Wikipedia:Wikipedia Signpost/2017-02-27/Op-ed

← Back to Contents

View Latest Issue

27 February 2017

Op-ed

Wikipedia has cancer

Contribute —

Author's Note

Author's note: At the request of the editors of The Signpost, this version has fewer citations and less information in the graphic. The original version is at User:Guy Macon/Wikipedia has Cancer. --Guy Macon (talk) 05:00, 28 February 2017 (UTC)

Expenses and pageviews

How much of this money went to payout a CEO that took down Wikipedia from the 3rd most used site in the world to wherever is it now? Her wikipedia page literally doesn't even show up in Google. Nergaal (talk) 01:35, 28 February 2017 (UTC)
- I get the general point, but [citation needed] on the specifics. I don't know that Wikipedia has ever risen above #5 on the ComScore ranking (lower on Alexa), and if there's been a recent change attributable to any specific person or time period I'm not aware of it. -Pete Forsyth (talk) 01:50, 28 February 2017 (UTC)
- If we do as I suggest in the second to the last paragraph ("We should make spending transparent, publish a detailed account of what the money is being spent on and answer any reasonable questions asking for more details") we will know how much we paid her. Of course we wouldn't reveal the [pay of lower-lever employees, but the CEO? Definitely. --Guy Macon (talk) 05:37, 28 February 2017 (UTC)
  - This remark is another illustration of the stunning ignorance that the op-ed author shows about his topic (below, I already mentioned the Foundation's actual software development processes as another example). Yes, publishing the executive director's salary is a reasonable demand - but sir, it *is* already being published year after year, alongside those of other executives, as part of the Form 990, as is standard practice for this type of nonprofit organization. These forms are listed one click away from the homepage of the organization about whose financial accountability you are so concerned about, and (besides being unaware of the basic legal requirements here), it appears you didn't even take a look at them before making bold claims about financial transparency.

This ignorance stand in astonishing contrast to the strength of opinion displayed in the op-ed. (Or maybe not that astonishing.) Regards, HaeB (talk) 06:21, 28 February 2017 (UTC)

Personal attacks do not strengthen your argument. Quite the opposite, actually. Is it your position that the WMF has already "made spending transparent, published a detailed account of what the money is being spent on, and answered all reasonable questions asking for more details"? Is it your position that such transparency is not desirable? Or is your only point that I don't have every document ever created by the WMF memorized, which appears to be your definition of "stunning ignorance"? --Guy Macon (talk) 13:29, 28 February 2017 (UTC)

If the information is already available, please point to it. The last published WMF 990 that I found is 2014-2015. Even a 990 for the fiscal year in which Lila was employed may not disclose severance payments, or may lump compensation in across the tenure of both the current and former ED. Nathan ^T 22:44, 28 February 2017 (UTC)

I imagine that the next 990 will be available whenever wmf does its taxes. I dont think its unreasonable for there to be a small delay before info gets posted. We are only 2 months out of 2016. See also meta:Wikimedia_Foundation_salaries. Bawolff (talk) 23:10, 28 February 2017 (UTC)

Thanks for the pointer, Brian. I haven't seen that before. Nathan ^T 23:28, 1 March 2017 (UTC)

Sounds like a success to me. Some tweaks are in order, but overall health seems vibrant and strong. Randy Kryn 01:38, 28 February 2017 (UTC)
- "The modern Wikipedia hosts 11–12 times as many pages as it did in 2005, but the WMF [...] is spending 1,250 times as much overall." Yeah, healthy indeed /s. Nergaal (talk) 01:40, 28 February 2017 (UTC)

I haven't been monitoring expenses and growth at the WMF for as long as Guy, but almost. My impression is similar. You can deduce the primary mission of an organization by where it devotes the greatest proportion of its resources; for the WMF, it appears as though continuously increasing the number of employed people is its main objective. For the tens of millions of dollars and tens of thousands of work hours invested each year, there is very little evidence of actual impact or contribution to the WMF's stated mission. Nathan ^T 01:54, 28 February 2017 (UTC)

A missing figure is something like [1] - can you get a longer timeline? We are seeing a hard to believe yet obviously well orchestrated shift of readers away from real computers viewing a free and open internet to ... "apps" where they are herded on their tiny screens like cattle under the all-seeing eye of some corporation, and surprisingly enough, fewer people want to go on to read a whole encyclopedia on a postage stamp. A decreasing total supply of eyeballs would seem to imply that the potential for fundraising has to decrease as well. To be sure though, the last four years in fundraising don't look like exponential growth at all - it isn't quite a static pool of money, but it's clear that the potential for increases is already pretty much tapped out - as we'd expect from the viewership figure. I'm not sure I'd call this cancer though - more like a transition from log phase to stationary phase. I'll leave it as an exercise to the reader to look up what comes after that ... but we should bear in mind that inherently stationary phase can last a *long* time. It is quite possible for other adverse cultural phenomena to tear it apart far sooner, but not simply a static level of resources. Wnt (talk) 01:53, 28 February 2017 (UTC)

Some relevant data:

Year	Planned spending	Actual spending
2007-08	$4.6M	$3.5M
2008-09	$5.97M	$5.6M
2009-10	$9.4M	$10.2M
2010-11	$20.4M	$17.9M
2011-12	$28.3M	$29.2M
2012-13	$42.1M	$35.7M
2013-14	$50.1M	$45.9M
2014-15	$58.5M	$52.5M
2015-16	$65M	$65.9M
2016-17	$63M	TBD

To me, it looks quite possible that spending has plateaued. Far from certain, since we're still short on data points. We'll know more when the first draft of the 2017-2018 plan is published on April 7. The community review of the plan starts on the same day, incidentally. --Yair rand (talk) 03:11, 28 February 2017 (UTC)

The WMF servers host not only Wikipedia, but many other sister projects. It would be nice to at least roughly factor them into the equations. --Felipe (talk) 03:21, 28 February 2017 (UTC)

From the "The Resources We'll Need" section of the 2010-15 strategic plan - compare with the numbers above

Among the many things wrong with this piece is the insinuation (implicit in the cancer metaphor) that the growth in expenses was unplanned and uncontrolled. If one compares the five-year spending projection from the 2010-15 strategic plan, one finds that the actual development stayed pretty close to the plan, which had been developed in a collaborative process with extensive community involvement and lots of discussion about what to spend money on - and what not. And underspending compared to plan has actually repeatedly been an issue.

Software engineering

"I have never seen any evidence that the WMF has been following standard software engineering principles" - that might be because you did not bother to look. Quarterly goals are available for all engineering teams on mediawiki.org, and the product teams largely follow the principles of agile software development (in particular Scrum). Agile might not use the exact same elements as your favorite 1980s era formulaic software process specification, but it is pretty much the standard today. BTW, is it correct to assume that the consulting business which is mentioned in the article intro (as evidence for the author's alleged expertise about the topic) is not about software, but mainly about other forms of engineering?

At https://phabricator.wikimedia.org/ , it's possible to follow along the work in detail, minute-by-minute. I would invite everyone to take a look in detail at the tasks that are being worked on right now - pick one at random from the "Activity Feed" or "Fresh tasks" boxes, for example, try to understand what its purpose is, what the downsides of ignoring the task might be, and then form an opinion about whether or not WMF should spend money on it. (Or, about how many of them could actually be imposed on volunteer developers, as the article tries to make use believe is possible for almost any useful software work for Wikimedia sites.)

The discussion about how much money WMF should spend, and on what, is of course absolutely legitimate and necessary. But simply defining growth as inherently bad is a fallacy. Obviously any organization has to grow from zero to its desired budget size, so with it will always be possible to cry "OMG 1000% growth!!". And yes, with all due respect to Jimmy, I think that that talk twelve years was dangerously mistaken about what would be needed even for the minimal goal of keeping Wikipedia online, stable and safe, without ever improving the interface, add new features etc.

Besides that, the amount of technical debt Wikipedia's software has accrued during those many underfunded years is enormous, and still felt today. (Just as a small example: To handle the rising traffic load, developers back then had to resort to very aggressive caching strategies, which succeeded in keeping Wikipedia available to readers at a bearable speed, but had the side effect of imposing a performance penalty on registered users - for which pages can't be cached - and anyone who edits a page. I.e. editors, who are the lifeblood of the project, were precisely those for whom it was the most sluggish. This only began to be mitigated a little over two years ago, when enough resources became available for seriously starting to work on performance.) And Wikipedia still has much less paid tech staff than basically any other website of its size.

(Personal comments, informed both by my experience as WMF employee and as community member since 2003, who for many years prior to joining WMF observed it from the outside, including writing about its activities and spending as former editor-in-chief of the Signpost.)

Regards, HaeB (talk) 03:27, 28 February 2017 (UTC) ┌────────────────────────────────────────────────────────────────────────────────────────────────────┘ Re: "is it correct to assume that the consulting business which is mentioned in the article intro (as evidence for the author's alleged expertise about the topic) is not about software, but mainly about other forms of engineering?", that is correct. I do a lot of work with software, but on chips that typically cost less than five cents, have 256 nybbles of RAM, and are not programmable -- meaning that we write raw assembly language code with no hardware to run it on other than a few crude simulation tools, then order masked-ROM parts (minimum order 50,000 chips with 6 to 8 weeks from code submission to chip arrival) and throw them all away if there is a bug. Clearly my experience isn't even close to the same as what the WMF developers do, and thus my opinions on this are suspect. Getting back to your main point, I remain 100% convinced that the WMF developers are doing the right thing and are using techniques that we who are in the business of producing 100,000 electronic toys per hour know nothing about. Nonetheless, the end result has been really, really awful. I have been around long enough to recognize the smell of good developers and good middle managers working hard and getting bad results because of bad top management decisions. So I may be completely wrong in defining what the solution is, but few would disagree that, as an organization, the WMF sucks at creating high-quality software.

