Wikipedia talk:Wikipedia Signpost/2015-04-01/In focus

← Back to In focus

Discuss this story

Genre of this piece / CA reorganization

• I'm not quite sure what this is intended to be: an analysis, an editorial, an opinion piece? Incidentally, only the "CA" part of "LCA" was pulled into the new Community Engagement department. Legal is separate. Risker (talk) 05:16, 3 April 2015 (UTC)
  • I suspect that In Focus is a little bit of each; that's what I like about it -- it keeps people guessing. It's experimental in its approach, and you don't know exactly what to expect in each paragraph. The only problem is the executive summary isn't comprehensive and it's impossible to navigate without a TOC of some kind. Otherwise, this is the way forward. Keep pushing the boundaries until you break on through to the other side... Viriditas (talk) 05:32, 3 April 2015 (UTC)
  • Good point about LCA; also, this reorganization happened in February 2015, not in late 2014. Regards, Tbayer (WMF) (talk) 10:20, 3 April 2015 (UTC)
• Much appreciation to the fine folk who put this together; lots of distilled yet sufficiently detailed information on a wide array of goings-on. I can literally smell the hard work put into it. I don't subscribe the Signpost, but I would subscribe to Special Editions that contain a section like this one. John Vandenberg ^(chat) 10:37, 5 April 2015 (UTC)

Mobile app

• "and a short description of the topic at the top of the articles taken from the corresponding Wikidata item (regrettably leading to increased vandalism on Wikidata)". How was this determined? Considering the descriptions of articles cannot be edited in the app at all and users are not informed where the descriptions came from, I find this a bit of a stretch, although I am happy to be proven wrong with conclusive data. --Dan Garry, Wikimedia Foundation (talk) 06:06, 3 April 2015 (UTC)

  I was told by the German Wikidata team that there was an edit button once that was disabled when in increase of vandalism was noted and led back to the mobile implementation. This still can be a cum hoc ergo propter hoc case. Cheers, —DerHexer (Talk) 09:53, 3 April 2015 (UTC)

  @DerHexer: That is incorrect. There has never been an edit button for the Wikidata description in any alpha, beta or production release of the Wikipedia app. This is something we're interested in, and something we've experimented with, but it has never gone into any release. --Dan Garry, Wikimedia Foundation (talk) 14:47, 3 April 2015 (UTC)

  I'll ask my contact person (a Wikidata developer), maybe I've misunderstood him. I changed my wording accordingly, are you okay with that? Anyway, thanks for clarification. Cheers, —DerHexer (Talk) 15:00, 3 April 2015 (UTC)

  @DerHexer: If I'm honest, I'm absolutely not okay with it, because there's zero evidence that the app has contributed to this at all. Your average reader doesn't even know what Wikidata is, much less know how to find the Wikidata item for the page they're reading then vandalise it using their mobile browser. There are zero links to or mentions of Wikidata in the app. This makes no sense to me at all. --Dan Garry, Wikimedia Foundation (talk) 15:18, 3 April 2015 (UTC)

  I've removed the parenthesis. Cheers, —DerHexer (Talk) 09:08, 4 April 2015 (UTC)

  Thanks! --Dan Garry, Wikimedia Foundation (talk) 22:58, 4 April 2015 (UTC)

• "The Signpost notes that the upload of images to Commons through one of the apps was disabled due to a massive spam attack of non-free and abusive images". This is incorrect. The Commons app was sunset because of its large number of bugs and lack of resources to support its development. This thread contains the full discussion. --Dan Garry, Wikimedia Foundation (talk) 06:16, 3 April 2015 (UTC)

  Well, this may not the reason for its deactivation but that most of these uploads had to be deleted cannot denied with mw:Mobile design/Uploads. Would you go with The Signpost notes that the upload of images to Commons through one of the apps that caused massive spam attack of non-free and abusive images was disabled its large number of bugs and lack of resources to support its development? Cheers, —DerHexer (Talk) 09:53, 3 April 2015 (UTC)

Hi DerHexer, you are linking to data about mobile web uploads, not app uploads. (IIRC there was a sense that contributors that used the apps were on the average far more experienced community members that those who used the mobile web upload without installing an app.) Unless you can point to strong evidence that the app caused a "massive spam attack of non-free and abusive images", I would suggest removing this claim. Regards, Tbayer (WMF) (talk) 10:31, 3 April 2015 (UTC)

Thank you both, I clarified that. Cheers, —DerHexer (Talk) 10:51, 3 April 2015 (UTC)

Thanks! --Dan Garry, Wikimedia Foundation (talk) 14:49, 3 April 2015 (UTC)

• "The Wikipedia app has been used more than six million times". I'm assuming that this is a misquoting an old monthly active users statistic, which stated that the Wikipedia apps for iOS and Android had a combined six million monthly active users. This is very different from it having only been used six million times. --Dan Garry, Wikimedia Foundation (talk) 06:19, 3 April 2015 (UTC)

