On Thu, Dec 3, 2015 at 2:32 PM, Oliver Keyes wrote:
> A metric that is based on a draft RfC that was only created this year
> and depends on JS? I'm going to go out on a limb and suggest it has
> problems of its own ;p
>
> On 3 December 2015 at 14:22, Gabriel Wicke wrote:
> > I have witnessed th
A metric that is based on a draft RfC that was only created this year
and depends on JS? I'm going to go out on a limb and suggest it has
problems of its own ;p
On 3 December 2015 at 14:22, Gabriel Wicke wrote:
> I have witnessed this discussion about what constitutes a page view
> repeatedly ove
I have witnessed this discussion about what constitutes a page view
repeatedly over the last months, and suspect that it is only going to
get murkier the more interactive and non-navigation features we add.
Some of these decisions are somewhat arbitrary, making the page view
metric a less accurate
>
> Sure. But in that, the answer to "who owns this?" is Analytics. It's
> Analytics who reached out and made sure everyone had a voice, it's
> Analytics brainstorming, it's Analytics implementing it, it's
> Analytics publicly logging it.
>
> Again, ownership does not mean code reviewing everything
On 3 December 2015 at 13:56, Dan Andreescu wrote:
> On Thu, Dec 3, 2015 at 1:13 PM, Oliver Keyes wrote:
>>
>> I have been informed MW Core no longer exists. Fair catch ;p. But this
>> is software development, not Oprah - product ownership is not
>> something under the seat of every audience membe
On Thu, Dec 3, 2015 at 1:13 PM, Oliver Keyes wrote:
> I have been informed MW Core no longer exists. Fair catch ;p. But this
> is software development, not Oprah - product ownership is not
> something under the seat of every audience member. Someone needs to
> actually own the definition. I don't
I have been informed MW Core no longer exists. Fair catch ;p. But this
is software development, not Oprah - product ownership is not
something under the seat of every audience member. Someone needs to
actually own the definition. I don't mind if it's AnEng, Research,
Readership, Search, whoever, bu
Sure, but ownership does not mean "knows everything" it means "Makes
sure it gets done". MediaWiki is owned by everyone, sure, but the
actual idea of what MediaWiki core is has a team. The MediaWiki core
team.
On 3 December 2015 at 13:07, Nuria Ruiz wrote:
>>So the pageview definition, one of our
>So the pageview definition, one of our core organisational KPIs, is owned
simultaneously by everyone?
Sure, just like mediawiki codebase is collectively own.
Personally I do not see that as a problem and regardless I think it
reflects reality, analytics team -as I mentioned before- doesn't have t
So the pageview definition, one of our core organisational KPIs, is
owned simultaneously by everyone?
On 3 December 2015 at 12:54, Nuria Ruiz wrote:
>>Who does own it?
> On our opinion every team should own the definition of a pageview in their
> product right and when in doubt analytics or resea
>Who does own it?
On our opinion every team should own the definition of a pageview in their
product right and when in doubt analytics or research can be involved to
provide feedback on lessons learned.
For example: who is best qualified than IOS team to decide what constitutes
a page in the IOS a
On 2 December 2015 at 15:38, Nuria Ruiz wrote:
>>It's true that MediaWiki supports search based solely on the ?search query
>> parameter. Regardless of whether title=Special:Search is specified.
>>This is mostly for legacy reasons as search predates the concept of special
>> pages.
>>However, woul
>I filed that ticked almost a month ago, without a reaction so far. I take
it that for the purpose of getting the attention of Analytics engineers
towards such widespread anomalies, this mailing list is a far better venue
than Phabricator.
Not really, it just means that on the last month we had hi
On Wed, Dec 2, 2015 at 3:38 PM, Nuria Ruiz wrote:
> According to Timo's comment it seems that "?search" requests should be
> counted as true pageviews but let us know otherwise via phab ticket or this
> list.
