Federico Leva (Nemo) wrote:
MZMcBride, 25/10/2014 16:16:
But again, the focus would be
integrating into the Wikimedia technical platform and fixing issues in
production, rather than trying to make Labs scripts and tools better.
False dichotomy IMHO. Usual example:*
https://bugzilla.wikimedia.org/42259 Quite clearly WMF's responsibility
to make it possible, but all the interfaces and value are in
consumers/tools.
Nemo
(*) Yes, I know
https://meta.wikimedia.org/wiki/Research:Page_view is
being worked on.
I think it's a limited view to suggest that the Wikimedia Foundation
should only provide raw dumps and/or queryable data and have volunteers
try to cobble together scripts and tools to interact with the data. That
certainly can and should be a piece of this, but there's no good reason
not to, for example, integrate page view graphs into MediaWiki's info
action, allowing regular users to see quickly and easily see an article's
page views over time.
We already have queryable revision information, but we rely on external
tools and services to try to graph edits over time, rather than having
visual functionality integrated into MediaWiki. The same is true of
visualizing a particular user's edits or other logged actions. We're
relying on external tools when we should be trying to create tools that
live within the technical platform that we've created. A generalized,
scaleable graphing/visualization tool would be an excellent use of
resources. Making such a tool could easily have a definable goal with
clear requirements, and implementing and deploying such a tool would have
a very clear benefit to all of our projects.
We have a real problem turning proofs of concept into stable
infrastructure. GLAMs and Chapters can invest in creating stable
infrastructure, but that probably doesn't mean investing in Labs, exactly.
Not if you want to have a long-term, substantive impact, in my opinion.
Regarding page views data specifically, the Wikimedia Foundation has done
a good deal of cookie-licking in this area and that really should be
addressed. The linked bug report demonstrates what I'm talking about. It's
been _years_ of waiting as various analytics team have come and gone (does
anyone remember when Open Web Analytics was going to save the day?). And
yet it's 2014 and we still can't answer the most basic analytics questions
such as "how many views did X article get this month?" There has been some
deep dysfunction in this area of the Wikimedia Foundation, but I don't
know what any GLAM institution or Wikimedia Chapter can do about the
analytics situation, so it may be best to focus on other areas in which
making an impact is feasible.
MZMcBride