Hi again!
Forgot to mention that I have currently stopped all Analytics jobs on
Hadoop (Camus, Oozie) as preparation step for a complete Hadoop cluster
reboot to install new Linux kernel upgrades. The reboots will be performed
in small batches to limit the blast radius but it might affect your jobs
anyway.
Please let me know if you'll encounter any issue during the next hours.
Thanks!
Regards,
Luca
Hello Analytics,
Is there any standard implementation for split testing ("AB-Testing") that
is used or should be used?
I researched a bit on mediawiki.org but could not find any information on
this. (Ideally, I'd use a javascript based solution but PHP will be OK, too)
Kind Regards,
Jan
--
Jan Dittrich
UX Design/ User Research
Wikimedia Deutschland e.V. | Tempelhofer Ufer 23-24 | 10963 Berlin
Phone: +49 (0)30 219 158 26-0
http://wikimedia.de
Imagine a world, in which every single human being can freely share in the
sum of all knowledge. That‘s our commitment.
Wikimedia Deutschland - Gesellschaft zur Förderung Freien Wissens e. V.
Eingetragen im Vereinsregister des Amtsgerichts Berlin-Charlottenburg unter
der Nummer 23855 B. Als gemeinnützig anerkannt durch das Finanzamt für
Körperschaften I Berlin, Steuernummer 27/029/42207.
Dear friends,
We've been investigating some Fundraising nastiness, and I'm beginning to
think that it comes down to a faulty "kafkatee" feed going into the frack
cluster.
https://phabricator.wikimedia.org/T132500
I see about half the number of webrequests in our files, compared to what I
believe are similar queries done on the stat1002 copy of the requests.
This is borne out by another source of data we're cross-checking
against--we currently see lower numbers of fundraising landing page
impressions than of actual donations, which is clearly wrong.
Any help would be appreciated, with reviewing the queries I've made, or
reasoning about what might be failing here.
Thanks,
Adam
Hi folks,
I can't get pageview analysis graphs to load on 2 wikis that I've tested,
and I've tried desktop and mobile on multiple browsers. Can someone take a
look at what might need fixing?
Thanks!
Pine
Hello!
(adding analytics@ public list)
>Opera Mini sends different user agent if it's in Mini or Turbo mode and UC
Mini has a little different user agent than the rest of the UC >browser. I
did a small writeup https://wikitech.wikimedia.org/wiki/ProxyBrowsers
Thanks for doing the write up, I have edited it a bit.
> I wonder if it would be possible to break up the stats for Opera Mobile
so that we could have mini with "mini data savings" in one bucket and >the
same with UC mini.
We use ua-parser to do browser detection so on this regard our preference
would be to do a pull request and add additional parser for Opera mini that
now doesn't look like it is in place. In our end we will just update UA
parser and get those changes. See:
https://github.com/ua-parser/uap-core/blob/master/regexes.yaml#L110
The raw data (available for 60 days only) does have the raw user agent and
you can use that to get an idea of how many people have remote rendering
enabled. We can help you query the dataset as needed be. It is documented
here:
https://wikitech.wikimedia.org/wiki/Analytics/Data/Webrequest
How to get access:
https://wikitech.wikimedia.org/wiki/Analytics/Cluster/Access
On Thu, Jun 23, 2016 at 8:40 AM, Peter Hedenskog <peter(a)wikimedia.org>
wrote:
> Hey Nuria,
> hope you are doing good, we talked last year when I was trying to get you
> to talk at a conference in Sweden :)
>
> Wow, I'm super impressed by https://analytics.wikimedia.org/ what a great
> work by you and the analytics team!
>
> I've been looking into how we handle proxy/remote browsers (where the
> browser vendor either tune or render the page server side and send a blob
> to the client). For Opera Mini (setting Data savings to Opera Mini instead
> of Opera Turbo) and for UC Mini (server side rendering) we have a lot of
> things we can do better. The search can work better on Opera and on UC the
> search doesn't work at all and our navigation is removed. And there are
> probably more things, I haven't looked into it too much.
