On Mon, 24 Jun 2024 at 14:44:18 +0100, Nathan Willis wrote: > And those factors would need to interact predictably with a specification > document that is free to read, implement, and share ... but the specification > should not be forked or modified (since that would defeat the purpose: > interoperability).
I don't think enforcing "no derivative works" via copyright law (like Creative Commons CC-ND) is the right way to ensure interoperability. Instead, I think the right way is to require modified versions to be clearly marked as not being the original - more like the approach taken by Creative Commons CC-BY. Consider this scenario: you publish your standard, version 1.0; time passes; you disappear and cannot be contacted any more; more time passes; now I want to produce a compatible version 1.1 of the standard containing clarifications or extensions, or an incompatible version 2.0 with significant changes (like HTTP -> HTTP2), or correcting a design mistake that cannot be fixed without a compatibility break. I think the legal situation that the community would want in that scenario is that I can modify and share draft copies of what I want to put in version 1.1 or 2.0, or even fork the specification to make my own intentionally incompatible specification under a different name if I really need to, but I have to label them as something clearly different (unless/until my changes have been agreed on and accepted by whatever is the most appropriate standards body that controls the "branding" on your behalf). A document that enforces "no derivative works" via copyright law is also likely to be forbidden from inclusion in Debian, because we require that everything we ship is Free Software (not just code, like e.g. Fedora, but also non-code like data and documentation, unlike Fedora). And if copy/pasting spec text into Free Software source code is a use-case for you, that is not a context where modifications can be disallowed, because if they were, it wouldn't be Free Software. Perhaps consider using <https://datatracker.ietf.org/doc/html/rfc5215#section-11>, the legal terms of the RFC that describes Vorbis-over-RTP: The authors agree to grant third parties the irrevocable right to copy, use, and distribute the work, with or without modification, in any medium, without royalty, **provided that, unless separate permission is granted, redistributed modified works do not contain misleading author, version, name of work, or endorsement information.** (my emphasis) If you want stronger protection against passing off an incompatible not-quite-Opentype-Shaping specification as being genuine Opentype Shaping, then trademarks rather than copyright are probably a better tool: not allowing "counterfeits" to harm your reputation or inappropriately benefit from your reputation is literally the purpose of trademarks! smcv