Depiction of Wikimedia Foundation destroying Wikipedia with Visual Editor, Flow, and Mobile App.

--Guy Macon (talk) 06:04, 28 February 2017 (UTC)

Hosting

I feel that the comparision with hosting costs is unfair. We do spend more money, but back in the day we also had plenty of down time and little redundency. We now have data centres in multiple places in the world, much improved performance, etc. We are not paying more money to get the same thing scaled to more page views, we are paying more money to get a better thing. Maybe that's worth it, maybe it's not, but comparing them like they are the same is unconvincing (which is not even to say you are necesarily even wrong in some aspects of wikimedia spending, the supporting evidence is just unconvincing). Bawolff (talk) 03:42, 28 February 2017 (UTC)

I really don't see any problem with WMF spending 33 times as much on hosting, and I wouldn't mind them having 33 times as many employees. 300 times as many employees and 1,250 times as much spending looks like cancer to me. --Guy Macon (talk) 06:04, 28 February 2017 (UTC)

Im also not sure if internet hosting includes only bendwidth or also other things. According to [2] (which is a much better source than some random comment by jimbo) hosting costs in the 2005-2006 year was $189,631. So thats only about a ten fold increase. Not only is that linear with page views, thats not even taking into account that we are now significantly more geographically distributed (instead of most of our eggs being in a basket prone to hurricanes), pages are probably longer on average, pages have more images and sometimes even movies. Bawolff (talk)

additionally for the 300x employee increase, im not sure its really fair to extrapolate from employee #1. At the very least you should try to account for software development donated (aka volunteer contributions). Since 2005 we went from an almost entirely volunteer developer model to a mostly paid developer model with some volunteers. There's some benefits to doing that and there are some cons. One could certainly argue that wikipedia is volunteer based maybe MediaWiki should be to. However thats an entirely separate argument. If we are comparing direct staffing costs we really need to take into account that there are people who are paid now who used to do the exact same thing for free, and that starting to pay these people is not the same as wmf expanding for expansion's sake. Bawolff (talk) 16:39, 28 February 2017 (UTC)

You make a compelling case. Given the above figures, I agree; we are not overspending on hosting. But then again, hosting is a relatively small part of our budget. --Guy Macon (talk) 17:14, 28 February 2017 (UTC)

While it would be nice to see an official WMF rebuttal, I am with HaeB on this one. Even the table at the top shows that net assets are increasing. Spending is less than revenue. This talk of 'debt' seems like fearmongering. This entire piece seems to me like a poor quality rant about WMF. I could never understand why some community members are so anti-WMF, but I was never particularly impressed by their 'arguments', and I am not impressed this time. I don't think such rants have place in The Signpost, neither. --_{Piotr Konieczny aka Prokonsul Piotrus| reply here} 10:25, 28 February 2017 (UTC)
- It sounds like you believe that revenue can increase forever. Please clarify. --Guy Macon (talk) 13:29, 28 February 2017 (UTC)

Exponential then linear?

The trend on the spending curve looks two phased to me - exponential then linear. I see Wnt has already noted this. Still unsustainable but can last much longer than exponential growth. Widefox; talk 12:13, 28 February 2017 (UTC)
- Could be. It is hard to plot a curve from one or two points, and even harder to predict future spending. Here are the raw numbers:

  SPENDING (Dollars):
  2003/2004 $23,463
  2004/2005 $177,670
  2005/2006 $791,907
  2006/2007 $2,077,843
  2007/2008 $3,540,724
  2008/2009 $5,617,236
  2009/2010 $10,266,793
  2010/2011 $17,889,794
  2011/2012 $29,260,652
  2012/2013 $35,704,796
  2013/2014 $45,900,745
  2014/2015 $52,596,782
  2015/2016 $65,947,465
  
  SPENDING INCREASE (Dollars):
  2004/2005 $154,207
  2005/2006 $614,237
  2006/2007 $1,285,936
  2007/2008 $1,462,881
  2008/2009 $2,076,512
  2009/2010 $4,649,557
  2010/2011 $7,623,001
  2011/2012 $11,370,858
  2012/2013 $6,444,144
  2013/2014 $10,195,949
  2014/2015 $6,696,037
  2015/2016 $13,350,683

Even if the curve now linear, can anyone who was here a year ago honestly say that fifty-two million dollars wasn't enough, that the WMF wasn't getting the job done a year ago, and that the WMF needed to spend an additional thirteen million dollars to meet its goals? --Guy Macon (talk) 14:10, 28 February 2017 (UTC)

(..Sound of Crickets...) --Guy Macon (talk) 09:53, 1 March 2017 (UTC)

Yes, from the numbers it's more clear that the increase 2011-2016 would fit a linear not exponential ie. annual increase of $9M+-$2M . Of minor interest is that there appears to be some "tick-tock" structure 8,11, 6,10, 7,13 . Sorry, it really doesn't fit exponential at all. Widefox; talk 12:41, 1 March 2017 (UTC)

Can anyone who was here a year ago honestly say that fifty-two million dollars wasn't enough, that the WMF wasn't getting the job done a year ago, and that the WMF needed to spend an additional thirteen million dollars to meet its goals? --Guy Macon (talk) 14:37, 1 March 2017 (UTC)

User:Guy Macon I think it great to raise the issue in of itself Guy. I still wouldn't be in a position to comment on the run rate, or the subtext of underlying efficiency of any organization with just these figures per se. But, I can say, that in the context of such a major web service and platform, plus core costs around global content, community, communication, outreach, chapter grants etc the numbers don't seem order of magnitude wrong to me. Resource curse, triple bottom line etc are considerations. User:Tbayer (WMF)'s comment about technical debt is an interesting insight into workload (and to disambiguate: User:Piotrus that's standard software language "technical debt", unrelated to scaremongering "financial debt"). Widefox; talk 01:14, 2 March 2017 (UTC)

Transparency

I think this article reflects the kinds of perceptions that go into circulation as a consequence of current WMF practices in budget communication and outreach. If there is another story to be told or another perspective worth having, then I think the WMF should do more to make that story available to be heard. The Wikimedia community elects 5 of 10 seats on the Wikimedia Foundation board which means that the community has a major interest in being informed about the budget and talks about it often. I think that the community has continually requested more and better visualizations for understanding what the budget is, how it compares with similar organizations, what parts are more flexible, and what parts are and are not discussed with the Wikimedia community. If more, easier-to-understand information were available, then I think the community would use it. Conversations about the budget happen too often without the context of clear budget information. Blue Rasberry (talk) 12:36, 28 February 2017 (UTC)

I think users in the comments section bring up some very good points, and I think they complement this op-ed well. My personal opinion is that, while I'm not overly concerned about the spending patterns of WMF as long as they are reasonable (which all seem to be), I do think that an endowment should be one of the absolute top priorities for WMF. Endowments are one of those things that are nothing but beneficial when you get them set up nice and early. Icebob99 (talk) 14:28, 28 February 2017 (UTC)

I'm actually leery of WMF getting too big of an endowment. The problem is that Wikipedia already has a bit of the resource curse - we have a whole lot of traffic, and that is worth a lot of money, if corporations or lobbyists can insinuate themselves into key roles and take it over. Now because of that, a small endowment won't matter - but if the endowment is large enough that the WMF doesn't have to go back to the public at all, it could contribute to the curse and allow infiltrators to carry out an unrepentant subversion of the project's purpose. (The editors are also a check, of course, but the more resources are accumulated the smaller the relative power of new editing) If WMF sets up an endowment, they're basically telling donors look, we can be trusted to spend your money in the future more than you can be. People will tolerate that to a degree, since they can forget and exigencies come up, but I don't think that they should allow it to the point where WMF budgets everything out of interest income. Wnt (talk) 15:30, 28 February 2017 (UTC)

The WMF actually makes a huge amount of information available; in terms of transparency, as a non-profit that is not a public corporation, it is already a leading model. So we can refer to some of the data it publishes when having factual discussions about how the WMF commits its resources and where its increases in spending have historically occurred. See this chart, which has a convenient breakdown of expense by category from 14-15 projected to 15-16 plan:

This chart is a little more clear than other financial statements that may have a lower level of specificity, often combing technology, grants and other categories of expense into "programs." In another graphic provided by the WMF, we can see that the growth in expenses is roughly 20% year over year for several years when grant funding is excluded:

But when you put grant funding back in, the line is quite different (note the change in scale):

So we know already that grant funding is a huge and growing chunk of all WMF expenses, and I think what benefit the Wikimedia movement derives from these expenditures remains an open question. According to the WMF, engineering expenses (roughly 40% of the budget in recent years) include bandwidth, hardware, salaries for engineering and product staff as well as contractors and consultants, and "other miscellaneous costs." I haven't been able to track down, but would love to see, a more granular breakdown - how much of these engineering expenses go to maintenance and keeping the lights on, and beyond that (what has sometimes been called "Core" costs) how are the remaining expenses apportioned to specific projects? For example, how much did we spend on the Knowledge Engine? How much of that expense is ultimately of no value, and how much was re-purposed into other work on the (now "paused") Discovery team? The fundamental gap here is in determining and communicating the ROI for Wikimedia expenses; this Op-ed, along with many other comments by many other observers over the years, continues to demonstrate that this need is not being met. Nathan ^T 23:29, 28 February 2017 (UTC)

