Brad Jorsch (Anomie) wrote:
On Wed, Jan 18, 2017 at 10:44 PM, Daniel Kinzler <
daniel.kinzler(a)wikimedia.de> wrote:
** Related discussion about whether new features
can require services
serparate from MediaWiki core.
That seems like it would be a decent RFC at some point.
Agreed.
There are often questions of debug and maintenance support and service
level for production services, from the Wikimedia (Foundation) operations
team and others. Clarifying the current situation and how it can be
modified in the future would be useful and alleviate some of the tensions
we've had, in my opinion.
The trickiness here is that some of this falls outside of the Architecture
committee's purview, particularly as no member of the operations team is
on the committee currently (Mark, Faidon, Alexandros, Giuseppe, et al.).
In my mind, there's what MediaWiki core and its extensions can do and can
require, but the relationship between MediaWiki and Wikimedia cannot be
completely sidestepped. With MediaWiki as the platform, there are also
questions about what
wikimedia.org,
mediawiki.org,
wikipedia.org, and the
various other Wikimedia projects are willing to host and support. RESTBase
seems like a decent example of this interplay.
Daniel Kinzler wrote:
We have to decide on a MO. I'm sharing this
summary here to keep you
posted, and give an opportunity for input. But in the end, the committee
has to decide how it wants to operate.
"You know what they call a leader with no followers? Just a guy taking a
walk."
In the past, the IRC dicussion was the only way to get
an RFC discussed
or approved.
It depends how strictly you view requests for comments. A decent-size
Gerrit changeset that gets accepted and merged in is, in some ways, an
RFC, usually with one or more associated Phabricator Maniphest tasks.
Sometimes with associated mailing list or on-wiki discussion.
MZMcBride