GPL up to version 3 – DragonFly BSD Digest

I don’t know what’s so amazing about that; the previous version had no revocation [because-we-came-up-with-something-new] clause, and including one would’ve been bad form and a big legal headache.

“This alternative is allowed only occasionally and noncommercially” is a nice tar pit; it’s noble to try to keep this special case from being sausage-making (“it happens, but you don’t want to watch”), I guess, so now everyone can worry about the meaning of ‘occasionally’ by jurisdiction… and ill-defined overreach can motivate a court to toss a license.

6e bumps up against 9’s concept of “ancillary propagation;” they obviously wanted to draw a distinction between being a BitTorrent seed and BitTorrent client, but this seems ugly and open to judicial misinterpretation. It also invites the creation of precedent for an unpleasantly narrow legal definition of ‘peer-to-peer,’ though maybe all the RIAA and MPAA lawsuits already have; I would’ve preferred a specific term like ‘decentralized file-sharing application’ and an explicit definition. (Accordingly, one could also bake a definition of ‘application’ as ‘use of a computer for specific means by use of specific object code or protocol’.)

3 throws in ‘users’ undefined, and speaks from an awkward vantage point; instead of the licensor directly declaring the work not to be such a mechanism, we get the broad ‘No covered work shall be…’ (Knee-jerk judicial response: “Oh yeah? Let’s see the definition from the statute.”) Similarly, some jurisdictions do not take kindly to contracts which attempt to restrict exercise of legal rights; this only, well:

Disclaim[s] any intention to [do specific things not permissible by the rest of the license anyway] as a means of enforcing … legal rights …

So it’s toothless language that manages to include an inflammatory phrase. Awesome. Why not just state that the *licensor* (aka ‘author,’ ‘copyright holder’) promises to waive his rights to invoke anti-circumvention law, and let this promise propagate with each modification to the work? Right now it’s defining a global whether or not it has permission, while defining a local for each individual work is sheltered under the author’s right to apply terms to his creation. O_o

5b’s second sentence is useless; a modified copy is logically not a verbatim copy.

5d: Intention is clear, language is flawed to the point of WTF. Someone forgot to define work as the specific changes added by the current licensor. In fact, the GPL’s always had that little problem with defining who exactly is granting the contract, originally addressed by the suggestion to attribute copyright to the Free Software Foundation (since, in theory, anyone writing GPLware can claim agency with the FSF, except for the part where they forgot to explain that, so people applied the license while asserting personal copyright as well)… I think that idea fell off the back of the cart. Something like “All those whose names appear hereby grant you…” could solve that, if only they’d tried.

Wow, back up to point 2, which refers in singular to ‘the term of copyright on the Program,’ while all ‘work’ added to the Program originates with copyright from that work’s author (in the US, though it can’t hurt to at least be consistent with the law of the land this beast is assembled in). There’s got to be a precedent for determining term on collaborative… um, “works”… Maybe they’re purposefully vague to stay applicable internationally, but the sparks are going to fly within 120 years or so, or whenever some party decides to assert that the work’s fallen into the public domain and a recent modifier begs to differ. *NIXes, like cockroaches, will still be around in 120 years, right? ;)

7e: The â€œofficial-mozilla-brandingâ€ clause, wherein the DMCA can’t restrict your right to use what you’ve been handed, but trademark law still can. Great. What was once only implied by external statute is now codified. Nothing to complain about legally, but if going this far, codifying the Mozilla compromise (terms added to restrict trademarks permissible *if* the source code as conveyed can produce functioning object code not using them) would have been one less practical headache. Look forward to Cisco/Linksys, Novell, etc plastering their trademarks all over the place, maybe even some schmuck utilities to insert code that combines to display a trademark in as many disparate locations as possible. (Easy to clean out? What if it’s put in with 1,000 time redundancy…?)

That’s a lot of disorganized nitpicking, and the sun’s coming up as I write this. Still, it goes to show that even I, entirely in favor of the license’s intentions, find it a mindf*** to actually work with. The Creative Commons tangle is starting to look simple by comparison.

Lots of Lessig logic in this one (well-intentioned, well-reasoned to a point, but disjointed and overextended in the end), not so much Stallman (recursive, obscure, yet internally consistent). Maybe it’s the hurd-of-lawyers effect at work; everyone gets their pet terms in, at which point it would be unconscionably rude for any one person to clean up (or pare down) the whole. After all, if you offend any large corporations who dared contribute, they might pack up and leave for a different bazaar.

6 Replies to “GPL up to version 3”

Wiger van Houten says:

2007/06/30 at 15:29

What amazed me, is that the Novell-MS deal is in no way hindered with this deal. It’s only relevant for >future
Wiger van Houten says:

2007/06/30 at 15:36

What amazed me, is that the Novell-MS deal is in no way hindered with this deal. It’s only relevant for Future deals.

http://linux.slashdot.org/article.pl?sid=07/06/22/1229213&from=rss

There’s some discussion as well about TiVO in this /. -article

(sorry for posting twice; i had the word “future” between “smaller than” and “bigger than” signs… which caused the message to cut off for some reason)

Anyway, I agree with the logo… it looks like a bee which will sting you in the ass ;).
Joe "Floid" Kanowitz says:

2007/07/01 at 04:46

I don’t know what’s so amazing about that; the previous version had no revocation [because-we-came-up-with-something-new] clause, and including one would’ve been bad form and a big legal headache.

