On Thu, Sep 1, 2016 at 6:54 PM, bawolff <bawolff+wn(a)gmail.com> wrote:
* In a big session on services-oriented architectures, a lot of time was
spent theorizing about what small wikis who do their hosting on
shared-hosting services do, and whether various solutions we were
proposing
would make it easier or harder for these non-WMF
users of mediawiki.
*But
none of these users were at the summit.* So no
decisions could
ultimately
be made, as the necessary affected parties were
not present.
I don't think that would change, regardless of what we do. Even if we
have more users at the summit, its not going to be a representative
sample of every type of user we have. I don't think its reasonable to
assume that just because a usecase isn't represented at the summit
that it doesn't exist. Anyone taking the time out of their day to
travel all the way to a MediaWiki event, is probably a power user, and
thus this will bias the representation of users at the summit.
Quite possibly! And thus perhaps we should just exclude those sort of
topics from the summit -- if we can't get representative participation, we
shouldn't have a decision-oriented summit session.
Sessions oriented around "learning from our community" might still work --
we could have invited a cross-section of shared hosting users for a
workshop where we gathered info about different hosting providers, walked
them through installs using vagrant (or what-have-you), listened to their
concerns & problems, and worked together with them to figure out what might
work. At the end, we can't say "this will work for all users of shared
hosts" but at least we can say, "for the 15 people we worked with, here are
the ways they managed to install a services-oriented prototype of
mediawiki" or something like that.
"Community workshop" is an underexplored dev summit format. But maybe that
would be a better fit at wikimania?
--scott