  You're right, that's what I meant when summarizing the document. Unfortunately, as a non-native speaker I mixed up “is (currently) used” and “has been used”. If no one objects, I would correct that. Cheers, —DerHexer (Talk) 09:53, 3 April 2015 (UTC)

  Fixed! Thanks to you both. Ed ^{[talk] [majestic titan]} 10:14, 3 April 2015 (UTC)

  Thanks! --Dan Garry, Wikimedia Foundation (talk) 14:49, 3 April 2015 (UTC)

Grantmaking

• Regarding "a shift in focus from money and process to impact and non-monetary support": a shift in focus from donate to 'impact' in which sense: (a), (b), (c ), (d ), or (e).
  • a: highly influential in the past, yet obscure in the future
  • b: influential now, but behind the scene
  • c: pervasive influence, for unknown reasons at present
  • d: reputed influence, for protection and self-survival
  • e: political strength

  I'm afraid this shift will not turn out well for the encyclopedia. What's wrong with the moral authority that comes from voluntary effort, transparency of mission, and a clear statement, say, from Jimbo? --Ancheta Wis   (talk | contribs) 06:20, 3 April 2015 (UTC)

That statement was also baffling for me. I am hopeful that what WMF has in mind here is that the grantmaking staff wants to start engaging more with the people who actually edit and create content, to work on creating a happier and healthier atmosphere. Right now the WMF "community engagement" function doesn't really speak to my needs as a content editor, and seems to be all about working with the developer / coding community on changes to site functionality. Since the grants staff gets direct feedback from people who edit openly in real life, it would be a logical place to start to building some organizational capacity for improved communications and understanding of the content editors. --Djembayz (talk) 14:24, 4 April 2015 (UTC)

Djembayz, I'm not sure that grantmaking recipients map closely onto content editors. Tony (talk) 04:35, 5 April 2015 (UTC)

• I think perhaps the quoted sentence has been significantly misunderstood. To quote from the overleaf page in the lede section, "While there was explicit information, there were also coded messages ("a shift in focus from money and process to impact and non-monetary support")." What the In Focus article fails to explain to its readers is that this section is about how grants will be given out (P.26 of the actual report); I don't actually understand why the Signpost staff thought this was a "coded message", especially since they were reading the whole report in context. I can speak wearing my Funds Dissemination Committee (FDC) member hat here in saying that there is genuine concern that the $8 million or so of grant money (about $6 million of it going to the 17 or so largest chapters and affiliates), has in the past been dispensed with relatively little emphasis on "value for money" until very recently when performance metrics were brought to bear; also in the past year, the processes for applications are being streamlined and much more focused on key metrics. I believe this quotation, when correctly applied to the various grants programs, refers to the need to continue building on shared knowledge tools such as learning patterns, appropriate goal-setting and progress reporting, and so on. It is definitely a work in progress, but speaking as an FDC member, the questions we most commonly ask, and the areas where we give very significant focus, are on how the programs offered by these very large organizations are having an effect on the Wikimedia family of projects. I will speak personally in saying that I was surprised at how little actual effect on our projects some of the programs of large chapters actually have; on the other hand, I've been impressed with several other programs, and I'm concerned that we should probably find better ways to fund multi-year programs like Kiwix and Wikidata instead of using unrestricted annual grants for these longterm programs that are well-aligned with movement priorities. Some of the FDC's grantmaking principles can be seen in the committee's 2014-15 Round 1 funding recommendations, and it is filtering through to other levels of the grants programs as well. Risker (talk) 03:01, 5 April 2015 (UTC)

Superprotect

• The report does not mention the superprotect issue, which resulted in a storm of heated discussions in the communities, including a letter that was signed by about 1,000 contributors – I think, nothing could demonstrate more clearly how distant WMF officials are from the community. We have no common perception of events and no common goals any more. This is not a look back at 2014, but rather a document of the distortion in perception of events on the WMF's side. They missed the opportunity to be critical of their own stance in hindsight. The arrogance that goes with it is unchallanged. I am rather disappointed by all this.--Aschmidt (talk) 08:57, 3 April 2015 (UTC)

+1. There is a section lessons learned, but when avoiding the superprotect disaster, obviously the lessons have neither been learned, nor is there any intention to learn them in future. --Magiers (talk) 11:47, 3 April 2015 (UTC)

I'd even say that the WMF's ignorance about its role towards the community is nothing more than a slam in the face of every active editor. The relationship between our work and the fundraising results is also lacking badly. We've spent thousands of hours editing Wikipedia, it is our work and it is us who must have a say in which way the project goes. This report is absolutely embarrasing.--Aschmidt (talk) 12:44, 3 April 2015 (UTC)

