them.
People keep giving advice on this issue of “it probably doesn’t matter, they put it on GitHub because they want people to use the code” which might be true today but it also might matter a whole lot tomorrow for whatever reason.
Bad, bad advice…
As I see it, currently the only sure-fire way to deal with unlicensed material is to treat it as "all rights reserved", until it hits the public domain. There are other reasonable approaches, but none universal and they sure imply a fair amount of due diligence and risk assessment.
A good solution could be a universal, retroactive, default-license. Any license would do. Because it's the absence which creates the uncertainty, the traps and the variable interpretations all over the world. Yet, while this would be very useful, I don't see it happening anytime soon...
So, what we can do in the meantime? We could gently ask authors to choose a license for their unlicensed artifacts whenever we stumble upon one. — Which I already do, with a few gratifying successes, if I must say myself. :)
Also, many code-repositories are already helping the problem by pushing for the choice of a license at the creation of new projects, but we all need to push more. Lobbying the code-repos to progressively give stronger nudges to the existing unlicensed stuff, which is indeed the more difficult to address otherwise...