+++ Burt_Harris [Sep 27 14 14:54 ]:
In my mind, this is not about applying some legal test for a derivative work, because accepted the legal tests change over time. This is more a matter of intellectual honesty.
Intellectual honesty requires giving credit for ideas you use. We have done that in the normal way, acknowledging John Gruber as the creator of Markdown, citing his syntax description, etc.
Licenses and copyright concern specific legal issues. So, here it is precisely a legal question, whether our work is “derivative” from Gruber’s.
Copyrights protect not ideas, but expressions of ideas. The idea of a lightweight markup language is not protected by Gruber’s copyright assertion, but the way he expressed the idea is. That form of expression expression included the concrete syntax of Markdown, which is common to both the prose and Perl he released under 2004 under license.
I would think that the “idea” is not that of “a lightweight markup language” (which is certainly not Gruber’s idea), but Markdown syntax – the specific combination of conventions for indicating structure in plain text. Gruber’s expression of the idea is the Perl script and the documentation. So the relevant question is whether our work is derivative from these. I think it is not. The fact that our work employs the same conventions for plain-text formatting is not sufficient to show that it is derivative from Gruber’s expressions of the idea, since this just reflects that fact that they are expressions of the same idea.
Consider this analogy. A mathematician, Smith, publishes a paper describing a new kind of mathemical structure that hasn’t been studied before – call it a “commutative foobar group” – by putting together some properties or axioms. (“A commutative foobar group is a commutative group that has properties X, Y, and Z.”) Other mathematicians can write papers about commutative foobar groups, and (by the norms of academic discourse) they should cite Smith for the idea. It would be plagiarism for someone else to claim to present commutative foobar groups as if they were their own original idea. However, nobody would claim that, just by discussing commutative foobar groups, proving things about them, and discussing variants like non-commutative foobar groups, their papers infringe Smith’s copyright.
Another interesting analogy is with typefaces, which aren’t copyrightable in the US. Fonts, which are essentially computer programs for rendering letters according to a typeface, are copyrightable, in the same way as other computer programs. But the abstract letter shapes that make up a typeface are not. Apparently the legal basis for this has to do with their “utilitarian purpose”; if they had been ruled to be works of art, they would have been copyrightable. It seems to me that Markdown syntax clearly has utilitarian purpose, and would naturally fall into the same category as typefaces.
Note also that most or all of the elements of Markdown syntax were already in wide use as conventions for formatting plain-text emails and contributions to usenet groups. (A few others came from earlier light markup systems, such as Setext and ATX.)
I’m not a lawyer, and I’m guessing you aren’t either. And the law has enough hard-to-understand rough edges that things may be more complicated than they seem to me. But I don’t think the license was ever the real issue. John Gruber didn’t like the idea of a group representing itself as determining a canonical syntax for Markdown, because he wants to be the one to do that. I understand that desire perfectly, though I wish he had been intersted in collaborating in or leading an effort of this kind himself. In my correspondence with him the name was the central issue, and he was satisfied when we changed it.