Second point is identifying the source of every commit. IMHO it is good practice to have a sign-off for ALL initial commits. Justt to make sure we have a full trail. Important for example for EAL certification. But also in case of disputes. Jan > On 10/18/2011 05:31 PM, Richard Fontana wrote: >> On Tue, Oct 18, 2011 at 01:41:04PM -0500, Anthony Liguori wrote: >>> > On 10/18/2011 12:35 PM, Carl Trieloff wrote: >>>> > > >>>> > >I would like to add to the site some text around applying the >>>> > >copyright >>>> > >statement per the license text as per >>>> > >http://www.ovirt.org/about/licensing/ >>>> > > >>>> > >In the piece at the top of the license, there is a section >>>> > > >>>> > >"Copyright [yyyy] [name of copyright owner]" >>> > >>> > This will remain a recommendation, right? It would be difficult for >>> > us to get approval to use something other than the IBM corporate >>> > standard copyright notice (which does change every couple years). >>> > It seems (C) comes in and out of style over time :-) >> That line is actually from the appendix of the Apache License 2.0, >> which gives a recommended copyright notice. >> >> I'm sympathetic to the view that precise formal details of copyright >> notices should not be mandated. At Red Hat, I think the "(C)" may have >> been historically recommended, but I have non-actively promoted >> non-use of it on principles of minimalism :-). >> >> I think, however, that Carl was suggesting a rule that *some* suitable >> form of copyright notice be placed by the author of an original file, >> not mandating what the form would be. > > correct. > > Carl. > I personally rather dropping the [name of copyright owner] I believe that's why we have history in the source control. There is one person adding the file and other X (X likely > 1) people changing it, I see no reason crediting the one who added the file. We had a discussion on this in the engine-core project some time ago and we decided to remove all credits to file creators. If not accepted as a general rule for all oVirt projects I would like to suggest that this can be decided/changed on a project level. Livnat _______________________________________________ Board mailing list Board(a)ovirt.org http://lists.ovirt.org/mailman/listinfo/board _______________________________________________ Board mailing list Board(a)ovirt.org http://lists.ovirt.org/mailman/listinfo/board

It sounds as if you are suggesting we assign copyright to oVirt, which doesn't make sense to me. oVirt isn't a legal entity that can hold a copyright.

I was following the Apache sample you gave. If ovirt.org is not a legal entity for intellectual property matters, and contributer does not belong to a company we'll get to a state which every file in the project will be copyrighted to a different person. This is too chaotic.

* Doron Fediuck (dfediuck(a)redhat.com) wrote: > I was following the Apache sample you gave. If ovirt.org is not a legal entity > for intellectual property matters, and contributer does not belong to a company > we'll get to a state which every file in the project will be copyrighted to > a different person. This is too chaotic. It is the norm for many open source projects. Diversity is healthy.

thanks, -chris _______________________________________________ Board mailing list Board(a)ovirt.org http://lists.ovirt.org/mailman/listinfo/board

On Tue, Oct 25, 2011 at 10:26:11AM +0200, Doron Fediuck wrote: > > I was following the Apache sample you gave. If ovirt.org is not a legal entity > for intellectual property matters, and contributer does not belong to a company > we'll get to a state which every file in the project will be copyrighted to > a different person. This is too chaotic. I'm confused about what the real issue is here. I haven't honestly studied a lot of FOSS source files, just when I needed to. The copyright and license header is there in a comment field, I skip by, and start hunting for whatever I need. Even when the copyright is several lines long. Having these details in the files and in the git record is important for when we need to contact copyright holders for any reason. The most important thing AIUI is that the files are _licensed_ properly. As long as the files are ASL, we can use them in the project, regardless of the copyright holder.

As someone else said in this list, how FOSS developers write their copyright notice is often a very personal thing. Some use their @work email, some their @personal email, some their @hackerspace email, and so forth. Since this is important to people and exercising that right does no real harm to the source files, it's best to leave the final decision up to people as to how they mark it. - Karsten _______________________________________________ Board mailing list Board(a)ovirt.org http://lists.ovirt.org/mailman/listinfo/board

On 10/25/2011 08:21 PM, Karsten Wade wrote: > On Tue, Oct 25, 2011 at 10:26:11AM +0200, Doron Fediuck wrote: >> >> I was following the Apache sample you gave. If ovirt.org is not a legal entity >> for intellectual property matters, and contributer does not belong to a company >> we'll get to a state which every file in the project will be copyrighted to >> a different person. This is too chaotic. > > I'm confused about what the real issue is here. > > I haven't honestly studied a lot of FOSS source files, just when I > needed to. The copyright and license header is there in a comment > field, I skip by, and start hunting for whatever I need. Even when the > copyright is several lines long. > > Having these details in the files and in the git record is important > for when we need to contact copyright holders for any reason. > > The most important thing AIUI is that the files are _licensed_ > properly. As long as the files are ASL, we can use them in the > project, regardless of the copyright holder. > I started the discussion about credits to file's author and it took a little turn, assuming no legal issue, my original request was - I rather not have the author name on the file. The main reason for that is that file changes over time and after a while it has little to do with the original version. There are many contributors to the changes and by having one name on the file we "hide" the contribution of other developers.