>
Well, make a phab ticket either way. This is a simple change, but I agree
with Nuria
>It's true that MediaWiki supports search based solely on the ?search query
parameter. Regardless of whether title=Special:Search is specified.
>This is mostly for legacy reasons as search predates the concept of
special pages.
>However, would it make sense to instead count these as page views for
Oliver Keyes, 02/12/2015 18:52:
Via Brian Davis we find out the responsible patch is
https://github.com/wikimedia/analytics-refinery-source/commit/05e5da92553dbd3e691eb45d40e559895337935f
Context:
https://meta.wikimedia.org/wiki/Research_talk:Page_view/Archive_1#Parameters_appended_to_short_UR
On Wed, Dec 2, 2015 at 9:38 AM, Joseph Allemandou wrote:
> Food for thoughts:
>
> Regarding the "???.." entries below, see also
https://phabricator.wikimedia.org/T117945 ; the list there shows the
prevalence of the "-"s across languages, too.
(I filed that ticked almost a month ago, without a re
It's true that MediaWiki supports search based solely on the ?search query
parameter. Regardless of whether title=Special:Search is specified.
This is mostly for legacy reasons as search predates the concept of special
pages.
However, would it make sense to instead count these as page views for
'
Via Brian Davis we find out the responsible patch is
https://github.com/wikimedia/analytics-refinery-source/commit/05e5da92553dbd3e691eb45d40e559895337935f
(I agree wholeheartedly with previous Oliver's assertion that
MediaWiki is trying to kill me)
I'll leave it up to y'all whether you remove it
Why the heck is search in
https://github.com/wikimedia/analytics-refinery-source/blob/master/refinery-core/src/main/java/org/wikimedia/analytics/refinery/core/PageviewDefinition.java#L74
? (I say that knowing it is probably my fault)
On 2 December 2015 at 12:38, Joseph Allemandou
wrote:
> Food fo
Food for thoughts:
SELECT
uri_host, uri_path, uri_query, COUNT(1) as c
FROM wmf.webrequest
WHERE webrequest_source IN ('text', 'mobile') AND
AND is_pageview AND pageview_info['page_title'] = '-'
GROUP BY uri_host, uri_path, uri_query
ORDER BY c DESC LIMIT 100;
en.m.wikipedia.org /w/index.ph
I mean, now I want to know how we can have a condition where there's
no page title but it registers as a pageview.
On 2 December 2015 at 12:14, Joseph Allemandou
wrote:
> Double checked:
> https://github.com/wikimedia/analytics-refinery-source/blob/master/refinery-core/src/main/java/org/wikimedia
Double checked:
https://github.com/wikimedia/analytics-refinery-source/blob/master/refinery-core/src/main/java/org/wikimedia/analytics/refinery/core/PageviewDefinition.java#L117
This value is the default when no page title is found.
I agree it's not very explicit.
Any suggestion on changing it, or
Can someone dig into it? We should really be excluding that (unless it
is the page on the dash ;p)
On 2 December 2015 at 12:00, Dan Andreescu wrote:
> I always wonder about that. There's also an actual page that could
> theoretically be hit:
> https://en.wikipedia.org/w/index.php?title=-&redirec
I always wonder about that. There's also an actual page that could
theoretically be hit:
https://en.wikipedia.org/w/index.php?title=-&redirect=no
On Wed, Dec 2, 2015 at 11:58 AM, Gabriel Wicke wrote:
> Historically, I vaguely remember that we have used that title for user
> script / style loadi
Historically, I vaguely remember that we have used that title for user
script / style loading with action=raw. I think that's gone from the
skin code, but it's possible that user scripts still reference this
title.
Gabriel
On Wed, Dec 2, 2015 at 8:41 AM, Oliver Keyes wrote:
> One of the most pro
One of the most prominent top articles has no page; it's "-". What is this?
--
Oliver Keyes
Count Logula
Wikimedia Foundation
___
Analytics mailing list
Analytics@lists.wikimedia.org
https://lists.wikimedia.org/mailman/listinfo/analytics
27 matches
Mail list logo