Discovery

I've been the lead product manager in Discovery right from the start, and I can say with confidence that there was no wasted value in terms of product or engineering work. Of course, there were designs and engineering experiments that ultimately were not turned in to products, and but no more so than other products I've worked on; iterating is a natural part of the process, and part of that occasionally includes throwing old work out and doing new work that is better. I should also point out that not all Discovery work is paused; work on search continues as normal, and right now we're actually working on getting maps deployed to more wikis before we halt development as the product has clear user value and is very popular with the communities in general. You can view our current goals on the Q3 goals page, and view historical goals all the way back to June 2015, which is a few months after Discovery was formed. Hopefully that is helpful. --Dan Garry, Wikimedia Foundation (talk) 09:32, 1 March 2017 (UTC)
- Thanks Dan, that is great to hear. I want to take this opportunity to offer feedback to you directly about your work - from my perspective and based on what I have been able to observe over the last decade, you have been a consistently fantastic Wikipedian and then Wikimedia employee and manager. I am always impressed by your ease and clarity in communicating with the community and contributors.Nathan ^T 23:19, 1 March 2017 (UTC)

@Deskana (WMF): Dan, I asked a question at mw:Talk:Wikimedia Discovery#Search prefix: but never got an answer. I understand that "old work was thrown out" (I'm unsure as to how well that "old work" worked, but the community seemed happy with it, as far as I can tell) without providing any replacement for that old work. I was asked to implement a way to search talk pages for old WP:Requested moves, and the clunky user interface at the top of WT:RM was the best I was able to come up with. As an aside, I thought the concept of "community engagement" was about finding better ways for the WMF to engage with their volunteer communities. It's taken me a while to understand that to the WMF, "community engagement" is more about finding ways to get the community to engage more with itself, or with the WMF projects (i.e., things). I think of engagement (diplomacy) rather than engagement (marketing). I suppose to TV programmers, "engagement" is all about inspiring people to engage more with their televisions. – wbm1058 (talk) 18:36, 1 March 2017 (UTC)

@Wbm1058: I think you've misunderstood what I wrote. I was talking about throwing out designs and engineering prototypes which never made it to production, not removing functionality from the site with no replacements as you implied. To my knowledge, Discovery have not removed any search functionality, with the exception of some small changes like implementing a maximum query length for performance reasons. Is there something specific that's been removed that you're referencing here? --Dan Garry, Wikimedia Foundation (talk) 22:42, 1 March 2017 (UTC)

@Deskana (WMF): Yes, there is functionality that was removed, as I explained on the Village Pump and on the talk page for the Discovery project. This functionality was implemented by Rainman in the Lucene-Search extension, but was dropped when search shifted away from Lucene-Search, best as I can tell from my research. Please read the previous discussions I initiated, which had limited participation. wbm1058 (talk) 22:51, 1 March 2017 (UTC)

@Wbm1058: Ah, so this is about the prefix pipes thing. That wasn't clear to me from your messages. Looking at that discussion, Cpiral seems to doubt that prefix ever supported pipes. I honestly wouldn't know whether it did or not, since all of this took place many years before Discovery was formed. So, no, Discovery did not remove this functionality, and this definitely has nothing to do with my first comment about "throwing out work". Regarding participation, indeed, feature requests that aren't posted in Phabricator, which is the proper venue for such requests, typically will get limited participation from the team. In contrast, we're normally very responsive on Phabricator. Please file your request in Phabricator and we can take a look. --Dan Garry, Wikimedia Foundation (talk) 23:03, 1 March 2017 (UTC)

@Deskana (WMF): I'm pretty sure that prefix pipes worked well enough at the time it was implemented, as if it didn't I wouldn't have found this barnstar left on Rainman's talk page:

		The Technology Barnstar
		This barnstar is for listening to our problem and so quickly coming up with a great Search Enhancement! Thank You!!! stmrlbs\|talk 05:59, 17 June 2009 (UTC)

I spent more time researching the history of this than Cpiral did, see also User talk:wbm1058#Templates for deletion for deletion. {{Search deletion discussions}} and {{Search prefixes}} are broken templates and I guess I'll eventually have to put them up for deletion if they can't be made to work again. Can't we just WP:IAR a bit, I've already put a lot of time into this, and don't want to go to the trouble of starting what would likely be just another WON'T_FIX. Can you locate Rainman's source code, I'm sure you know your way around that stuff better than I do. If you can find the code, maybe this can be implemented quickly. Rainman's initial release of this was pretty fast. If you can't find the code, then I'll just initiate further cleanup and deletion of User:Stmrlbs' templates. Oh, yes, I understood you. This case is a variant of throwing out your own stuff, called "not invented here syndrome", and "can't take time to understand the retired developer's code, because I prefer greenfield development". I'm that oddball type of developer who actually enjoys working on, and perfecting, legacy systems. – wbm1058 (talk) 23:55, 1 March 2017 (UTC)

So stumbled across this and I can give a bit of history. As with most things Cirrus, the answer to "why doesn't it do X like lsearchd did" is that lsearchd's behaviors were not very well documented so when we re-implemented them in Cirrus there was a lot of guesswork. I can confidently say that I did not know that prefix: even supported multiple arguments or I would've looked to replicate the behavior at the time (so yes, you can blame me and throw me in the stocks). The old code, if you're interested, is archived in Phabricator over here. Note that the code won't be able to just copied over (or even copied) as the logic for query handling is mostly in Cirrus now as opposed to in the backend itself--this was a conscious design decision at the time to allow us to mostly use Elasticsearch out of the box instead of maintaining our own custom Lucene-based systems. I don't want to speak for Discovery as that's not my team or project anymore, but it should be--at least in theory--possible to implement similar behavior again. I've taken the liberty of at least filing a task ^demon^{[omg plz]} 02:48, 2 March 2017 (UTC)

Thanks ^demon, or should I call you a fine daemon for your excellent background servicing of my request.

You wrote up the request in a nice concise way, better than what I would have done had I attempted to file the phab myself. The link I made in December is 404 now, but I observed "no commits between August 2007 and March 2010" so I think you're pointing at what I found back then. So where is the June 2009 change to support this? Not important though if as you say the code can't be reused. wbm1058 (talk) 05:41, 2 March 2017 (UTC)

Wbm1058, my assumption would be that the code for this never made it into a commit (the history was migrated from Subversion to Git) -- either that or it already existed and was enabled at a configuration level. Or...we're just looking in the wrong place! It could perhaps be handled in the PHP/MW side after all, in which case we would want to be looking at MWSearch instead. There's a couple of commits that look related but I'm going to bed now instead of digging any further :) ^demon^{[omg plz]} 06:45, 2 March 2017 (UTC)

@Guy Macon: Respect. The WMF's employees are convinced they are our bosses. In fact they are supposed to provide the services for us (that is, the creators of the Wikimedia sites) to work. Their fund-raising is on the basis of their supporting us. Anything beyond that, and especially the claim that we work under them, is grossly inappropriate and leads to what we see: their raising ever more money on false pretenses and harming the projects. The cancer metaphor is harsh but appropriate. Yngvadottir (talk) 23:57, 28 February 2017 (UTC)
- Re: "false pretenses", see [3]. --Guy Macon (talk) 09:58, 1 March 2017 (UTC)

Bariatric surgery?

I would not have gone with "cancer"; my preferred headline would have been "Wikipedia has overeating problem; May need bariatric surgery if it cannot control high-fat, caloric diet; Also must lay off the Tim Tams in any event." Guy points out a real potential problem, hopefully it does not become a real one. With proper management, it should be controllable. Having an endowment large enough to guard against revenue/expense swings is important.--Milowent • ^hasspoken 15:21, 1 March 2017 (UTC)

From what I understand spending is leveling this year and this topic has been discussed a fair bit at the WMF level. We will eventually know how much the prior ED cost (18 months after the most recent fiscal year) Doc James (talk · contribs · email) 02:27, 2 March 2017 (UTC)

It's worth noting that the WMF has recently hired a strategic planning firm, so much of these concerns may already be on the minds of the board. While I agree with many of Guy's insights the general thrust of his argument, the view is a tad alarmist in place. It's dubious that the WMF could be forced to sell off the encyclopedia. Because it operates under U.S. nonprofit law, its fiduciary practices provide some constraint on the way finances are handled. As best I know, the org has never been in "debt" (or even remotely close to it) and even under a catastrophic decline in revenue, could scale back its operations and remain operative. It's entirely common for the non-profit sector for (even healthy) organizations to expand and contract like that. It's jsut a reality, especially for those organizations that get the lion's share of their funding from donations, fundraising channels and grants--but if handled well, one can minimize the period of contraction. I think the cancer metaphor is a little heavy handed, but its clear the WMF has been in a crisis of planning for a while now, and there (which I think is the main point of Guy's arguments) I am in agreement. I also kind of agree that it sometimes feels like the outgoing grant money doth run a little too freely at times. Snow ^{let's rap} 09:01, 2 March 2017 (UTC)

Reconsider the Title, Please

I have been asked by a current employee of the WMF who wishes to remain anonymous that the title of this article be reconsidered. They are a cancer survivor, and I can understand why they might prefer not to be declared equivalent to the thing that almost killed them.

