[Openstack] RFC: Thoughts on improving OpenStack GIT commit practice/history

Daniel P. Berrange berrange at redhat.com
Thu Jun 28 17:17:10 UTC 2012


On Thu, Jun 28, 2012 at 09:21:20AM -0700, Johannes Erdfelt wrote:
> First off, I wanted to say I think these are a great set of
> recommendations.
> 
> On Wed, Jun 27, 2012, Daniel P. Berrange <berrange at redhat.com> wrote:
> >     Fixes: bug #1003373
> >     Implements: blueprint libvirt-xml-cpu-model
> >     Change-Id: I4946a16d27f712ae2adf8441ce78e6c0bb0bb657
> >     Signed-off-by: Daniel P. Berrange <berrange at redhat.com>
> >
> > As well as the 'Signed-off-by' tag, there are various other ad-hoc
> > tags that can be used to credit other people involved in a patch
> > who aren't the author.
> 
> What is the Signed-off-by tag used for?
> 
> Your examples have yourself, but isn't that kind of implied by
> submitting the patch for review in the first place?

Yes, you are technically correct in this respect.

It is an idiom originating from the Linux Kernel community, which is
auto-added by GIT if you pass the '-s' arg to 'git commit'. Basically it
is a statement that you have read & are complying with the projects
contributor guidelines (eg "Developer's Certificate of Origin"[1] in
the kernel), or a more formal contributor license agreement such as
that used by OpenStack.

OpenStack obviously has a formal CLA which all contributors *must* agree
with prior to submitting patches, which serves the same purpose. Thus
using a Signed-off-by: line is pretty much redundant for any contributions
to OpenStack, since you must have signed the OpenStack CLA before you
can even get access to post patches to Gerrit.

The only case where I see that it might be considered relevant, is if
the person submitting the patch to OpenStack, is not the same as the
person who wrote the patch. For example, if someone in Red Hat's QA
team (who isn't an OpenStack contributor) writes a patch for OpenStack
& gives it to me, then they'd typically include their own email addr
in a 'Signed-off-by' tag to indicate that they are the author & they
understand the contribution requirements of OpenStack. This indicates
to me that I can trust their patch & thus I'd be happy to add my own
email 'signed-off-by' line & submit it to OpenStack in my role as
someone who has agreed to the formal CLA.

Since this tagging is a standard feature of GIT, it is quite typical
for people to add Signed-off-by tags on all their commits, to any
project, regardless of whether the project actually mandates this
as their submission policy. I certainly just do it out of habit
for all projects I contribute to.

So in summary, you are perfectly ok to just ignore the whole
Signed-off-by concept in OpenStack, given the formal CLA reqired
for contributors already.

Regards,
Daniel

[1] https://www.kernel.org/doc/Documentation/SubmittingPatches
-- 
|: http://berrange.com      -o-    http://www.flickr.com/photos/dberrange/ :|
|: http://libvirt.org              -o-             http://virt-manager.org :|
|: http://autobuild.org       -o-         http://search.cpan.org/~danberr/ :|
|: http://entangle-photo.org       -o-       http://live.gnome.org/gtk-vnc :|




More information about the Openstack mailing list