I don't know if saying it's a "slam in the face" is the right set of words. I'm someone who's been here since near the beginning. When I read thru this summary of this document I feel as if I'm reading about the latest antics of a childhood friend who's become an international celebrity, & wondering how this person could have changed so much -- & if I really ever knew that person. With that perspective, I find it hard to be insulted -- just baffled, & more than a little sad. -- llywrch (talk) 07:44, 5 April 2015 (UTC)

Wikipedia Zero

The WMF has defended itself by pointing out that they do not pay carriers, nor offer it as part of a bundle: Facebook's Internet.org lists multiple app launches that include Facebook Zero and Wikipedia Zero (along with a variable range of other sites): Zambia (31 Jul 2014), Tanzania (29 Oct 2014), Kenya (14 Nov 2014), Colombia (14 Jan 2015), Ghana (22 Jan 2015), India (10 Feb 2015). Are these not bundles?

That aside, I appreciate the candidness of this WMF report. It's a positive development. Andreas JN466 09:22, 3 April 2015 (UTC)

Yes, but it didn't seem fair to accuse them of lying when it's not specifically Wikipedia Zero. It's a big difference, given that Wikipedia operates under a CC license. Ed ^{[talk] [majestic titan]} 10:08, 3 April 2015 (UTC)

Acgtually, I don't see Wikipedia Zero mentioned in any of these links. FYI, the Signpost author was referring to one of the Wikipedia Zero Operating Principles, which reads in full: "When encouraging carriers [Mobile network operators] to participate in the Wikipedia Zero initiative, we will maintain our commitment to the following principles: [...] Wikipedia Zero cannot be sold as part of a bundle. Access to the Wikimedia sites through Wikipedia Zero cannot be sold through limited service bundles." To my knowledge, this doesn't mandate suing third-party organizations to stop them from offering access to Wikipedia for free in separate (non-WMF) programs, whatever one thinks of Internet.org. Regards, Tbayer (WMF) (talk) 10:20, 3 April 2015 (UTC)

You're right, Internet.org doesn't mention Wikipedia Zero, only that the app "provides people with free access to basic internet services so they can browse useful health, education, finance, employment, communication and local information and services without data charges" (to quote one of the pages). It then lists Facebook and Wikipedia among the sites the app will provide free access to. The specific bundles in question are not mentioned on http://wikimediafoundation.org/wiki/Mobile_partnerships either.

I thought there was some active collaboration between Wikipedia Zero and Internet.org, having read a related Quora reply by Jimmy Wales. A few months prior to the start of these bundles, Jimmy Wales was asked on Quora "What does Jimmy Wales think about Mark Zuckerberg's Internet.org project, especially in light of Wikipedia Zero? Is there a chance for it to become a collaborative project between Facebook and the Wikimedia Foundation?",

He replied (my emphases):

I like what they are doing. I have spoken to both Mark Zuckerberg and Sheryl Sandberg about it, and the internet.org team is in contact with our Wikipedia Zero team.
Because Wikipedia/Wikimedia is somewhat "the Switzerland of the Internet" (i.e. with a strong tendency to be very vendor neutral) we are always going to be supportive of efforts like this, which are broad industry coalitions to do something useful particularly relating to broad access to knowledge, our core value. But we won't generally be tied up in any one thing per se. But we'll work with them where it makes sense, of course.
In my personal capacity, I am a big fan of what they are trying to do and support it fully.

(On an unrelated matter, Tilman: any updates on the 2012 gender stats?) Andreas JN466 10:57, 3 April 2015 (UTC)

Incidentally, people may not be aware that Facebook includes a fairly complete mirror of Wikipedia (articles only, no talk or project pages). See e.g. [1][2][3][4]. I can't swear that every last article is there (though all of these are fairly obscure), and I don't know how often the database is updated (articles only just created seem to be missing), but I believe this means Facebook Zero users automatically have access to pretty much all Wikipedia articles already. (The report mentions this as one of the possible reasons for the downturn in pageviews.) Andreas JN466 11:06, 3 April 2015 (UTC)

So we agree that Facebook does actually not list Wikipedia Zero as part of Internet.org in these announcements. Yes, I had seen you quoting Jimmy's Quora response in your mailing list postings already. One board member expressing an opinion about something, or teams being "in contact", does not automatically translate into official partnerships.

Facebook reusing Wikipedia articles is really old news. I wrote a Signpost article about this back in 2010, and it has been listed at Wikipedia:Mirrors and forks since that year too. Regards, Tbayer (WMF) (talk) 14:15, 3 April 2015 (UTC)

Yes, we agree that Facebook does not actually list Wikipedia Zero as part of Internet.org. (Nor does it list "Facebook Zero", of course; Facebook doesn't use that moniker on internet.org.) What it does say is that people using the internet.org app will incur zero data charges for accessing Wikipedia. In practice, it seems like a distinction without a difference. Regards, Andreas JN466 16:49, 3 April 2015 (UTC)

It probably doesn't make much of a difference for the amount of data fees that these people are incurring while accessing Wikipedia, yes.