Aside from that, inflammatory, low-rent article titles like this only help serve the idea that the Signpost has continued its relentless trek towards irrelevance. Keepin' it classy, I see. --Jorm (talk) 18:19, 1 March 2017 (UTC)

I'm reminded of the inflammatory, low-rent campaign of Donald Trump. You can dismiss the community of us older white guys as a bunch of deplorable people, or you can engage in diplomacy. wbm1058 (talk) 18:55, 1 March 2017 (UTC)

Jorm, I've asked the team to consider this request, and will return to it after hearing some input. To give it full consideration, we'll need at least 12 hours, as we are volunteers spread across many timezones. As always, we welcome op-ed submissions from community members reflecting any perspective pertaining to Wikipedia or Wikimedia. -Pete Forsyth, Editor in Chief (talk) 19:06, 1 March 2017 (UTC)

(EC) Both the cancer analogy and the hyperbole about exponential growth were challenged in the drafting stage. The result is what it is - an opinion expressed in a knowingly exaggerated and offensive way. As the author put it "I doubt that a cancer alogy will cause as much offense as our images of Muhammad". We are in an era of hyperbole not diplomacy. If you want to encourage this sort of Trump like communication then treat this sort of Op-Ed seriously. Ϣere SpielChequers 19:19, 1 March 2017 (UTC)

Peteforsyth I am confused as to what the purpose of having editorial oversight is, then. --Jorm (talk) 19:32, 1 March 2017 (UTC)

If your confusion goes beyond what's explained in the lead sentence of op-ed, I'll need a clearer articulation of what it stems from. -Pete Forsyth (talk) 19:38, 1 March 2017 (UTC)

I am confused as to why you need to wait 12 hours to make a change to the text. You are the editor, yes? Edit.--Jorm (talk) 19:48, 1 March 2017 (UTC)

Ah, I see -- thanks for clarifying. This is how I edit. I find the request, which is for a significant change, to be reasonable enough to consider. Consideration will take a little time. -Pete Forsyth (talk) 20:11, 1 March 2017 (UTC)

Hey, it's only the integrity of your paper you've got up on the block, but okay. Take your time. --Jorm (talk) 20:37, 1 March 2017 (UTC)

I would like second the request about the title, but for a completely different reason. Yes, it is an attention-grabber, but it also produces a knee-jerk overdefensive emotional reaction from the criticized side, rather than a thoughtful response to criticism. I remember a while ago something similar happened to some Signpost article where the author deliberately chose to use a foul language. As a result, AFAIR the talk was 80% discussion of the language rather than of content. (And today this is the only thing I remember from the article :-) Staszek Lem (talk) 19:28, 1 March 2017 (UTC) ┌────────────────────────────────────────────────────────────────────────────────────────────────────┘ I have given careful consideration to this and am I will not agree to change the title.

Anyone who is offended by cancer analogies should stop reading at the title. There are many things on Wikipedia that some people find deeply offensive, images of Muhammad being just one example. We don't censor Wikipedia because someone is offended.

To properly remove the cancer analogy, I would have to rewrite the op-ed with another analogy, which itself will offend some. Should I use obesity? AIDS? Rats multiplying and eating all the grain?

Given the many previous failed attempts by many people in many places to get the WMF to change the way they do things, I simply do not believe that I will get a "thoughtful response to criticism" from the WMF no matter what analogy I use.

I completely reject Jorm's characterization of the title, and note that Jorm has a clear COI as a former WMF employee and developer and a long history of engaging in personal attacks towards me. I really don't want to have any interactions with Jorm.

I know that some people want to make everything be about Donald Trump, but that doesn't mean that I have to respond. I refuse to address any comment that references Trump.

Re: "I've asked the team to consider this request", No. The Signpost editorial team does not have my permission to change the title (unless of course someone puts forth a compelling argument and I agree to the change). Several changes were requested prior to publication, and I agreed to every change. Now that my words have been approved and published, you are free to delete the op-ed. You are free to write your own op-ed under your name and replace the one I wrote (including a note that you have done so and why would be appreciated). You do not have permission to change my words other than fixing obvious errors or formatting. Not while my name is on them.

BTW, I am a cancer survivor myself. --Guy Macon (talk) 20:07, 1 March 2017 (UTC)

Cool story, bro.--Jorm (talk) 20:37, 1 March 2017 (UTC)

Speaking about potentially offensive comments - he says he's a cancer survivor, and your response is "Cool story, bro"? Hopefully readers can interpret his use of the analogy as a useful and illustrative analogy of the problem he is describing. I'm not sure the same redeeming value can be found in your response. Nathan ^T 23:32, 1 March 2017 (UTC)

Despicable comment. Guy: Excellent article. I too am a cancer survivor and feel you would be minimizing cancer awareness by removing it from the title, which offends me. [4]. D.Creish (talk) 00:55, 2 March 2017 (UTC)

Re: to get the WMF to change the way they do things - Are there any mechanisms at all? Staszek Lem (talk) 20:21, 1 March 2017 (UTC)

Well the next board elections are in about three months. See Meta:Wikimedia Foundation elections/Board elections. Ϣere SpielChequers 11:28, 2 March 2017 (UTC)

I find the objection of this section by an "anonymous employee" to be at best overly sensitive and at worst manipulative. There is nothing wrong with using cancer as a metaphor, especially in the sense used by the author as a deep concern about the future of the project that could lead to its demise. The author wrote a well-thought out piece and gave a thesis that suggests the aptness of the metaphor. As many of us have dedicated a large chuck of our lives to the project, the very idea that the fruit of our volunteered labor could end up in the hands of a for-profit is abhorrent. Cheers to Guy for trying to bring attention to this issue. Tell the employee that despite their personal battle, the author clearly is not disparaging those who have battled a biological form of cancer and is therefore being too sensitive. The "I find your metaphorical use of cancer" angle is an easily abused appeal to emotion that can be used to censor an unwanted point of view; so I am not convinced there's negligible chance the employee's objection is without ulterior motive. Jason Quinn (talk) 06:53, 26 March 2017 (UTC)

WMF Board

Don't get your hopes up regarding the board. As a test of our often-claimed "financial transparency", in July of 2015 I asked the board to answer a simple, noncontroversial question about spending. I am still waiting for an answer -- any answer, even if the answer is "we refuse to tell you" -- to the following question:

"I would really like to see an itemized list of exactly what computer equipment and office furniture was purchased with the $2,690,659 spent in 2012 and the $2,475,158 spent in 2013. Verifying that those purchases were reasonable and fiscally prudent would go a long way towards giving me confidence that the rest of the money was also spent wisely. As a non-profit that accepts donations and is committed to openness, the WMF should have no problem with telling me exactly what was bought, the price paid, and in general where it went. No specific names, of course, just a general description of which department got what."

See:

Please note that the actual question I asked is not important. I chose it to be easy to answer. I encourage anyone reading this to pick an easy-to-answer question of your own and to try to get an answer from the WMF. --Guy Macon (talk) 14:38, 2 March 2017 (UTC)

Signpost response to title request

In response to the request relayed by Jorm above, to change the title of this op-ed:

I am sad to learn that the title we ran pains a reader. It is never our intent at the Signpost to cause distress or ill feelings. While it may happen from time to time in the course of vigorous debate, it's never something we relish, and should be avoided whenever possible.

In this case, the article in question is an op-ed, and intended principally to allow the author to express a point of view, with no implication of endorsement or agreement of the Signpost team. Even so, there are lines we shouldn't cross, and causing needless harm or distress is one of them. The possibility of this outcome was discussed by others ahead of publication; I reviewed and considered what was said there prior to publication and, for better or worse, chose to heed the one comment that cited personal experience, which also happened to align with my own experience and instinct. I felt that the intent of the metaphor was clear, if stark; and that the biological underpinning of the point would outweigh any personal trigger for our readers.

If and when I face such a decision again, I will proceed with more caution. However, in this case, we will not be making the requested change. There are two reasons:

Making any substantive change after publication is something we only do in extreme cases. The bar is even higher when it involves the title of an article. "Extremity" always involves a judgment call. In this case, we have not heard from multiple parties; and the one complaint we have heard appears to involve a misunderstanding of the author's thesis. I do not read in this piece that the WMF is itself comparable to cancer, but that its spending patterns may be comparable to cancer. That offense was taken is real and legitimate, and something I regret; but the offense can't be fully undone; moreover, I see no evidence that it might extend to a large number of readers.
We never expect to fully agree with the authors of op-ed pieces. The title was written by Guy, and he has a strong preference to keep it. We don't endorse any of the reasons he presents above, but his preference in itself is a significant factor. (As a side point, Guy has not responded to any of the three private emails we sent in the process of composing this op-ed, including a request to discuss this specific request. All of his communications to us are part of the public, on-wiki record.)

On the overall suitability of this piece for the Signpost: in general, we seek to run op-eds that stimulate worthwhile discussions, without regard to whether or not we agree with their positions. We should not do so, however, at the expense of anybody's well-being, or at the expense of fairness or accuracy. While I'm generally pleased by the depth of discussion above, I do find WereSpielChequers' point compelling; we hope never to exclude any voices due to a sense that the Signpost offers an unsuitably serious forum for discussion.

