Open Stack

On 2017-02-23 18:37:49 +0000 (+0000), Cathy Zhang wrote:
> I strongly support the proposal to mandate this. To be fair, I
> think TC should mandate this across all projects. In many
> complicated and technically hard commits, co-author does not make
> any less amount of technical contribution to the commit. If just
> the owner is counted, people will start to fight for the ownership
> of a commit which is not healthy for the open source community.
> 
> For my own case, it is well known that I am the initiator and
> project lead of this networking-sfc project and have contributed a
> lot to this project on the technical side and project management
> side. I have done many reviews and approvals in this cycle and
> co-authored quite some commits. It is a surprise to me that
> co-author is not counted as technical contributor in Neutron.

The technical limitations for this in the past have been twofold:

1. Gerrit did not provide a usable API for querying arbitrary
substrings from commit messages.

2. Voters must be foundation individual members and we had no way to
query the foundation member database by contributor E-mail address.

The first is less of an issue in the version of Gerrit we're running
now and the second is a situation I'm collaborating with the
foundation's development team to attempt to resolve. In the
meantime, the solution has been that PTLs should entertain requests
from co-authors to be added to the "extra ATCs" list for their
project. I don't personally have any objection to letting change
co-authors vote in elections, we just don't (yet) have a solution to
be able to automatically verify whether they're authorized to vote
under our bylaws and charter.

Separately, there was a problem back when we used to provide free
conference passes to code contributors, where someone at a company
would submit a punctuation fix to a comment in some project, add
half a dozen of their co-workers as co-authors, and then ask for
free admission for all of them (this really happened). Relying on
PTLs to vet extra ATCs before adding them was how we mitigated this.
Now that we no longer rely directly on code contributions to decide
who should get free/discounted conference admission this issue
should hopefully be purely historical. People seem to be far less
interested in gaming elections than going to conferences (or in some
cases scalping free tickets as a money-making scheme).
-- 
Jeremy Stanley