However, since this article is about the actions of the Wikimedia Foundation, I do think that it is relevant here whether the initiative that enables this access is by the Wikimedia Foundation or by a different organization.

Regards, Tbayer (WMF) (talk) 23:21, 3 April 2015 (UTC)

Fair point. Regards, Andreas JN466 00:28, 4 April 2015 (UTC)

90% page views from Wikipedia Zero going to English Wikipedia

The analysis of this data is flawed, since we don't know for sure what "South Asia" is when you look at the dataset. See the mail about the problems tracking this data on the Analytics mailing list posted by Erik Zachter here: https://lists.wikimedia.org/pipermail/analytics/2015-April/003712.html So probably most measured ip addresses are coming from English speaking areas. Jane (talk) 10:14, 3 April 2015 (UTC)

Structured data and goalsetting

• Regarding "the theme of inadequate data is lurking in a number of places throughout the document, to the extent that in some respects the organization seems hamstrung by it ('An overwhelming percentage of project knowledge is not available as structured data'...": This paragraph seems a bit confused about what kind of structured data and project knowledge the quote refers to. It is not, as the surrounding text suggests, the information about the projects that is underlying the organization's data-based decisionmaking (which is of course important). Rather, it's about the huge amount of knowledge collected in the project themselves by editors for readers, as explained in more detail further below in the report ("Wikipedia content lacks consistency in styling, and is often not available in structured formats. ... Lack of clear structure makes reader and editor engagement features more difficult to build."). The "Engineering and Product" section of this Signpost story interpreted the same quote correctly.
• With respect to the article's criticism of "vague terms" etc., one may want to be aware that (apart from the mentioned annual plan, which uses broader strokes), the Foundation uses a quarterly goalsetting and review process which goes into much more detail (see e.g. the Q4 goals for the Engineering and Product departments which have just been published, and the documentation of past quarterly review meetings). Besides his work on the 2015 Call to Action that is mentioned in the article, Terry is driving and improving this quarterly process too.

Regards, Tbayer (WMF) (talk) 10:20, 3 April 2015 (UTC)

I found parts of this report, at least (I still have not read it in its entirety), unusually and refreshingly candid (specifically, those on Fundraising and WP0). But there is really no excuse for the persistence of "coded messages" such as those quoted in the intro. More plain, transparent speaking is required. Andreas JN466 11:12, 3 April 2015 (UTC)

Andreas, I think this parlance is a clear indicator that we are not the adressees of this report. It is a report aimed at future investors and other stakeholders.--Aschmidt (talk) 12:46, 3 April 2015 (UTC)

Yet it is being published publicly, which means that numerically the predominant audience is going to be an external one. It's an interesting strategic direction indicative that they're serious about communications with the community, but it does reduce the utility of the report. For instance, here Bayer is criticizing us for not correctly interpreting one bullet point buried in six in a different section from which the analysis is about. The report isn't intended to be confusing and I imagine that the people working in these sectors day-to-day don't have trouble parsing but at times to an external observer it is just that. ResMar 13:54, 3 April 2015 (UTC)

• While I noted some areas of vagueness myself, I don't agree with the assessment that there were a lot of "coded messages". There's a difference between jargon (which yes, I saw quite a bit of especially in certain areas) and obfuscation. If a physician tells a patient with a literature degree but limited sciences background that she has lymphocyte-depleted Hodgkins lymphoma Stage III and he would like to start a trial of ABVD, it may sound like a secret code to the patient, but it's simply a bunch of medical jargon that the speaker and others who are similarly informed would find entirely understandable. Perhaps it would be more effective and useful to point out areas where the jargon in this report gets in the way of comprehensibility, instead of assuming that there is a deliberate attempt to hide meanings. Risker (talk) 03:38, 5 April 2015 (UTC)

IIPM closed?

Quoting the article "Indian Institute of Planning and Management, a now-shuttered for-profit business school" I'm not aware that this is the case. Can you provide a reference? I'd doubt that an ArbCom decision could do that. There may be a mixup here with the Indian courts saying that IIPM doesn't offer degrees (and never could) but only offers certificates. In any case, if the school now is closed we should add that to our article on them.

BTW, I agree with comments above that it should be made clear that this piece is an editorial, not a news story.

Smallbones_(smalltalk) 13:32, 3 April 2015 (UTC)

Thanks. Changed to "controversial". Andreas JN466 17:19, 3 April 2015 (UTC)

Save time

A lot of the reduction in save-time must be attributed to the work of Wikid77, Mr Stradivarius and other coders who have (by and large) made a lot of highly used templates more efficient. All the best: Rich Farmbrough, 16:26, 3 April 2015 (UTC).