For anyone wishing to demonstrate that non-confrontational, newsworthy stories are worthy of discussion, might I suggest the current edition's Recent Research, which features reviews of several interesting studies about Wikipedia and education. Thus far, first post is up for grabs. -Pete Forsyth (talk) 08:02, 2 March 2017 (UTC)

Good going! You've shown such courage here. I wish you all the success in turning this "paper" into the wiki equivalent of Brietbart, as that appears to be your goal.--Jorm (talk) 16:38, 2 March 2017 (UTC)

I would advise ignoring Jorm. He is clearly trying to start a fight. He has been following me around calling me names ever since I expressed the opinion (an opinion shared by many others) that the WMF has a history of creating low quality software. As a former paid WMF software developer and a big fan of WP:Flow [5], he naturally takes exception to that opinion. His snarky comments about The Signpost are, IMO, just an extension of his WP:BATTLEGROUND mentality regarding the quality of WMF's software. I think http://mollywhite.net/wikimedia-timeline/ tells us pretty much everything we need to know about that era of Wikipedia's history. --Guy Macon (talk) 17:05, 2 March 2017 (UTC)

I support Pete's decision here. While the author doesn't want to be compared with Trump/Brietbart et al. I think the analogy is apt. Perhaps another analogy is to WP:Fringe. We should be very careful about where we draw the line between acceptable, publishable minority opinion and fringe opinions of crazy persons that should be censored. The WMF may dismiss this opinion piece as "not representative of the community as a whole" at their peril. Trump was dismissed that way, and look where we are now. – wbm1058 (talk) 16:21, 3 March 2017 (UTC)

And this "paper" could only become the wiki equivalent of Brietbart if it began rejecting mainstream opinions and facts as "fake news". wbm1058 (talk) 16:33, 3 March 2017 (UTC)

A word, if I may

Editors are perfectly within in their rights to criticize Pete Forsyth and the editorial board of the Signpost when they make a bad call, but please keep in mind that running the Signpost is a tremendous job and we should be grateful to anyone willing to take on that role. It is a role requires dozens of decisions every week, and they aren't going to get all of them right. Please remember that one article you do not like does not represent the entirety of their work and this publication, or even of a single issue of The Signpost. The editors working on Traffic Report or Recent Research, for example, didn't compare anyone to cancer and deserve acknowledgement for their hard work.

Guy Macon, considering that you compared an entire organization of people to cancer, I'm surprised the reaction has been as muted as it has been, and it was certainly much less than the reaction we got last year merely for putting some swear words in an op-ed. If you consider Jorm's mild rebukes to be personal attacks, then the op-ed business is not for you, you sweet summer child.

I am, as you know, a big fan of hyperbole and attention-getting gestures, so I am sympathetic (to a point) to the failed attempt here at message delivery. But flamboyant deliveries have to be coupled with reasonable, coherent messages and informed opinions. I don't expect everyone on the Signpost to know what a Form 990 is, but I do expect someone writing about WMF spending, salaries, and transparency to be aware of them, otherwise why am I reading their opinion on this topic? This is a very minor quibble compared to the counter-productive message here, which I think is just one example of many that I see in community forums by people who appoint themselves voices of that community. If someone came up to you on the street and called you a cancer, would you listen to them? No, because they would sound like a crazy person and you don't listen to crazy people. If you care about the WMF being responsive to the community, you should consider the effect of having community members respond to them in the manner of a crazy person. The WMF should, as stewards of a community resource, always be responsive and answerable to the community, but the organization is also full of human beings who will have a natural, human tendency to respond to crazy people as if they are crazy people. Crazy behavior has a deleterious effect on the relationship between the WMF and the community, discourages transparency and responsiveness, and encourages the WMF to write off the sane majority of the community based on the behavior of loud, attention-getting outliers. What do you care about more? Do you care about having a productive relationship between the WMF and the community, or do you just care about getting attention for yourself? Gamaliel (talk) 02:55, 3 March 2017 (UTC)

You may not consider calling someone an asshole [6][7] and a liar[8] a personal attack, but I do.

A productive relationship between the WMF and the community is impossible at present, because the WMF does not respond[9][10] in any way to any suggestions or criticisms, no matter how they are phrased. Martijn Hoekstra has been unfailingly polite and restrained,[11] yet he also has never been able to engage in any discussion with anyone at the WMF regarding his complaints.

I encourage you to do your own experiment. Pick any subject area, large or small, regarding how the WMF does things. Ask any question with an answer that isn't already published. Avoid all of the things that you think are wrong about my approach. Post a link where they engage in any sort of dialog with you. You will be stonewalled. Do the experiment and prove me wrong. --Guy Macon (talk) 04:54, 3 March 2017 (UTC)

I believe you are confusing "This individual did not receive the specific response they demanded" with "The WMF is not responsive to the community as a whole." Gamaliel (talk) 15:33, 3 March 2017 (UTC)

Good job setting up a straw man and knocking it down, but my actual claim was that no individual who asked a similar question has received any response. My claim may be wrong, but it was not unclear. I was quite clear in saying that I was asking for any response, even if the response was "go away". I have asked for examples of anyone asking a similar question and getting a response and no example have been forthcoming, plus I encouraged you to do your own experiment and see if you get a response. --Guy Macon (talk) 16:14, 3 March 2017 (UTC)

Frequently Unanswered Question

For those following along at home, try asking the following question:

In the following document...

~~https://upload.wikimedia.org/wikipedia/foundation/b/bf/Audit_Report_-_FY_13-14_-_Final_v2.pdf~~

https://upload.wikimedia.org/wikipedia/foundation/6/6e/FINAL_12_13From_KPMG.pdf

...on page 10 12 (~~10th~~ 12th page of the document: they numbered it ~~"8"~~ "10") it says "Furniture: 2013=$439,562, 2012=$277,312"

Please provide a detailed accounting of how that $439,562 and $277,312 was spent.

Ask the question wherever you think best, using whatever wording you think best.

See if you get an answer, even a "no, we will not release that information" answer. --Guy Macon (talk) 16:25, 3 March 2017 (UTC)

I believe you mean page 11, and you're off by a year in terms of which report you copied those figures from. wbm1058 (talk) 17:00, 3 March 2017 (UTC)

I apologize for the error. I have stricken the incorrec information above and added the correct information. --Guy Macon (talk) 04:40, 8 March 2017 (UTC)

The data follow as well the exponential growth part of a sigmoid Logistic function. Yes, an indefinite exponential would eventually eat up all the resources of the USA and later of the Earth, thus exterminating humanity and saving us from global nuclear war or whatever. So, WMF spending growth is likely to go linear after another decade and someday stop. And eventually decline as the 18th century concept of an encyclopedia distinct from other literature falls victim to some wonderful new development. The questions for the next decade are, when and how will the belt be tightened? We don't always have free pizza at my local chapter meeting, and before mid-century it will probably be a fond memory, but who's going to look decades ahead in a high tech industry? Jim.henderson (talk) 19:54, 3 March 2017 (UTC)

@Guy Macon: To answer your question: The money spent on furniture was to upgrade from a 1980s cubicle office to a semi-modern open floor-plan office. The WMF is not going to give you "detailed accounting" of what furniture it buys, however. Frankly such a request seems more like sea lioning than constructive questioning. Your criticism of the WMF's overall budget increases is a valid point (although perhaps a year or two outdated), and is a worthy topic to discuss, but I think you're shooting yourself in the foot with this furniture crusade. Have you seen photos of the WMF office? It certainly isn't extravagant. Kaldari (talk) 22:30, 7 March 2017 (UTC)

You have *cough* an interesting definition of "answer to your question". --Guy Macon (talk) 01:19, 8 March 2017 (UTC)

Also, you do realize that Jorm hasn't worked for the WMF since 2014, right? Kaldari (talk) 22:33, 7 March 2017 (UTC)

If you think asking Jorm's replacement will get you an answer, go for it. After about the tenth time someone who is not a WMF employee suggested that I ask somewhere else -- only to recieve the same stonewalling I always get -- I gave up. --Guy Macon (talk) 01:19, 8 March 2017 (UTC)

Kaldari, it is actually you who is shooting yourself in the foot with your answer. You say "It certainly isn't extravagant." Then an interesting question is where did "$2,690,659 spent in 2012 and the $2,475,158 spent in 2013" go. For $5M I can quite lavishly sit about 1,000 personnel with $5K per face. Staszek Lem (talk) 00:27, 8 March 2017 (UTC)

Good point. The basic issue here is one of trust and of the WMF doing what they say they will do. Trust: if we can see what furniture was bought and how much was spent on it, we can determine for ourselves whether the WMF is wasting money. If the money spent on furniture in 2012 and 2013 was reasonable, that will give us all good reason to trust that the WMF is spending wisely in other areas. The WMF doing what they say they will do: I can give you several quotes from WMF leaders promising a commitment to financial transparency. I am simply asking them to act in accordance to those commitments. --Guy Macon (talk) 01:19, 8 March 2017 (UTC)

Why do you think that $5M is an unreasonable amount to refurnish a 3 floor office? What is your basis of comparison? What other companies have done refurnishings at a lower cost per capita? Kaldari (talk) 02:01, 8 March 2017 (UTC)

Collapsed a late-night ill-thought attempt of a joke. Apologies. Staszek Lem (talk) 18:40, 8 March 2017 (UTC)

Kaldari, are you a Jew? Staszek Lem (talk) 03:28, 8 March 2017 (UTC)

...what the fuck, dude?--Jorm (talk) 04:12, 8 March 2017 (UTC)

