[PATCH 0/2] Use SPDX-License-Identifier for rockchip devicetree files

[Heiko Stübner]

Am Freitag, 15. Dezember 2017, 15:42:48 CET schrieb Philippe Ombredanne:
> On Fri, Dec 15, 2017 at 3:28 PM, Heiko Stübner <heiko at sntech.de> wrote:
> > Am Freitag, 15. Dezember 2017, 14:45:34 CET schrieb Philippe Ombredanne:
> >> Klaus,
> >> 
> >> On Fri, Dec 15, 2017 at 12:44 PM, Klaus Goger
> >> 
> >> <klaus.goger at theobroma-systems.com> wrote:
> >> > This patch series replaces all the license text in rockchip devicetree
> >> > files text with a proper SPDX-License-Identifier.
> >> > It follows the guidelines submitted[1] by Thomas Gleixner that are not
> >> > yet merged.
> >> > 
> >> > These series also fixes the issue with contradicting statements in most
> >> > licenses. The introduction text claims to be GPL or X11[2] but the
> >> > following verbatim copy of the license is actually a MIT[3] license.
> >> > The X11 license includes a advertise clause and trademark information
> >> > related to the X Consortium. As these X Consortium specfic points are
> >> > irrelevant for us we stick with the actuall license text.
> >> > 
> >> > [1] https://patchwork.kernel.org/patch/10091607/
> >> > [2] https://spdx.org/licenses/X11.html
> >> > [3] https://spdx.org/licenses/MIT.html
> >> 
> >> FWIW, the X11 license name was not always something clearly defined.
> >> SPDX calls it clearly MIT which is the most widely accepted name for
> >> the corresponding text. And this is also what we have in Thomas doc
> >> patches that should be the kernel reference.
> >> 
> >> Also, as a general note, you want to make sure that such as patch set
> >> is not merged by mistake until you have collected an explicit review
> >> or ack from all the copyright holders involved.
> > 
> > Just for my understanding, is it really necessary to get Acks from _all_
> > previous contributors?
> > 
> > I see that Thomas patches moving license texts into the kernel itself do
> > not seem to have landed yet, but when the actual license text does _not_
> > change and only its location to a common place inside the kernel sources,
> > it feels a bit overkill trying to get Acks from _everybody_ that
> > contributed to Rockchip devicetrees for the last 4 years.
> > 
> > If we would actually want to change the license I would definitly feel
> > differently, but the license text does not change.
> 
> Well you are technically right. But there is a social and politeness
> angle to this too. So may be getting the ack of all contributors is
> not always needed, but getting it is best and the right to do and at
> least getting for the named copyright holders should be there.
> 
> That's only only my take: leaving aside any technical legal issue, say
> I would be on the receiving end as one of the holder or contributors:
> I would find it really great and nice to have my ack requested. And I
> would be a dork not to give it. So I like to do to others the same I
> would appreciate done to me (within reason, as I sometimes shoot
> myself in the foot ;) )

Hehe ... I didn't plan on merging this without ample time for people
to either ACK or NAK the change, so was planning on keeping to social
protocol ;-) . Just the "all" threw me for a loop.

And having that as PATCH without RFC also communicates that people
should take a look, as RFC patches are often overlooked.

As Klaus seems to have included most people that have contributed in the
past, I would guess we should receive any existing complaints about that
change :-) .

So I'll definitly let this simmer for quite a bit and do a best-effort Ack 
collection.


Thanks
Heiko