I am aware of some of the problems of the concept of "releasing a work to the public domain" (as mentioned here and here; I haven't listened to this specific discussion yet). IANAL, but the main problems seem to be as follows:
- Releasing to the public domain doesn't exist in many jurisdictions; rather, a work becomes public domain after copyright expires.
- Some moral rights are inalienable due to (certain ratifications of) article 6bis of the Berne Convention.
However, aren't you still allowed to publish a work anonymously? If you anonymously publish a work, in such a way that authorship can't be proven or even distinguished, wouldn't that effectively "unlicense" your code? I would think so: since then any claim of authorship can't be validated, the rights implied by authorship can't be enforced, not even by the original author.
If so, I'd really like to know which permissive license (ISC/MIT/BSD-style) would be fit for such publications. (Especially when they've managed to get rid of the lengthy
THIS SOFTWARE IS PROVIDED IN "SCREAMING CAPS"-style liability waiver.) If no such license exists yet, how would one modify an existing license to be suitable for an anonymous publication?