“This alternative is allowed only occasionally and noncommercially” is a nice tar pit; it’s noble to try to keep this special case from being sausage-making (“it happens, but you don’t want to watch”), I guess, so now everyone can worry about the meaning of ‘occasionally’ by jurisdiction… and ill-defined overreach can motivate a court to toss a license.

6e bumps up against 9’s concept of “ancillary propagation;” they obviously wanted to draw a distinction between being a BitTorrent seed and BitTorrent client, but this seems ugly and open to judicial misinterpretation. It also invites the creation of precedent for an unpleasantly narrow legal definition of ‘peer-to-peer,’ though maybe all the RIAA and MPAA lawsuits already have; I would’ve preferred a specific term like ‘decentralized file-sharing application’ and an explicit definition. (Accordingly, one could also bake a definition of ‘application’ as ‘use of a computer for specific means by use of specific object code or protocol’.)

3 throws in ‘users’ undefined, and speaks from an awkward vantage point; instead of the licensor directly declaring the work not to be such a mechanism, we get the broad ‘No covered work shall be…’ (Knee-jerk judicial response: “Oh yeah? Let’s see the definition from the statute.”) Similarly, some jurisdictions do not take kindly to contracts which attempt to restrict exercise of legal rights; this only, well:

Disclaim[s] any intention to [do specific things not permissible by the rest of the license anyway] as a means of enforcing … legal rights …

So it’s toothless language that manages to include an inflammatory phrase. Awesome. Why not just state that the *licensor* (aka ‘author,’ ‘copyright holder’) promises to waive his rights to invoke anti-circumvention law, and let this promise propagate with each modification to the work? Right now it’s defining a global whether or not it has permission, while defining a local for each individual work is sheltered under the author’s right to apply terms to his creation. O_o

5b’s second sentence is useless; a modified copy is logically not a verbatim copy.

5d: Intention is clear, language is flawed to the point of WTF. Someone forgot to define work as the specific changes added by the current licensor. In fact, the GPL’s always had that little problem with defining who exactly is granting the contract, originally addressed by the suggestion to attribute copyright to the Free Software Foundation (since, in theory, anyone writing GPLware can claim agency with the FSF, except for the part where they forgot to explain that, so people applied the license while asserting personal copyright as well)… I think that idea fell off the back of the cart. Something like “All those whose names appear hereby grant you…” could solve that, if only they’d tried.

Wow, back up to point 2, which refers in singular to ‘the term of copyright on the Program,’ while all ‘work’ added to the Program originates with copyright from that work’s author (in the US, though it can’t hurt to at least be consistent with the law of the land this beast is assembled in). There’s got to be a precedent for determining term on collaborative… um, “works”… Maybe they’re purposefully vague to stay applicable internationally, but the sparks are going to fly within 120 years or so, or whenever some party decides to assert that the work’s fallen into the public domain and a recent modifier begs to differ. *NIXes, like cockroaches, will still be around in 120 years, right? ;)

7e: The â€œofficial-mozilla-brandingâ€ clause, wherein the DMCA can’t restrict your right to use what you’ve been handed, but trademark law still can. Great. What was once only implied by external statute is now codified. Nothing to complain about legally, but if going this far, codifying the Mozilla compromise (terms added to restrict trademarks permissible *if* the source code as conveyed can produce functioning object code not using them) would have been one less practical headache. Look forward to Cisco/Linksys, Novell, etc plastering their trademarks all over the place, maybe even some schmuck utilities to insert code that combines to display a trademark in as many disparate locations as possible. (Easy to clean out? What if it’s put in with 1,000 time redundancy…?)

That’s a lot of disorganized nitpicking, and the sun’s coming up as I write this. Still, it goes to show that even I, entirely in favor of the license’s intentions, find it a mindf*** to actually work with. The Creative Commons tangle is starting to look simple by comparison.

Lots of Lessig logic in this one (well-intentioned, well-reasoned to a point, but disjointed and overextended in the end), not so much Stallman (recursive, obscure, yet internally consistent). Maybe it’s the hurd-of-lawyers effect at work; everyone gets their pet terms in, at which point it would be unconscionably rude for any one person to clean up (or pare down) the whole. After all, if you offend any large corporations who dared contribute, they might pack up and leave for a different bazaar.
Joe "Floid" Kanowitz says:

2007/07/01 at 04:47

Kill me, I used an emoticon again.
Captain Sonic says:

2007/07/01 at 05:58

What a horrifying logo….
Pingback: GPL v3 analysis here · DragonFly BSD Digest

Comments are closed.