ESR has opined that those who have been making the argument that copyright-holding linux-kernel-programmers have the right to rescind their licenses regarding linux-kernel code they own; have a good factual case, but need help with public relations. The lengthy explanations come off as "obsessive" to the public, but how to explain the foundations of the law and how such relates to the fact-pattern succinctly?
>ESR: The post a few posts above which starts with “Subject Threats of “blackballing” from industry if copyright-holders rescind.” has the links at the end.
Then my thoughts are
(1) It didn’t belong in this post threat.
(2) I think you have a decent factual case. Now, if you could just stop sounding enough like an obsessed crank to be easily written off, you might get somewhere with it.
The "Rescind" discussion has arisen after the implementation of a Code of Conduct, which threatens to penalize gratis programmers for utterances they make anywhere, if deemed offensive. Effectively treating contributors as utterly replaceable employees++. The crux of the matter being that many contributors never signed over ownership of their code, and did not ever receive nor seek any form of payment for their licensing of their code under the GPL version 2 terms; making the license grant a gratuitous license: which under US licensing law are revocable not being supported by an interest, according to the "Recind"(ers).
++ (See statements by Matthew Garret and Bruce Perens to such effect)
Most large projects require contributors to sign over ownership of the copyrights to the project. The FSF has always followed this policy, to avoid contributor rescission, the "Rescind"(ers) contend.
The "lol no" contingent argues that: Indeed contributors are utterly replaceable, and the "Rescind"(ers) are simply wrong because opensource simply wouldn't work if that were the case. They argue that all software licenses are non-revocable because the paralegal "PJ" said so on grok-law. The "lol no" contingent (including the SFConservancy) also claims that a promise exists within the GPLv2** that the grantor will never revoke the license and thus argue that promissory estoppel applies.The "lol no" contingent also says that anyone who tried to rescind would be blackballed from the industry.
**The "Rescind"(ers) note that the cited clause simply declares that should an upstream-licensee have his license automatically revoked due to a violation of the license by the upstream-licensee, the down-stream licensee' license is not in-turn automatically revoked. "Recind"(ers) maintain that this clearly is not a promise by the grantor of irrevocably, amongst other things. The "Recind"(ers) also note that promissory estoppel was created to secure grants to charities and for situations where a person who was promised a transfer of land then invested to improve the land based on that specific promise of transfer. The "Recind"(ers) contend that even in a theoretical case of explicit no-revocation-clause, because the promise is not supported by any consideration, and there is no meeting-of-the-minds, the grantor cannot be held to that supposed promise. The most that could happen, the "Recind"(ers) contend, if there was an unsupported promise not to revoke, is that the licensee may be allowed to continue to use whatever specific version of the Project that contains the rescinded code, that he is currently using, but would not be able to add the code to future version of the Project. The "Recind"(ers), however, contend that there simply is no such promise within the GPLv2 so the issue is moot.
The "Recind"(ers) note that the GPLv3 contains an explicit no-revocation-by-grantor clause, and contend that this was added to adapt to the way the GPLv2 was being used in the wild; and stands as an implicit acknowledgement by the drafters regarding the inadequacies of the GPLv2 when used without a contributor assignment agreement .
(Note: The linux kernel code is licensed under the GPLv2, and does not contain the familiar "and any later version" clause)