Throughout the migration, the cluster comprised a mixture of Zend and HHVM servers, and we separated the metrics and reviewed them side-by-side. The difference between HHVM and Zend when they were running concurrently is on par with the overall reduction in save time shown in the graph. That said, you are right to praise these coders, who quietly and without fanfare work on improving the performance of the Encyclopedia for everyone. --Ori.livneh (talk) 19:10, 3 April 2015 (UTC)

• Lua faster speed predated save-time quickening but data-hoarding is danger: Most of the prior speed improvements as noted and pioneered by Rich, through faster templates and migration to Lua script, occurred through 2013, such as the wp:CS1 Lua cite templates formatting pages 3x-4x faster and work done on many special slow templates as with Template:Weather_box quickened by User:Dragons_flight. Even then, the WMF developers had worked with us, by making the Lua-module interface run 2x faster soon after the initial deployment. Now, the recent 2014 quickening of save-time operations is yet another speed improvement: the Lua-based templates were the genesis of 2013's "7-second Wikipedia" to allow edit-preview of most pages within 7 seconds (typical attention span), but now the save-time speed is more like "5-second Wikipedia" or even faster. Before Lua, a heavily-footnoted page such as "Israel" formatted in perhaps 42 seconds, now previewed in 5 seconds or such. Unfortunately, the growing danger (the "elephant in the room") is the escalated massive wp:Data hoarding, such as article "United States" now with 555 cite footnotes. The Foundation needs to encourage a Micropaedia project to create mini-pages (perhaps below 1,000 words each) for popular topics and reduce the wp:Grandstanding which stuffs a page with 555 cites about pet theories of some POV-pushers wanting to wp:Coatrack into major pages. We have confirmed how the lower visibility of pages can reduce POV-pushing, but faster save-time speed has encouraged even more wp:data hoarding, perhaps 50% more than in 2012. -Wikid77 (talk) 18:57, 4 April 2015 (UTC)

@Wikid77: This is an interesting perspective. Have you written about it at length anywhere before? ResMar 19:53, 4 April 2015 (UTC)

I have for some time thought that we should re-think referencing. Certainly {{|Dragons_flight}}'s footnote listed references is a major step forward in simplifying source pages that we haven't capitalised on. Some of our more conservative interpreters of referencing style have left the project, perhaps making this whole issue more approachable.

The HMI aspect is that a user primarily wants to be able to see a footnote, without moving to the references section, and navigate to the resource if it is online. The gadget "Reference tooltips" provides this functionality, albeit in a slightly fragile way.

Given this, and the fact that references should ideally be either at the very end of the (visible) page (we currently have them before "External links", succession boxes, navboxes and "Authority control", for historical reasons), it would seem to be worth having a system that loads the main part of the page first and the references second. Indeed most web2.0 sites load the page in increments.

All the best: Rich Farmbrough, 21:36, 4 April 2015 (UTC).

HHVM impact

"Contrary to expectations, implementing HHVM did not increase the number of edits saved or reduce the time spent on the same amount of edits; more research is needed to achieve these goals" No: the analysis we were able to conduct was scoped to the productivity of new users during their first week as registered editors. The analysis did not find an impact, but extrapolating from that to conclude that the HHVM rollout had no effect is groundless. Absence of evidence is not evidence of absence. I would greatly appreciate if you print a correction.

I will say that as project lead of the migration, I deeply regret not spending more time on instrumentation and data analysis. A positive result would have made it easier to get buy-in from editors, donors, and management for further, ambitious performance work on site performance. --Ori.livneh (talk) 19:26, 3 April 2015 (UTC)

@Ori.livneh: We took that information from m:Communications/State_of_the_Wikimedia_Foundation#Engineering_Core. Is the report incorrect? Ed ^{[talk] [majestic titan]} 22:40, 3 April 2015 (UTC)

(ec) For me, according to current research was implied, more research can always prove former research wrong, there is no absolute truth. I hope I could clarify it with changing the second sentence into more research is needed to reflect whether or not these goals were achieved which itself implies that the first sentence depends on only some research that was made so far and led to this conclusion you also share in your document (“[…] research has yielded no evidence that HHVM has increased the number of edits saved or reduced the time spent completing the same number of edits for new editors. More research should be done in 2015 to determine what impact, if any, HHVM has had on editor engagement.”). Thanks for pointing me on that. Cheers, —DerHexer (Talk) 22:49, 3 April 2015 (UTC)

Thanks for correcting the separate misquote "more research is needed to achieve these goals".

