The article is for Technology Review ... so, I would expect some simplification :-)
While it is true that most free software projects don't get anywhere, I think the FOSS community has already proven that the open source process can produce software of equal or better quality than proprietary products, and in a reasonable timeframe (or, at minimum a *comparable* timeframe to closed-source proprietary products).
The FOSS community does not mean unpaid hackers in a basement or dorm room. Big corporations (IBM, for example) have paid employees working on software that they modify as part of the open source community! Their interests are profit-driven, and, in their cases, the collaborative model produces usable software with the features their customers want faster and cheaper than they could if they were working with another company or their own closed-source product (this is something along the lines of an Eric Raymond/OSI type argument).
As for the Supreme Court, I think that the case undermines the Constitutional argument, but not the social argument. In order to make any changes, Lessig &company will have to appeal to the public and the lawmakers to do something that goes against a lot of lobbying money.
I disagree on the "rights clearance" not being an issue for source code. GPLFlash was saved from the junk-heap to become useful by the GPL. And, wouldn't it be nice if you could still support any program you used to use and like, but can't find anymore, because the company went out of business or killed the product?
I don't understand the contract law point, so I can't really comment.
In case you don't believe me, here are a few examples: There was no Linux kernel at the beginning of 1991; For reference, the Linux kernel became usable, and GPL'd about the same time Windows 3.1 was released (MS had a big head start!). Now, it can run everything from mobile phones to supercomputers.
The GNU project started in 1984, the same year as the first Apple Macintosh was released. By the time Linux came around in 1990, the GNU userland was an almost complete operating environment, lacking only a kernel.
Almost everyone in the mathematical sciences (and some outside those fields) uses LaTeX as a text processing and typesetting tool for creating structured documents. It started with bare typesetting software in 1977, the first uasbility macros (LaTeX) in 1983, and is still gradually improving. I don't even know of any widely used non-free competing product.
For more, see: The X-Window system, the Apache web server, MySQL and PostgreSQL database products, OpenOffice.org...
(Historical references were all found by trivial Google searches.)
Well, I still don't think we disagree all that much :-) I did not mean to say that a collaborative effort pursued strategically can't work. On the contrary, I think inviting collaboration in the right context can be extremely savvy -- although I have worked with companies gravely concerned that while doing so they not inadvertently subject their limited proprietary "crown jewels" to the GPL or another open-source license.
Overall, I was responding to what I perceive to be fuzzy, naive arguments to the effect that collaborating basement and dorm-dwellers would typically be just as efficient working "li'shma," as it were. I think Lessig's article seeks to feed that myth.
On the issue of rights clearance, I think that strong arguments can be made that, in addition to fair use arguments under Section 107 of the Copyright Act, Section 117 already permits system maintenance, as long as the user already owns a licensed copy. Many software licenses include software escrow provisions with events like insolvency or dissolution entitling the customer to obtain source code. Where no such provision exists and the company disappears so thoroughly that no rightsholder can be identified, a true system maintenance kind of use would be extremely unlikely to result in any liability or allegations of infringement. If it proves to be a problem, statutory fixes more narrowly focused than copyright term changes could be implemented without great difficulty.
On Lessig's Eldred case, the constitutional argument was, in my view, a very thinly veiled policy argument. The Constitution refers to "securing for limited Times to Authors and Inventors the exclusive Right to their respective Writings and Discoveries." Lessig argued for Eldred that Congress was not entitled to extend copyright terms by a finite (albeit long) period, because doing so effectively amounted to a perpetual copyright term, rather than "limited Times." Although I think he's right from a policy standpoint, the legal argument is quite flimsy. The Supremes didn't buy it either, and the two dissenters focused on the policy arguments. So, Lessig is using the same policy arguments to attempt persuade enough people to press Congress on the issue.