Sigh... Here we are, having a calm reasoned discussion. There is some vigorous disagreement, as one would expect from the subject matter, but pretty much everyone on both sides wants what is best for Wikipedia, even as we disagree as to what that is. Suddenly, we get some racist trolling. I am not going to suggest completely ignoring the troll, but might I suggest that each editor react no more than once and then ignore him? I say we leave him shouting into an empty hall. Here is my only and last comment on this: Staszek Lem, I refer you to the reply given in the case of Arkell v. Pressdram. --Guy Macon (talk) 04:40, 8 March 2017 (UTC)

Comment struck. this appears to have been a joke that I did not "get". --Guy Macon (talk) 19:43, 8 March 2017 (UTC)

racist trolling - wrong answer. The correct answer would be: "Why are you asking?" The I would answer: "Why do you want to know?" - followed by: "Are you an antisemite?" answered by another question ... ad infinitum. Staszek Lem (talk) 17:41, 8 March 2017 (UTC)

wtf dude?" - my question is a well-known (looks like not) element of Jewish humor, a stereotype that answering a question with a question is a Jewish trait. If you think that it is an insult, then you are racist, because actually asking a clarifying question is a trait of a smart person who, before jumping to a wrong answer, tries to clarify the question. Of course, this penchant to scholarship is a subject of derision by anti-Semites. <sigh> Staszek Lem (talk) 17:41, 8 March 2017 (UTC)

It is not well known, dude, and if you are calling me an anti-semite, or even fucking implying at that, you can go fuck yourself and we can have words that will end in fisticuffs.--Jorm (talk) 18:05, 8 March 2017 (UTC)

Kaldari, you know that they spend that money to refurnish a 3 floor office -- how? Nobody here knows what that money was spent on. Maybe they were fiscally prudent and got a lot for their money. Maybe they overpaid. Maybe a big chunk of the money went to redecorating some WMF executive's home. We simply do not know, because the WMF will not give us a detailed accounting. We just have to trust them. --Guy Macon (talk) 04:40, 8 March 2017 (UTC)

Kaldari, Firstly, may I assume that the above is not an official WMF answer telling me that they refuse to give me a detailed accounting? If you are an official WMF spokesperson, please say so, otherwise my point that the WMF refuses to answer reasonable questions -- even if the answer is "no, we will not tell you that" -- remains. Secondly, is it any question asking the WMF for details on how they spend money that you are calling sea-lioning, or is it just this one question? If the latter, be advised that I chose the question to be easy to answer and uncontroversial and feel free to suggest a different question about spending that you believe the WMF will answer. --Guy Macon (talk) 01:19, 8 March 2017 (UTC)

No I'm not an official WMF spokesperson. I think where we differ is in our opinion of what constitutes a "reasonable" request. The WMF publishes its budget every year and unless there is an actual discrepancy, I don't think it makes sense for them to spend time creating itemized accounting based on the requests of individual community members. The WMF doesn't micro-manage the community and the community shouldn't micro-manage the WMF. That isn't a way to build trust, it's a way to build frustration and antagonism. Kaldari (talk) 02:01, 8 March 2017 (UTC)

I don't think it makes sense for them to spend time creating itemized accounting based on the requests of individual community members - (A) itemized accounting is a normal way of accounting, as opposed to sloppy accounting. (B) I dont think that WMF is overwhelmed with "requests of individual community members". In the case if they do, their job is to classify these requests, sort out most common ones and post a kind of FAQ, unless "community members" is ~~cattle~~ wordless slaves for them. Staszek Lem (talk) 03:33, 8 March 2017 (UTC)

I second what Staszek Lem writes above. If the WMF has to do any actual work to give an itemized list for "Furniture: 2013=$439,562, 2012=$277,312"[12] instead of simply cutting and pasting the answer from the existing detailed accounting, that would imply that the WMF does not keep track of basic accounting like confirming that a desk arrived after one was ordered and paid for. Which I don't believe for a second. --Guy Macon (talk) 19:43, 8 March 2017 (UTC)

I'm puzzled as to why nobody from (WMF) is defending themselves here, so I will try. It isn't that difficult to understand these financial statements if you've at least taken an Accounting 101 course. I've never worked as a professional accountant, but I did take several accounting courses as a college student. The first thing you need to understand is the difference between an income statement, a balance sheet and a cash flow statement. These are basic financial statements. When you ask the question in bold, "Please provide a detailed accounting of how that $439,562 and $277,312 was spent", you are exposing your lack of knowledge of basic accounting. You're implying that $439,562 was spent on furniture in 2013 and $277,312 was spent on furniture in 2012. This is wrong. These figures are provided as a more detailed supplement to the balance sheet line item "Property, plant, and equipment, net". So, as of June 30, 2013 the WMF had spent $439,562 on all of the furniture it owned as of that date. This is before accumulated depreciation. I don't have the figure for the number of employees on June 30, 2013 but assuming it was about 200, then the furniture per employee was $439,562 ÷ 200 or about $2200 per employee. Taking the difference between $439,562 and $277,312 I find that a net of $162,250 was spent on furniture in 2013. They may have spent more than this if some of the furniture owned on June 30, 2012 had been sold or discarded. The upgrade from a 1980s cubicle office to a semi-modern open floor-plan office seems like a reasonable explanation for this. I suppose it's reasonable to debate the need for this upgrade. As I type this, I'm sitting on furniture I bought in the 1980s. I don't think it's reasonable to speculate that a big chunk of the money went to redecorating executives' homes. I think we discussed earlier that roughly $2K wasn't a totally invalid figure for the cost of a professional workstation, and there's probably some office furniture for common areas, too. Regarding, "where did "$2,690,659 spent in 2012 and the $2,475,158 spent in 2013" go", these cash flow figures are for purchase of computer equipment and office furniture, and the vast majority of this is for computer equipment. Back to the balance sheet, $9,504,243 less $8,338,207 equals $1,166,036 net spending on computer equipment in the year ended June 30, 2013. $1,166,036 for computers plus $162,250 spent on furniture totals $1,328,286 net spent on computer equipment and office furniture. The cash flow statement shows $2,475,158 total spent on both computers and furniture – it doesn't break out how much on each. Presumably a lot of the computer equipment are servers that aren't even located in San Francisco. So I'm not sure about the difference between $2,475,158 and $1,328,286. Presumably that's old furniture and computers that were either sold off (would be included in the line "proceeds from sales" or donated to another charity or recycled or thrown out. wbm1058 (talk) 20:50, 8 March 2017 (UTC)

Thank you very much for the detailed write-up. I understand it took you some time. And of course it may be a burden for a WMF employee. In fact the question was much simpler: "how much was spent on furniture". And the answer could have been much simpler: "Sorry you misread the meanings of the $$$$$$: it was equip+fun+else. The turn itself was only $$." And of course ~2K per employee looks reasonable. And we shouldn't have had all this nasty bickering wasting many people's time and spreading "false news" :-), if only a WMF person ha some decency to spend 15 minutes for an answer. Nobody asked for something stupid as itemizing each sofa and coffee table. Staszek Lem (talk) 02:35, 9 March 2017 (UTC)

Odd. When I asked the WMF treasurer, he seemed to think it was the numbers were spending, not assets.[13][14][15] That being said, Wbm1058 could very well be right. Or it could be that we somehow endup looking at different kinds of financial documents. So let me revise the question: We spent some unknown amount on furniture in 2012 and 2013. How much did we spend? What furniture did we purchase? --Guy Macon (talk) 04:58, 9 March 2017 (UTC)

20 April 2015 Signpost Op-ed discussion: Of the amounts quoted most of the money was spent on servers and other equipment for the data center. The next biggest amount was for computer equipment and software for staff. Only about 9% of the amount in 2013 and about 7% of the amount in 2012 was spent on furniture and fixtures.GByrd (WMF) (talk) 18:01, 20 April 2015 (UTC)

Meta-Wiki, 31 May 2015 I repeated this to you. He never seemed to think the numbers were spending, not assets – only you did. There may have been others who were similarly confused; I recall a somewhat misleading blog post and/or criticism site discussion too.

Doing the math.

$2,690,659 spent in 2012: about 7% of the amount in 2012 was spent on furniture and fixtures. $2,690,659 × 0.07 = $188,346

$2,475,158 spent in 2013: about 9% of the amount in 2013 was spent on furniture and fixtures. $2,475,158 × 0.09 = $222,764

Compare with my calculation above of a net increase of $162,250 in the total cost of furniture in use in 2013 on the balance sheet. So there must have been about $60,000 worth of old furniture that was retired from service.

Wiki version of the finacial report: Upon retirement, sale, or other disposition of property, plant, and equipment, costs, and accumulated depreciation are eliminated from the accounts and any resulting gain or loss is included in operations.

Hopefully, after nearly two years, we can finally put this one to rest. Or, by "What furniture did we purchase?" are you now going to insist on an itemized list of each sofa and coffee table? wbm1058 (talk) 07:24, 9 March 2017 (UTC)

No. I am not going to stop asking. No, I am not going to pretend that a non-answer by an ordinary Wikipedia editor somehow makes up for the continued refusal to answer by the WMF. The question of what the money was spent on is an important question. As I first asked two years ago, some people have, quite reasonably, asked "where does the money I donate to the Wikipedia Foundation go?" Verifying that the money spent on furniture purchases in 2012 and 2013 were reasonable and fiscally prudent would go a long way towards giving me confidence that the rest of the money was also spent wisely. As a non-profit that accepts donations and is committed to openness, the WMF should have no problem with telling me exactly what was bought, the price paid, and in general where the furniture was installed. I shouldn't have to repeat this, but I have always been crystal clear that I will accept any answer from the WMF, even if the answer is "no, we will not tell you where that money was spent".