But, as Ori said, absence of evidence is not evidence of absence. Therefore, the sentence in the report:

• "research has yielded no evidence that HHVM has increased the number of edits saved or reduced the time spent completing the same number of edits for new editors"

does not support the sentence in the Signpost story:

• "implementing HHVM did not increase the number of edits saved or reduce the time spent on the same amount of edits"

What's more, the Signpost also extended that claim to all editors instead of just new editors without justification. E.g. consider the possibility that the HHVM speedup has a more significant impact for more experienced editors who make a lot of small edits in rapid succession, say for categorization or vandal fighting.

(BTW, Aaron, Ori and I also considered the idea to measure whether, quite separately, HHVM might have a positive effect for readers by reducing the number of vandalized pageviews, because a vandal fighter's reverts save faster. Looking at the revert times charts on p.3 of Aaron's paper here and comparing them with the size of the HHVM speedup, such an effect does not seem impossible. But testing this hypothesis would involve quite a bit of work.)

Regards, Tbayer (WMF) (talk) 23:40, 3 April 2015 (UTC)

I see, for new editors must be read for both, the increased number of edits saved and the time spent completing the same number of edits. That was not obvious for me. A wording like “research on new editors (behavior) has yielded no evidence that …” would have avoided this ambiguity. I hope that the new text finally nails it. Cheers, —DerHexer (Talk) 00:22, 4 April 2015 (UTC)

(edit conflict) The most bewildering part of that statement in the report was in fact the need of the WMF to justify seemingly everything, even core engineering improvements as important as HHVM, in terms of "new editors". Frankly if you are unhappy with our interpretation of the events presented then part of the plane is on the WMF for failing to have provided the necessary context for independent reviewers to fully grasp your meanings. I don't see a huge leap from "research has yielded no evidence" to "did not increase", and I doubt that there is a huge rift between "new editors" and, um, everyone else, when it comes to the impact that such a basic infrastructure improvement generates. And what if the change didn't budge anything for new editors, but only old ones? By differentiating so—and maintaining the importance of such a distinction in the comments—you make apparent which body of editors matters more in your decisions, something well-known at this point, perhaps, but which still smarts. See also my response to your response to "Structured data and goal-setting", above. If these meanings are self-apparent to WMF staff, fine, but then why the hoopla? Why release this report publicly at all? ResMar 00:25, 4 April 2015 (UTC)

(edit conflict) Your assumption that we only care about new editors is baseless. The analysis was limited to new editors because existing users could opt in (via the HHVM beta feature) to having their requests handled by HHVM, and we worried that it would be unethical (and complicated) to quietly revoke the ability to toggle the beta feature from a subset of editors and to potentially clobber their preference. Aaron and I could not see an elegant or simple solution to this problem, and the window we had for conducting a comparison was closing. By restricting the study to new users, we could dodge these issues, and keep the implementation simple and robust. In no way did this decision reflect a value judgement on the relative "worth" of editors. As for the need to justify core engineering improvements: the migration to HHVM was the top priority for WMF Engineering for much of the project's lifespan, based on internal consensus that we had to make the sites faster for everyone. The resources that were committed to HHVM were not contingent on it increasing contributions. That said, I think it is proper for the Wikimedia Foundation to relate all initiatives to its core mission and to express the goals of these initiatives in human rather than mechanical terms. --Ori.livneh (talk) 01:26, 4 April 2015 (UTC)

That's a great explanation and I wish (cough) that it was something that was clearly indicated in the report. ResMar 01:55, 4 April 2015 (UTC)

I still don't know where I was talking about evidence on the absence while my comment intended to imply from what we know now/until there's evidence to the contrary, so the very opposite of evidence, especially when further research was announced. But maybe you can help me understand why I gave both of you that impression. It's probably a matter of different thinking or non-nativeness again. Cheers, —DerHexer (Talk) 11:31, 4 April 2015 (UTC)

Feedback culture

Hi! I'm a bit astonished how feedback was made on this summary. It's volunteers who spent hours on working through the longish report, trying to provide a summary in easy words for outsiders too and adding some points that they think are missing. Especially when they have been asked by the Wikimedia Foundation to do that. Even if there are (small) mistakes, they should be addressed by assuming good faith. Instead of claiming that something was incorrectly said or evidence to the contrary should be provided, we should better assume that something was mixed up for unknown reason or through an ambiguous phrase, or was misinterpreted. Reading phrases like I'm absolutely not okay with it, This makes no sense to me at all., No […] is groundless, or without justification does not convince me to spent time again on that. Cheers, —DerHexer (Talk) 01:03, 4 April 2015 (UTC)

If you are referring to my comments, you are right; I could have and should have made my appreciation explicit. The volunteers that publish the Signpost each week are amazing and wonderful to me. I am sorry if my comments made it seem otherwise. --Ori.livneh (talk) 01:29, 4 April 2015 (UTC)

