Hi!
I agree. I do think that as a community of practice we
have many
unwritten rules and numerous expectations of how we work together. We
don't explicitly define the expectation of a README.MD file in repos
either.[0] It's a best practice and cultural expectation in our spaces
to include one. The code works the same with or with out it.
I think there's several aspects here to consider:
a) In WMF technical spaces, and more widely, in the Mediawiki/Wikimedia
universe, I think there's universally acknowledged expectation of
certain standards of behavior, which in the Wikimedia space have been
codified in the CoC. The purpose of these expectations, as I understand
them, is to build and maintain an open, welcoming, productive and
inspiring community that would support development of Mediawiki and
Wikimedia projects. And the CoC is the instrument that we chose to
codify and implement those expectations in the Wikimedia spaces, which
applies to all of them regardless of technical means chosen to publish
or document it.
I do not think there is much disagreement about that.
b) How exactly the spaces are managed within this wide framework has a
lot of complex and tricky details. Some of which may seem trivial to
some people and highly sensitive to other people. Including which files
are placed in which repositories, who is allowed to change which
repository and for which reasons and procedures, and so on. I think
having more clear expectations on that would certainly help.
But beyond that, I think when designing and enforcing the rules for
these minute details, we should not lose the sight of why it is done,
and not make the process of CoC enforcement go against the goal of
having CoC - namely, the welcoming community. If that means sometimes
being more flexible, or having a bit more patient discussion and
resisting the urge to force your point through, even if I am completely
sure I am correct, I think it is still worth it in the long run.
c) Specifically about CoC.md file, I personally think having redundant
pointers to the documentation (both technical and about societal norms)
is highly welcome, as locating proper docs is notoriously hard and
largely unsolved problem with most code. Having the docs is half of the
problem (which we also sometimes fail at ;), having it where people
would find them is the other half. So adding of the CoC file from this
point of view is a smart move.
On the other hand, maintaining a rigid "one size fits all, no
exceptions, no discussions, shut up and comply" approach to it feels a
bit counter-productive to me. Yes, I foresee the question "if it is a
good thing, why not make everybody do it?" - and I could probably easily
write a 20-page essay on this topic, but I would limit myself here to
this - people have different points of view, and I think being more
accommodating in this case is better than having a nice set of
checkboxes checked.
What it specifically means for the specific file? I admit I don't have a
better proposal than "let's have a community discussion on it". But I
think making an open and welcoming community including sometimes being
patient in figuring out how exactly to do it. Enforcing having the file
in every single repo does not seem to be a pressing concern that would
do any harm if not brought into compliance right now. So let's see if we
can reach some consensus here.
--
Stas Malyshev
smalyshev(a)wikimedia.org