I have no problem with using your numbers for the total amount spent on furniture. Just to confirm, Is is your claim that in 2012 the WMF spent $188,346 on furniture and in 2013 the WMF spent $222,764 on furniture? If you really meant "furniture and fixtures", do you have a definition for "fixtures" and some idea as to how much was spent on furniture? --Guy Macon (talk) 15:03, 9 March 2017 (UTC)

It is not my claim that in 2012 the WMF spent $188,346 on furniture and in 2013 the WMF spent $222,764 on furniture; it is the claim of former Chief of Finance and Administration GByrd (WMF) that about that much was spent, give or take a few thousand. He gave his answer to the nearest percent; one percent of $2,690,659 is $26,907 so you can assume they spent $188,346 ± $26,907. Fixtures would be stuff like the lights in the ceiling, drinking fountains, toilets and sinks. 19 June 2015: I have responded to the question with information that most of the money listed in "Purchase of computer equipment and office furniture" is for servers. This information is derived from our audited financial statements. If there were any issues on any of the purchases of computer equipment and office furniture then it is my responsibility to report such issues to the Board of Trustees. We have found no such issues and I have not received any inquiry from the Board of Trustees on this issue.GByrd (WMF) (talk) 00:30, 19 June 2015 (UTC)

I take that answer as "no, we will not tell you in further detail than we already have where that money was spent".

How will you verify that the money spent on furniture purchases in 2012 and 2013 was reasonable and fiscally prudent without personally reviewing the purchase receipts and doing a physical inventory of the furniture and fixtures at the San Francisco headquarters? You want to make sure all the furniture purchased was actually delivered to the right place, right? Is an $800 workstation reasonable? How about an $1100 workstation? What about a $2000 workstation? Where do you draw the line? Do we need a community consensus to determine whether each individual item in their furniture inventory was a prudent purchase? Basically I hear you saying that you want an audit of the furniture. I believe that would be the remit of the audit committee of the Board of Directors. You could run for that, but my guess is that you would get booted from the board before they'd elect you to serve on the audit committee. I could be wrong about that though. wbm1058 (talk) 17:53, 9 March 2017 (UTC)

No. I do not require an audit or a physical inventory. Someone -- anyone -- who works for the WMF simply saying that the furniture arrived (or that they have no idea whether it actually arrived) would be quite sufficient. Where do I draw the line? I draw the line at asking the following quite reasonable question: We spent some unknown amount on furniture in 2012 and 2013. How much did we spend? What furniture did we purchase? If the WMF answers those questions I will have nothing further to say but "thank you". --Guy Macon (talk) 20:22, 9 March 2017 (UTC)

The amount is not unknown. You have been given a ballpark answer. Another accounting concept comes into play here: materiality. As a simple example, an expenditure of ten cents on paper is generally immaterial, and, if it were forgotten or recorded incorrectly, then no practical difference would result, even for a very small business. However, a transaction of many millions of dollars is almost always material, and if it were forgotten or recorded incorrectly, then financial managers, investors, and others would make incorrect decisions as a result of this error. The assessment of what is material – where to draw the line between a transaction that is big enough to matter or small enough to be immaterial – depends upon factors such as the size of the organization's revenues and expenses, and is ultimately a matter of professional judgment. GByrd gave you what he judged to be a sufficiently precise amount, based on his professional judgment. What would it take to satisfy you. A figure to the nearest $1000? Or do you need to know the exact dollar amount? Surely you know what kinds of furniture furnish a typical software development office. Again, is it really material to know how many chairs and flowerpots they bought? wbm1058 (talk) 22:01, 9 March 2017 (UTC)

Re: "The amount is not unknown. You have been given a ballpark answer"; I am going to keep asking, so do I have your permission to replace "and unknown amount" with "Approximately $188,346 ± $26,907 (estimate from Wikipedia user Wbm1058, who is not a WMF employee)"? and direct anyone who thinks the numbers are wrong to you? Or, if you think that "GByrd gave me what he judged to be a sufficiently precise amount" for the amount spent on furniture in 2012 and 2013, just post the numbers and a link where he gave me those numbers and I will use them.

Re: materiality; I have no idea why you think that there might be a discrepancy. A discrepancy requires two sets of numbers that do not match. I am asking for a single set of numbers from the WMF (not you) and promising to not dispute whatever numbers they give me.

Re: "Again, is it really material to know how many chairs and flowerpots they bought?" Flowerpots no, chairs yes, unless flowerpots are furniture. How many times have you asked me the above question now? Three? Four? The answer is not going to change, so please stop asking. How much we spent on furniture and an accounting of what furniture was purchased is designed to be a spot check. If the WMF was fiscally prudent regarding furniture purchases, we can presume that the` are being fiscally prudent in other areas. Right now it really looks like they have something to hide, all the while bloviating about "transparency".

You misunderstood the concept of materiality. If they had entirely omitted furniture purchases from their report of cash flows, that would be a material omission. If they reported $400,000 in furniture purchases but neglected to report that they spent $1,000 on decorative plant stands, then that's probably not a material omission. Let's assume that they spent a total of $411,000 on furniture in those two years, for a significant workstation upgrade. Now they have not given you a precise figure, just given you a percentage to the nearest percent. So our assumption may be a bit off. Maybe they spent $430,000. That's not a material difference. You can't say that $420,000 is a prudent and reasonable amount to spend on furniture in two years, but that $430,000 is not prudent, because the difference between those figures is not material in respect to the overall budget. Adding up my calculations from above, $188,346 + $222,764 = $411,110 ÷ $439,562 balance sheet cost of all furniture, I find that 93.5% of all furniture at the end of the reporting period was purchased within those two years. So it does appear to be a near total replacement of furniture. It's possible they bought more workstations than the number of current employees, in anticipation of future staffing growth. You wouldn't want to hire someone without already having an empty desk to assign them. Let's assume then their answer is either (A) we bought 200 workstations, @ $2100 each, total $420,000 or (B) we bought 400 workstations, @ $1050 each, total $420,000 or (C) we bought 50 workstations, @8400 each, total $420,000. And regardless of their answer, upon seeing that, you would say "thank you" and quietly walk away, regardless of whether the answer was A, B or C. I don't get the point, as usually you would ask questions like this in the course of decision making, and then make a decision as to how to respond after evaluating the answer to your information-gathering question. The idea that you are doing this as a "spot check" on the suspicion that they "have something to hide" belies the idea that you will have nothing further to say but "thank you". wbm1058 (talk) 17:40, 10 March 2017 (UTC)

The reason I say that` if I get any answer I will say nothing more other than thank you is because of the multiple false accusations that incorrectly claim that I will not accept any answer. I will accept any answer that tells me how much we spent on furniture in 2012 and 2013 and what furniture was purchased. I trust the Wikipedia community to evaluate the answer and to ask any followup questions needed. I just want to see whether the WMF actually has a commitment to financial transparency. --Guy Macon (talk) 21:48, 10 March 2017 (UTC)

You are never going to get an answer from anyone at WMF about this because you are a) Clearly wrong, as anyone who knows about this stuff has been telling you; b) clearly won't accept any answer you're given (you've been answered now multiple times by my count); c) you're aggressive and hyperbolic about every engagement; and d) you've probably (read: absolutely) been thrown into the "Crank File". Congratulations! You played yourself.--Jorm (talk) 17:21, 9 March 2017 (UTC)

Go away, Jorm. I do not want to have any interactions with you and I will not respond to anything that you write. --Guy Macon (talk) 20:22, 9 March 2017 (UTC)

Reconsidering the title

Is cancer the best analogy? Consider the alternative, "Wikipedia has a parasite".

Parasitism is a non-mutual relationship between species, where one species, the parasite, benefits at the expense of the other, the host
Parasites typically do not kill their host, are generally much smaller than their host, and will often live in or on their host for an extended period
Parasites reduce host biological fitness by general or specialized pathology
Parasites increase their own fitness by exploiting hosts for resources necessary for their survival
In many cases, it is difficult to demonstrate harm to the host
Parasites show a high degree of specialization, and reproduce at a faster rate than their hosts
Parasitism is part of a continuum of types of interactions between species
Parasitism can take the form of isolated cheating or exploitation among more generalized mutualistic interactions

The above was copied from Wikipedia's Parasitism article.