Media Viewer

re Because the Media Viewer was deployed with too many bugs, <snip> In a lessons learned section,

May be I am flogging a dead horse, but I cannot keep wondering whether the "lessons learned" section comments on the way how this product was forced down our throats by WMF even more brutally than the previous programming gem of wysiwyg editor. Staszek Lem (talk) 01:06, 4 April 2015 (UTC)

Or how about when MV took over your entire mobile device when you opened an image? How could they have not seen this as a problem on mobile? Boggles the mind, really. I'm guessing there was virtually no testing. Viriditas (talk) 01:50, 4 April 2015 (UTC)

Not to downplay such a bug, but MediaViewer was never active on mobile devices, cf. [5]. Regards, Tbayer (WMF) (talk) 02:37, 4 April 2015 (UTC)

Right, so all of those times I opened up an image on iPhone's Safari browser on my phone only to have Media Viewer take over my screen are just fantasies I created, right? Are you serious? Has it possibly occurred to you that one can be using a mobile device but be logged on as a desktop viewer? Hmmm? Think about what you are going to say before you reply, please. I'm really sick of the continuing stream of denials, deflections, and distractions. Your reply indicates an extreme level of pathological denial. Viriditas (talk) 02:46, 4 April 2015 (UTC)

Let's not devolve into personalized dialogue, Viriditas. I've known Tilman for quite some time; he's not trying to bullshit you. In fact, if I were him, I'd emphasize that you're using a website designed for a large desktop computer on a small phone. Your experience will obviously be less than ideal. Ed ^{[talk] [majestic titan]} 03:07, 4 April 2015 (UTC)

Can you name one other feature that doesn't work with this setup or that interferes with using the site or even the browser? No, you can't, because there isn't one, only MV. I've been using Wikipedia on mobile devices since 2005 (BIP [Before iPhone], the PPC-6700), and I've written several GA's just on such a device. I've never encountered any major problems except for MV. Viriditas (talk) 03:17, 4 April 2015 (UTC)

I fail to see how taking up most of the screen is a "major problem." If you clicked on a photo before, you'd go to the file page ... where most of your screen is taken up by the image in question. Ed ^{[talk] [majestic titan]} 03:24, 4 April 2015 (UTC)

It's a major problem because you can't see the image at the resolution displayed, nor can you do much of anything else. It's all but useless. Of course, that's what I saw when MV was released, at which point I disabled it and haven't used it since. Viriditas (talk) 03:28, 4 April 2015 (UTC)

Thanks for the free psychiatric diagnosis ;) (No worries, heat of the moment etc.)

Your comment was indicating that this was a problem that was specific to mobile devices (as opposed to, say, particular screen sizes), and took the fact that there was a problem in this situation to conclude that there was "virtually no testing". So it entirely makes sense to point out that MV was never active on the normal mobile-optimized site. As said, I did not mean to downplay such a problem (it would be hard anyway to dispute or confirm it meaningfully in the absence of specific information like a Phabricator/Bugzilla ticket reference) and I certainly did not say you were imagining things. But indeed, while I have accessed the desktop site on smartphones quite a few times myself, it was not without expecting that it would present many a problem on mobile, e.g. requiring excessive scrolling at times. Regards, Tbayer (WMF) (talk) 03:34, 4 April 2015 (UTC)

At the time, there were what, 1 billion Android users and perhaps 300 million iOS users? I think arguing that MV was never for mobile-optimized browsers misses the entire point and deflects the problem. Why would you add a default feature that 1.3 billion people couldn't use in their desktop browser view, and more to the point, why would you force them to opt-out of it instead of opt-in? Is there anyone minding the store? Viriditas (talk) 03:46, 4 April 2015 (UTC)

I fail to understand your point. Like on most modern websites, these users will normally be served the mobile-optimized Wikipedia site (https://en.m.wikipedia.org/ ). Unless they switch back to the desktop version deliberately (after finding out how to do so); which is what I thought we were talking about above. Anyway, I think we are veering a bit off-topic at this stage. Regards, Tbayer (WMF) (talk) 03:52, 4 April 2015 (UTC)

And most modern websites (Google Images comes to mind) allow you to browse images using a media viewer-like interface on mobile without problem. You keep comparing apples and oranges. Viriditas (talk) 04:03, 4 April 2015 (UTC)

As mentioned in the link given above [6], the mobile-optimized site has its own image viewing function (which is separate from MediaViewer). If there are problems with it, they haven't been mentioned here yet. - Or we have been talking past each other and the issue you encountered did actually happen on the mobile site https://en.m.wikipedia.org/ instead of the desktop version https://en.wikipedia.org/, in which case your objection to my first comment would be even harder to understand. Regards, Tbayer (WMF) (talk) 04:25, 4 April 2015 (UTC)

I have filed phab:T95211 for this problem. —TheDJ (talk • contribs) 19:49, 6 April 2015 (UTC)

Not befuddled-- justfiably wary!