>
> Opera Mini sends different user agent if it's in Mini or Turbo mode and UC
> Mini has a little different user agent than the rest of the UC browser. I
> did a small writeup https://wikitech.wikimedia.org/wiki/ProxyBrowsers
> (but I only tested the latest versions of the browsers so it isn't
> perfect). I think you have better insights than me on this :)
>
> I wonder if it would be possible to break up the stats for Opera Mobile so
> that we could have mini with "mini data savings" in one bucket and the same
> with UC mini. It's the same browser family but the behavior is totally
> different. If we can do that, it will make it easier to see if our changes
> will have impact
>
> All the best
> Peter
>
You may want to follow https://phabricator.wikimedia.org/T68699 , since
it will affect how often user_touched is updated for users that choose
"Keep me logged in".
Matt
Hi Analytics folks,
Now that the pageview graphs are working again (thanks Toby!) I decided to
take a quick look at our graphs of readership of main pages of a few wikis.
Most wikis were relatively flat, but there were a few exceptions:
https://tools.wmflabs.org/pageviews/?project=en.wikipedia.org&platform=all-…https://tools.wmflabs.org/pageviews/?project=es.wikipedia.org&platform=all-…https://tools.wmflabs.org/pageviews/?project=commons.wikimedia.org&platform…
I was guessing that the declining numbers on ESWP could be related to the
ending of the school year, but (1) that same pattern isn't present on other
wikis, and (2) the school years of a number of predominantly
Spanish-speaking countries don't end in June as US school years often do.
The increasing numbers on ENWP are still happening despite the school year
having ended for most schools, so this is a pleasant surprise. Any idea
about what could be causing that?
Similarly, we're seeing some nice growth in the Commons main page stats.
Any ideas about what could be causing these three sets of numbers to have
these patterns?
Thanks,
Pine
We just released a new version of Research:FAQ on Meta [1], significantly
expanded and updated, to make our processes at WMF more transparent and to
meet an explicit FDC request to clarify the role and responsibilities of
individual teams involved in research across the organization.
The previous version – written from the perspective of the (now inactive)
Research:Committee, and mostly obsolete since the release of WMF's open
access policy [2] – can still be found here [3].
Comments and bold edits to the new version of the document are welcome. For
any question or concern, you can drop me a line or ping my username on-wiki.
Thanks,
Dario
[1] https://meta.wikimedia.org/wiki/Research:FAQ
[2] https://wikimediafoundation.org/wiki/Open_access_policy
[3] https://meta.wikimedia.org/w/index.php?title=Research:FAQ&oldid=15176953
*Dario Taraborelli *Head of Research, Wikimedia Foundation
wikimediafoundation.org • nitens.org • @readermeter
<http://twitter.com/readermeter>
People are now working to conclude the "Code of Conduct/Cases" part of
the draft Code of Conduct:
* Section:
https://www.mediawiki.org/wiki/Code_of_Conduct/Draft#Page:_Code_of_Conduct.…
* Talk:
https://www.mediawiki.org/wiki/Talk:Code_of_Conduct/Draft
* Alternatively, you can provide anonymous feedback to
conduct-discussion at wikimedia.org .
This is the best time to make any final necessary changes to this
section (and explain why, in edit summaries and/or talk) and discuss it
on the talk page.
After this last call, I will send out another email seeking approval,
like before.
Thanks,
Matt Flaschen
Hello!
We would like to announce http://analytics.wikimedia.org the domain where
the analytics team hopes to consolidate dashboards and data downloads (a
work in progress)
We have two dashboards to announce with data we hope you find interesting:
1. The new traffic reports with browser data for mobile and desktop domains:
*http://tinyurl.com/zqqr3zd <http://tinyurl.com/zqqr3zd>*
2. The vital signs dashboard that displays several metrics that are of
interest like Pageviews and Unique Devices:
*http://tinyurl.com/jr4oou3 <http://tinyurl.com/jr4oou3>*
Thanks,
Nuria