Re: DEP5: Public domain works
Jonas Smedegaard <firstname.lastname@example.org> writes:
> Here's how I currently treat PD in Debian packaging:
> Copyright statement + "is in the Public Domain": DFSG-nonfree, because
> either a) it really is in the Public Domain and then cannot at the same
> time be upheld by a copyright, or b) is not really in the Public Domain
> and then lack a free license.
I think it's excessively picky to treat this as DFSG-nonfree. Unless
there's some clear reason to believe that it's going to be a problem, I
treat code released by the author with a statement "this is in the public
domain" as if it has an all-permissive license.
Of course, as opportunity arises, I try to clarify that with the original
author and ask them to use a better license. But of all the things I'm
worried about when it comes to free software licenses, treating code with
an explicit statement that it's in the public domain (that was added by
the original author, not by some "helpful" third party) as if it were
released under a permissive license isn't high on the list.
> If not, what is bad in marking as "PD" and leaving to others to
> investigate if it really is a True Public Domain or just a bogus
> upstream claim? I mean - aren't we generally trusting upstream
I don't generally trust upstream statements that something is in the
public domain because they're wrong 90% of the time, which is unusually
high for upstream statements about licensing. It's much less likely that
upstream will be wrong when they declare the license to be an MIT or
2-clause BSD license. But misunderstanding of public domain is
Hell, I have stuff of my own that's distributed under a license that says
"this is public domain" that I need to go back and fix, and I'm relatively
clued about this sort of thing.
Russ Allbery (email@example.com) <http://www.eyrie.org/~eagle/>