Stated above: "If anyone's equipped to deal with issues of information creation, access, and dissemination, it's librarians. But librarians with many years of professional experience can be completely befuddled by simple Wikipedia concepts or freeze up before hitting the submit button to make a minor edit."

Librarians have proven techniques for dealing with any sort of information or any sort of person: clear instructions for gathering, packaging and delivering information; clear rules and enforcement regarding disruptive behavior; and a commitment to upholding the standards of the profession and their institutions.

By contrast, the rules on Wikipedia are not clear, the enforcement on disruptive behavior is arbitrary or non-existent. Online game players, vulgarians, and sea-lioning randos who congregate here can be as disruptive and outrageous as they wish, with impunity. They don't care, because they don't have to.

In short, librarians have better options available than dealing with a software company that is attempting to exploit their free labor in an unprofessional atmosphere, and every reason to be wary of exposing themselves to the trolls and attack sites so active here.

It is pretty common at editing events to find that people don't want to hit that "save" button because they aren't sure they want to go on record, or bear responsibility for what appears on this website. They sense that we are attempting to recruit more obsessive editors, and they don't want Wikipedia editing to take over their lives. Enthusiasm is fine, but this constant pressure to push people to participate past their comfort zone because it creates "more impact with less dollars" is off-putting. If the Wikimedia movement wants to be sustainable, it needs to do better at understanding what it means to be a volunteer, and at accepting the level of editing contribution that individuals are in a position to give.

Expect polite participation at social editing events, and understand that if we do not make the atmosphere more professional here, most cultural professionals will have other tasks awaiting them that are much more important.--Djembayz (talk) 14:01, 4 April 2015 (UTC)

Agreed, the atmosphere here is definitely a reason many people stay away, especially those with professional or expert knowledge. But I think these conceptual issues come into play as well, and people have to get past those before they even get to the point that they are turned off by the negative atmosphere. You're right when you say "If the Wikimedia movement wants to be sustainable, it needs to do better at understanding what it means to be a volunteer", and that means dealing with those issues head on. Gamaliel (talk) 21:06, 4 April 2015 (UTC)

I just have to say that I completely agree with the paragraph including

" the rules on Wikipedia are not clear, the enforcement on disruptive behavior is arbitrary or non-existent. Online game players, vulgarians, and sea-lioning randos who congregate here can be as disruptive and outrageous as they wish, with impunity." I just had to think "What would happen if the conversations at Jimbo's talk page happened in a library." I hate to think of the police being called to a library, but I guess that's what would happen. Actually, people would never act that way at a library, so perhaps our job is to figure out why they act that way on Wikipedia and what we can do to stop it. Smallbones_(smalltalk) 21:25, 4 April 2015 (UTC)

Alas, people do act that way in the library sometimes. In many places, police officers are regularly stationed at some libraries. But that's a tangent. There are norms of behavior that people in general associate with the library, and the trick is getting people to adhere to them on the encyclopedia as well. Gamaliel (talk) 21:47, 4 April 2015 (UTC)

security

The article states that the signpost feels that security is understaffed. I would be curious as to the reasoning why the signpost feels that way (im not neccesarily disagreeing, just want to know reasoning). Are there specific issues the signpost would want to see better and faster responses in? Would it like to see more preventive actions? Are there aspects of site security it is specificly concerned about? Something else? Bawolff (talk) 21:26, 4 April 2015 (UTC)

Mostly with a view to security-related bugs. As said in the article, the operations part with respect to outside security threats have always been handled both quickly and accurately from what I can see. But talking to people who work on security-related bugs in whatever regard, always gave me the impression that more (wo)man power would be appreciated. Mainly, my comment was to lead to this job offering. I'm quite sure that other departments would also appreciate further support. Cheers, —DerHexer (Talk) 00:39, 5 April 2015 (UTC)

On the operations side, a person was just hired for security. On the platform side, Tim and last year mostly Chris Steipp have been doing it as I think you are aware. The last quarterly review meeting did mention that they are looking for someone to help Chris, but it's a difficult position to fill. —TheDJ (talk • contribs) 19:43, 6 April 2015 (UTC)

Fundraising

Wikimedia Foundation financial development 2003–2014. Green is revenue, red is expenditure, and black is assets, in millions of dollars.

Based on the data on https://frdata.wikimedia.org/ and the info that the Foundation reached its $58.5 million target for the 2014–2015 year on January 7, the Foundation has so far taken about $62 million in the financial year to date. If the pattern observed over the past three months continues, the year-end total will exceed $65 million (compared to $52.6 million in 2013–2014). Andreas JN466 09:02, 5 April 2015 (UTC)