Finally, in terms of the contract law point in my prior post, I mean that the enforceability of open source license "agreements" like the GPL will hinge mostly on whether courts are persuaded that users actually understood and affirmatively assented to its terms, and if so, the extent to which the terms (which are somewhat novel from a legal standpoint) should be enforced. All of these determinations flow from the law of contracts, not from the copyright statute. A license agreement is a kind of contract, after all, even if its subject matter happens to involve copyright.
I don't think we disagree that much either, I just like prolonging arguments. :-)
In populist articles, the tendency is to try to get the reader to "favor the underdog," and that's why the emphasis is placed on the "garage-band" set. People are more likely to identify with them. I think that's more necessary in arguing for free culture than in arguing for free software. That said, it was the basement and dorm-dwellers who are now the big names in both the proprietary and free-software worlds.
The "system maintenance" of the type I'm referring to may not involve a critical business system. Let's say you just really liked using WordStar, but, it can't handle some random, but important feature that more modern word processors have. I don't think that would make a sufficient legal argument to force the defunct company or its successors to give you source code. Your documents were still accessible using other compatible software (or your old, licensed versions of the software).
And, thanks for the explanation of your contract law point. In terms of "understanding", I could see a case making it to the court involving the meaning of "linkage." I don't see how a company like Fortinet could claim that they didn't quite get it.
If, by some chance, a court did find that the GPL's copyleft/share-alike provision (probably its most "novel" part) were unenforcable, I don't see how that would help someone violating it. Wouldn't it go back to a standard copyright, where the violator would have no rights to distribute it at all?
It only "goes back to copyright" if the plaintiff establishes: (1) ownership of a valid copyright in the code at issue; and (2) that the copying/modification at issue was not authorized by a license agreement or otherwise; AND (3) the defendant fails to establish a valid defense, such as fair use or other statutory exemption (such as under Section 117).
All of this is easier said than done (as with any kind of litigation).
Also, the strongest defense argument regarding enforceability would be lack of affirmative assent, which more courts have been requiring lately to enforce "clickwrap" or "browsewrap" license terms. Note: affirmative assent can be accomplished by clicking an "I Agree" button.
What would be considered "affirmative assent" when the work in question is on a computer, readable, but not directly executable? The GPL primarily licenses how one may use and distribute source-code. You can't "click-wrap" source code, or, an e-book for that matter. So, wouldn't a copyright notice at the top of the file and the location of the copyright license be sufficient? That is where GPL notices are usually put.
For the sake of binaries compiled from GPL source code, I'm not sure how the --version, --copyright/--copyleft switch and/or the startup banner fit into all this; in a case of a GPL violation, the banner and copyright notice are usually removed.
While it is true that most free software projects don't get anywhere, I think the FOSS community has already proven that the open source process can produce software of equal or better quality than proprietary products, and in a reasonable timeframe (or, at minimum a *comparable* timeframe to closed-source proprietary products).
The FOSS community does not mean unpaid hackers in a basement or dorm room. Big corporations (IBM, for example) have paid employees working on software that they modify as part of the open source community! Their interests are profit-driven, and, in their cases, the collaborative model produces usable software with the features their customers want faster and cheaper than they could if they were working with another company or their own closed-source product (this is something along the lines of an Eric Raymond/OSI type argument).
As for the Supreme Court, I think that the case undermines the Constitutional argument, but not the social argument. In order to make any changes, Lessig &company will have to appeal to the public and the lawmakers to do something that goes against a lot of lobbying money.
I disagree on the "rights clearance" not being an issue for source code. GPLFlash was saved from the junk-heap to become useful by the GPL. And, wouldn't it be nice if you could still support any program you used to use and like, but can't find anymore, because the company went out of business or killed the product?
I don't understand the contract law point, so I can't really comment.
In case you don't believe me, here are a few examples:
There was no Linux kernel at the beginning of 1991; For reference, the Linux kernel became usable, and GPL'd about the same time Windows 3.1 was released (MS had a big head start!). Now, it can run everything from mobile phones to supercomputers.
The GNU project started in 1984, the same year as the first Apple Macintosh was released. By the time Linux came around in 1990, the GNU userland was an almost complete operating environment, lacking only a kernel.
