[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [Repo-criteria-discuss] B2 for Gitlab?
From: |
Aaron Wolf |
Subject: |
Re: [Repo-criteria-discuss] B2 for Gitlab? |
Date: |
Sun, 1 May 2016 10:56:49 -0700 |
On 05/01/2016 10:39 AM, Mike Gerwitz wrote:
> On Sun, May 01, 2016 at 18:46:47 +0200, Sytse 'Sid' Sijbrandij wrote:
>> Personally I think that just a license file in the project can be
>> sufficient if you use version control that keeps the files together
>> and discoverable.
>
> It's easy for a source file to become separated from others: files may
> be copied to another project, for example, or may simply be separated on
> a filesystem or website. If that file does not have its own copyright
> and license header, then it's origin and license will be incorrectly
> assumed to be that of the project that received it; if there is no
> containing project, then it's in a state of licensing limbo: effectively
> proprietary unless its origin can be discerned (works are proprietary by
> default unless rights are granted back by the author, so the absence of
> such a notice means all rights are reserved).
>
I'm not an authority to make conclusions here, but I don't get this
logic. The same lack-of-license could happen with copy/paste of content
from a file even when the header is included. It seems the only
reasonable view here is to say that files getting separated is more
likely than copy/paste (though I don't know if that's true, I'll accept
it seems a good guess), and the extra layer of caution is nice.
> A project might incorporate code from another into a single source file;
> in this case, the license of the file must clearly state dual licensing,
> unless they can be relicensed (e.g. Expat -> GPL).
>
That's a more interesting point. Of course, a header could be added just
for those cases. Incidentally, how does one determine which parts of a
file are under which license?
signature.asc
Description: OpenPGP digital signature