[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [GNUnet-developers] question on copyrights in gnURL / cURL
From: |
ng0 |
Subject: |
Re: [GNUnet-developers] question on copyrights in gnURL / cURL |
Date: |
Wed, 13 Sep 2017 19:13:55 +0000 |
Christian Grothoff transcribed 4.8K bytes:
> On 09/13/2017 08:32 PM, ng0 wrote:
> > No matter if we end up with wget2 or stick with gnURL,
> > the code on gnURL is out there. However since I picked
> > up its maintenance I touched many many files, but only
> > added one copyright header in a new file.
>
> That should be fine.
>
> > FSF licensing is a small team, so maybe we have someone
> > here with a little bit of expertise: I think I need to
> > add copyright header (or at the very least a name+email
> > and a way to display that this file was changed within
> > gnURL) by now.
>
> I don't see why, and while I am not a lawyer, I'm pretty sure what I say
> here and below is sound. cURL itself does not credit every contributor
> in every file, they list them in docs/THANKS. GNU says that every
> source file should have a copyright header, but the cURL ones do
> already, so no need to change that. You said you added one for a new
> file, so great, you're done with that!
> > With the license cURL uses:
> >
> > Would I be allowed to just grep the "cURL" in the header
> > of all files in cURL or simply within the files I touched?
>
> I would leave the cURL license headers completely untouched,
> even if you touched the files. It just muddles the diff.
>
> > And: is it required at all? I have tasks to do which are
> > way more fun (and necessary) than git grep + sed'ing through
> > an ever-growing fork.
>
> Exactly, I see no reason why you should do this.
>
> > For example neomutt, an effort to add new features and
> > more programmatical changes to mutt has kept all the
> > original headers and added no copyright notes (no header name
> > changes, no copyright lines, etc) (GPL2 licenses iirc).
> > One could argue where a copyright'able contribution
> > begins.
>
> But copyright*able* does not mean that you must claim it, it just means
> you _could_. And attribution is again separate, under the cURL license I
> don't see that you'd even have a right to require attribution, and you
> certainly have no duty to provide it in any particular form either.
> Besides, Git is good enough for that for anyone who really wants to know.
>
> > I just have a questionmark on the amount of
> > changes I have in gnURL compared to cURL based on the
> > git commits and if it requires any statement somewhere
> > (not everyone reads git logs).
> >
> > At its core all the changes I made are with the build system
> > of cURL and how it behaves, renamed files, renamed documents,
> > renamed occurences, references of files, and some more
> > renames and fixes still pending.
>
> Sure, but none of this requires you to make any further changes just
> because of copyright.
>
This matches up with what I assumed and partly knew.
Thanks for taking the time to clear it up!
--
ng0
GnuPG: A88C8ADD129828D7EAC02E52E22F9BBFEE348588
GnuPG: https://krosos.org/dist/keys/
https://www.infotropique.org https://www.krosos.org
signature.asc
Description: PGP signature