Almost everyone in the mathematical sciences (and some outside those fields) uses LaTeX as a text processing and typesetting tool for creating structured documents. It started with bare typesetting software in 1977, the first uasbility macros (LaTeX) in 1983, and is still gradually improving. I don't even know of any widely used non-free competing product.
For more, see: The X-Window system, the Apache web server, MySQL and PostgreSQL database products, OpenOffice.org...
(Historical references were all found by trivial Google searches.)
Reply
Overall, I was responding to what I perceive to be fuzzy, naive arguments to the effect that collaborating basement and dorm-dwellers would typically be just as efficient working "li'shma," as it were. I think Lessig's article seeks to feed that myth.
On the issue of rights clearance, I think that strong arguments can be made that, in addition to fair use arguments under Section 107 of the Copyright Act, Section 117 already permits system maintenance, as long as the user already owns a licensed copy. Many software licenses include software escrow provisions with events like insolvency or dissolution entitling the customer to obtain source code. Where no such provision exists and the company disappears so thoroughly that no rightsholder can be identified, a true system maintenance kind of use would be extremely unlikely to result in any liability or allegations of infringement. If it proves to be a problem, statutory fixes more narrowly focused than copyright term changes could be implemented without great difficulty.
On Lessig's Eldred case, the constitutional argument was, in my view, a very thinly veiled policy argument. The Constitution refers to "securing for limited Times to Authors and Inventors the exclusive Right to their respective Writings and Discoveries." Lessig argued for Eldred that Congress was not entitled to extend copyright terms by a finite (albeit long) period, because doing so effectively amounted to a perpetual copyright term, rather than "limited Times." Although I think he's right from a policy standpoint, the legal argument is quite flimsy. The Supremes didn't buy it either, and the two dissenters focused on the policy arguments. So, Lessig is using the same policy arguments to attempt persuade enough people to press Congress on the issue.
Finally, in terms of the contract law point in my prior post, I mean that the enforceability of open source license "agreements" like the GPL will hinge mostly on whether courts are persuaded that users actually understood and affirmatively assented to its terms, and if so, the extent to which the terms (which are somewhat novel from a legal standpoint) should be enforced. All of these determinations flow from the law of contracts, not from the copyright statute. A license agreement is a kind of contract, after all, even if its subject matter happens to involve copyright.
Reply
In populist articles, the tendency is to try to get the reader to "favor the underdog," and that's why the emphasis is placed on the "garage-band" set. People are more likely to identify with them. I think that's more necessary in arguing for free culture than in arguing for free software. That said, it was the basement and dorm-dwellers who are now the big names in both the proprietary and free-software worlds.
The "system maintenance" of the type I'm referring to may not involve a critical business system. Let's say you just really liked using WordStar, but, it can't handle some random, but important feature that more modern word processors have. I don't think that would make a sufficient legal argument to force the defunct company or its successors to give you source code. Your documents were still accessible using other compatible software (or your old, licensed versions of the software).
And, thanks for the explanation of your contract law point. In terms of "understanding", I could see a case making it to the court involving the meaning of "linkage." I don't see how a company like Fortinet could claim that they didn't quite get it.
If, by some chance, a court did find that the GPL's copyleft/share-alike provision (probably its most "novel" part) were unenforcable, I don't see how that would help someone violating it. Wouldn't it go back to a standard copyright, where the violator would have no rights to distribute it at all?
Reply
All of this is easier said than done (as with any kind of litigation).
Also, the strongest defense argument regarding enforceability would be lack of affirmative assent, which more courts have been requiring lately to enforce "clickwrap" or "browsewrap" license terms. Note: affirmative assent can be accomplished by clicking an "I Agree" button.
Reply
For the sake of binaries compiled from GPL source code, I'm not sure how the --version, --copyright/--copyleft switch and/or the startup banner fit into all this; in a case of a GPL violation, the banner and copyright notice are usually removed.
Reply
Leave a comment