--- Comment #25 from Steven Walling <swall...@wikimedia.org> ---
(In reply to comment #24)
> This is actually quite important. Bunching different changes together into
> same experiment can easily invalidate the usefulness of the results, and is
> certainly likely to do so here - because then you have no idea which part was
> the part that people actually objected to, or which parts were the ones that
> really made things better. You're left with only speculation and a few
> specific reports such as this one that are, it seems, easily disregarded.
I think you and MZ are a bit confused about what kind of experiment this is.
This is not an A/B test, and we're not opting anyone in to this by default.
We're also not collecting *any* quantitative data beyond how many people have
opted in to a particular beta feature.
All of the feedback is qualitative via the Talk page, etc. so it's perfectly
fine to bundle many changes together and let people comment about what they
like/don't like. That's what's happening, if you look at Talk:Typography
refresh on mediawiki.org.
It doesn't really matter how many times you want to reopen this bug. We're not
going to roll back a part of the beta that we literally just launched days ago.
We're going to collect feedback on it for at least a week or two, then meet
again to compile all perspectives on all the latest changes. At that point
we'll consider what to retain as beta, what to graduate to stable, and what to
throw out. The best thing you can do, if you don't like something about the
current release, is to express clearly and calmly why you don't like it on the
Talk page. Playing tug of war on Bugzilla is not going to do anything except
waste our time.
You are receiving this mail because:
You are the assignee for the bug.
You are on the CC list for the bug.
Wikibugs-l mailing list