Consider this is relation to the Foundation's new interest in political advocacy for non-education-related causes, such as travel and immigration policies and climate change, which is in conflict with one of the Five pillars of Wikipedia, neutral point of view:

We avoid advocacy and we characterize information and issues rather than debate them. Emphasis mine. – wbm1058 (talk) 14:40, 4 March 2017 (UTC)

Describing the WMF as a parasite is as hyperbolic and offensive as calling it a cancer. It has the disadvantages of being offensive and hyperbolic - if you wanted to make valid points at the same time they get obscured by the discussion of how offensive and incorrect your message was. So just as the cancer analogy was rebutted by people pointing out that the growth was not following an exponential growth path, so describing the WMF as a parasite is Trump like in falseness, and can be responded to by people saying "more like Commensalism, how long would Wikipedia last without servers, bandwidth and electricity?" Signpost Op Eds can be effective, but in my view the effectiveness is unrelated, possibly even inversely related to the gratuitous offensiveness, and as strong as the most easily refuted falsehood. Lots of people remember when Signpost had a pottymouth issue, anyone remember what their ostensible cause was in that issue? Ϣere SpielChequers 19:05, 4 March 2017 (UTC)

You make fair points. But, consider "how long would LeBron James last without a basketball court and arena?" That would seem like a silly question. As a figure of speech, hyperbole is usually not meant to be taken literally. But when the WMF uses hyperbole in its appeals for donations (e.g. the well-worn "buy a programmer a cup of coffee" appeals) how many donors realize that's not meant to be taken literally? I suppose opinions vary as to how offensive such appeals are. Good behavior should be modeled from the top down. I remember the "pottymouth issue", though I'd like to forget it. Had I been part of the pre-publication review team, I would have strongly encouraged a major rewrite of the piece. Praying that never comes back to bite the author. wbm1058 (talk) 23:56, 4 March 2017 (UTC)

I'm a Londoner. When the WMF was telling people in the US that $5 would buy a cup of coffee they were saying £5 here, and in days when that would have got you $8 on the exchanges. I complained at the time, I still think it a classic example of a marketing campaign that trashed its own brand. Both because it fed the stereotype of programmers being overpaid "give us money and we will waste it", and because such a blatant disregard for the facts alienated the WMF from the rest of the community. The £5 coffee is far from the sole cause of tension between the WMF and the editors. But it did reinforce three memes that will take a long time to dispell:

The WMF wastes money
The WMF doesn't share the values that all true members of the community share
The WMF is capable of pissing the community off and then ignoring it.

Personally I don't believe that those three are all entirely true or fair, and that it was a mistake by the WMF to give themselves that image. But I don't envy the task of whoever at the WMF has to live that down, their best bet is to put up as many "under new management" signs as they can justify and try to approach the community differently for a few years. Self inflicted tarnish on a brand can be very stubborn. Ϣere SpielChequers 01:18, 5 March 2017 (UTC)

What kind of curve?

The S-curve (sigmoid curve) frequently matches the growth pattern of many processes, including children. A growth curve of this nature is to be expected and doesn't necessarily indicate a systematic problem. But it may be best for all concerned to understand that it will level off at some point, and plan accordingly. Possibly some portion can be set aside to serve as a non-profit investment fund for future needs? Praemonitus (talk) 22:05, 5 March 2017 (UTC)

I really don't think that anyone can be certain of the curve based upon only the last one or two data points. To my eyes, the curve at User:Guy Macon/Wikipedia has Cancer (which I think is much better than the simplified curve that the Signpost editors insisted be in this op-ed} looks like an exponential curve with one glitch (lower than curve revenue in 2013-2014 followed by lower than curve spending in 2014-2015, going back to exponential in 2015-2016). I can buy the idea that the curve might be exponential changing to linear (one or two data points are not enough to tell the two apart) but I see zero evidence that the curve is sigmoid. Of course we all know that exponential curves cannot continue forever, but it appears to me that in real life exponential spending growth is far more likely to end up with a crash rather than a leveling off. Again, none of us can predict the future. --Guy Macon (talk) 09:16, 6 March 2017 (UTC)
Yes there are plans for an endowment to be created. Running the fundraiser for shorter periods and stopping when the target is reached seems a positive thing to me. In theory the spending is driven by the budget not the the maximum income they can raise and they aren't raising as much as they possibly could. Is it possible we have, or could have had if the WMF hadn't been ending the fundraiser when they hit target, exponential fundraising and sigmoid curve on spending? Ϣere SpielChequers 09:28, 6 March 2017 (UTC)

Sharing some additional information

Greetings - the Wikimedia Foundation wanted to address some of the points raised and share some additional information.

"We should make spending transparent, publish a detailed account of what the money is being spent on and answer any reasonable questions asking for more details."
- We post our draft plans on-wiki for community review and feedback, following the format requested by the Funds Dissemination Committee, before it is adopted by the Board. We recently posted our mid-year check-in, and we are in the midst of preparing for the publication of our draft plan for fiscal year 2017-18 on-wiki. Our financial reports are posted on the Wikimedia Foundation website and announced to the community each year.
"We should freeze spending increases to no more than inflation plus a percent or two"
- Our current budget includes total expenses of up to $63,000,000 - a reduction from last year's total expenses. While final spending will vary from our budget, we have historically come in below or near our planned expenses. As reported in our mid-year check-in, we are currently $5,417,166 under budget in expenses for the period from July through December 2016, but we are expecting a majority of this will be spent during the second half of our fiscal year. Regarding an inflation-only cap on all future spending, we agree it is impossible to accurately predict what the future holds, and that is in part why we feel any such restrictions would be dangerously limiting.
"build up our endowment"
- Some of the previously mentioned plans include efforts to build up our endowment - which is now just over one year old. We are on-track to reach our goal of $100 million in principal funding by 2026.
"restructure the endowment so that the WMF cannot dip into the principal when times get bad"
- Fortunately, no restructuring is needed. Our endowment is housed under a separate organization (Tides) with a separate advisory board. This setup, among other advantages, protects the endowment and does not allow the Wikimedia Foundation Board to misuse funds.

It is possible that one day our donors, for whatever reason, will not be as supportive as they are today. We have a "rainy day" fund that provides us with approximately 20 months of funding to make any necessary adjustments to our spending, should the situation happen unexpectedly. Additionally, we launched our endowment last year to eventually help cover the most basic expenses required to sustain access to the contributions of volunteers.

Any organization working in technology is going to face challenges and experience failures. The Wikimedia Foundation continues to analyze and learn from ours. Most recently, we have been drafting Technical Collaboration Guidance and a code of conduct for Wikimedia technical spaces. We also recently appointed Victoria Coleman to Chief Technology Officer. Coleman helped write some of the documentation on technical engineering best practices - including designing the software engineering and theoretical strands of the Royal Holloway undergraduate computer science program and creating the Masters program on Dependable Computer System.

The Wikimedia Foundation does mission driven programmatic work beyond hosting the projects, and we support wiki projects beyond Wikipedia. We have grown to support Wikimedia Commons, Wikivoyage, Wikidata, and many other projects. We also work with over 100 regional and thematic affiliates around the world, help support numerous community events, hundreds of community activities, and awareness efforts to grow our volunteer and reader communities. Using the metrics of this article, some of our projects have grown by factors of millions. However, the reality is that these are poor metrics to use when considering a starting point of 1-2 staff and startup nonprofit funding levels of thousands rather than millions. We are currently a small operation in comparison to nearly every other website with an audience of our size who also utilizes user-generated content.

Measuring and increasing impact remains an ongoing and important topic. We are publishing our quarterly check-ins, and recently completed our mid-year check-in to help provide insight into the impact of our programs this year. How we measure our impact in the future is a topic that we hope the community will engage with us on during the upcoming movement strategy discussions.

We look forward to these conversations continuing, and appreciate the efforts of others to help provide a more complete set of facts to help inform each other. --GVarnum-WMF (talk) 00:03, 8 March 2017 (UTC)

Related question: [16] --Guy Macon (talk) 14:14, 8 March 2017 (UTC)

(Sound of Crickets...) --Guy Macon (talk) 21:50, 10 March 2017 (UTC)

(...Chirp...) --Guy Macon (talk) 19:21, 14 March 2017 (UTC)

@GVarnum-WMF: --Guy Macon (talk) 23:07, 14 March 2017 (UTC)

I finally got an answer of sorts, but it appears to not actually address any of the specific questions I asked. Please see Meta:Talk:Wikimedia Endowment#How should we select members of the Wikimedia Endowment Advisory board? --Guy Macon (talk) 10:02, 26 March 2017 (UTC)

┌────────────────────────────────────────────────────────────────────────────────────────────────────┘ It has been 30 days, and still nobody at the WMF has answered the three specific questions I asked.[17] --Guy Macon (talk) 04:12, 7 April 2017 (UTC)

Guy, thank you for your op-Ed. I share your concerns. You may find these questions of interest. Jus da fax 09:44, 8 April 2017 (UTC)

Keep up with The Signpost on Twitter.

Home

About

Year	Support and Revenue	Expenses	Net Assets at year end
2003/2004	$80,129	$23,463	$56,666
2004/2005	$379,088	$177,670	$268,084
2005/2006	$1,508,039	$791,907	$1,004,216
2006/2007	$2,734,909	$2,077,843	$1,658,282
2007/2008	$5,032,981	$3,540,724	$5,178,168
2008/2009	$8,658,006	$5,617,236	$8,231,767
2009/2010	$17,979,312	$10,266,793	$14,542,731
2010/2011	$24,785,092	$17,889,794	$24,192,144
2011/2012	$38,479,665	$29,260,652	$34,929,058
2012/2013	$48,635,408	$35,704,796	$45,189,124
2013/2014	$52,465,287	$45,900,745	$53,475,021
2014/2015	$75,797,223	$52,596,782	$77,820,298
2015/2016	$81,862,724	$65,947,465	$91,782,795
Based upon a table created by Simplicius on the German Wikipedia.

Wikipedia:Wikipedia Signpost/2017-02-27/Op-ed

Wikipedia has cancer

Author's Note

Expenses and pageviews

Software engineering

Hosting

Exponential then linear?

Transparency

Discovery

Bariatric surgery?

Reconsider the Title, Please

WMF Board

Signpost response to title request

A word, if I may

Frequently Unanswered Question

Reconsidering the title

What kind of curve?

Sharing some additional information

Navigation menu

Personal tools

Namespaces

Variants

Views

More

Search

Navigation

Interaction

Tools

Print/export

Languages