# Could Wikipedia get too big to fail? **URL:** \ **Category:** Great Debates\ **Created:** [December 31, 2009, 1:41am UTC](https://boards.straightdope.com/t/could-wikipedia-get-too-big-to-fail/523130 "2009-12-31T01:41:02Z")\ **Posts on this page:** 20\ **Page:** 3
**Author:** ![Martin\_Hyde](https://avatars.discourse-cdn.com/v4/letter/m/47e85d/32.png) [@Martin\_Hyde](https://boards.straightdope.com/u/Martin_Hyde)\ **Post date:** [January 1, 2010, 7:05am UTC](https://boards.straightdope.com/t/could-wikipedia-get-too-big-to-fail/523130/41 "2010-01-01T07:05:09Z")
> [@Kobal2](#): > > The real money sucker is the serving part, I’ll grant you that. Phat pipes don’t grow on trees, and renting the bandwidth adds up. However, if the information added by “unimportant” pages really is trivial, then it won’t be consulted much, will it ? No significant increase in costs there. And if it does drive up the costs, then by definition it’s not trivial - information that lots of people want to know is important information. Even if it’s Captain Picard’s bio. The more I think about it the more I doubt the government could fund Wikipedia directly as it stands now. I think it’d be hard to for the government to be in the position of paying for something that could ever be construed as “promoting” a specific television series or et cetera. I think there’s too many cans of worms. Also, you shouldn’t underestimate the costs of running a site like Wikipedia. The whole user system, the editing, the discussion pages et cetera have underlying database and programming that were done by software developers/database guys. Those guys make money, usually at least $60,000/year, and generally speaking if you want to keep something like that running long term you have to keep those guys around even once most of the real development work is done. I guess you could always phase them out in favor of lower paid “administrators”, but then you’d also be saying you wouldn’t be much interested in substantially improving Wikipedia over time. You also need people that will babysit the whole thing, in addition to your development team. You also need people who will be in charge of those people, allocating important tasks to specific people and et cetera. Even if you outsource a lot of this stuff, that still means you’re cutting fat checks whenever that stuff has to get done. ---
**Author:** ![Martin\_Hyde](https://avatars.discourse-cdn.com/v4/letter/m/47e85d/32.png) [@Martin\_Hyde](https://boards.straightdope.com/u/Martin_Hyde)\ **Post date:** [January 1, 2010, 7:07am UTC](https://boards.straightdope.com/t/could-wikipedia-get-too-big-to-fail/523130/42 "2010-01-01T07:07:05Z")
As a non profit you also need to have people that can bring the money in, WikiMedia foundation actually has a posting right now for a “Chief Development Officer” who would be in charge of such things. (Similar to the sort of career Michelle Obama was involved in.) Those people tend to be people who are great at networking and are well connected, meaning they don’t get paid in peanuts. ---
**Author:** ![Kobal2](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/kobal2/32/20_2.png) [@Kobal2](https://boards.straightdope.com/u/Kobal2)\ **Post date:** [January 1, 2010, 7:27am UTC](https://boards.straightdope.com/t/could-wikipedia-get-too-big-to-fail/523130/43 "2010-01-01T07:27:44Z")
> [@Martin\_Hyde](#): > > The more I think about it the more I doubt the government could fund Wikipedia directly as it stands now. I think it’d be hard to for the government to be in the position of paying for something that could ever be construed as “promoting” a specific television series or et cetera. I think there’s too many cans of worms. I dunno about that. After all, governments funds libraries, and those make “Hooking up with Tila Tequila” available. There even was a Pit thread about it not so long ago :p. In essence, Wikipedia is not much different from a library - not all the books are written by dead white men. That being said, yeah, never happens. For one thing, the open source fanatics who comprise the deep core of the foundation would eat their own gonads rather than accept being in the pocket of a government. I wouldn’t like it either, for that matter - would \*you \*trust the impartiality and objectivity of a “War in Afghanistan” page hosted, edited and monitored by the US government ? > [@](#): > > Also, you shouldn’t underestimate the costs of running a site like Wikipedia. > > The whole user system, the editing, the discussion pages et cetera have underlying database and programming that were done by software developers/database guys. Those guys make money, usually at least $60,000/year, and generally speaking if you want to keep something like that running long term you have to keep those guys around even once most of the real development work is done. I guess you could always phase them out in favor of lower paid “administrators”, but then you’d also be saying you wouldn’t be much interested in substantially improving Wikipedia over time. > > You also need people that will babysit the whole thing, in addition to your development team. You also need people who will be in charge of those people, allocating important tasks to specific people and et cetera. > > Even if you outsource a lot of this stuff, that still means you’re cutting fat checks whenever that stuff has to get done. All of this is true, but that overhead exists whether the “Captain Kirk’s date in episode #46” page exists or not, which was my point. I never said that the whole Wiki enterprise is cheap (the site does run on a 10 million bucks budget a year, after all), merely that sheer volume of available information isn’t a serious financial concern in and of itself. ---
**Author:** ![toofs](https://avatars.discourse-cdn.com/v4/letter/t/4491bb/32.png) [@toofs](https://boards.straightdope.com/u/toofs)\ **Post date:** [January 1, 2010, 7:34am UTC](https://boards.straightdope.com/t/could-wikipedia-get-too-big-to-fail/523130/44 "2010-01-01T07:34:48Z")
> [@Curtis LeMay](#): > > And this is an occasional Wikipedia vandal that says so. Wha…huh? ---
**Author:** ![Bryan\_Ekers](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/bryan_ekers/32/183_2.png) [@Bryan\_Ekers](https://boards.straightdope.com/u/Bryan_Ekers)\ **Post date:** [January 1, 2010, 7:42am UTC](https://boards.straightdope.com/t/could-wikipedia-get-too-big-to-fail/523130/45 "2010-01-01T07:42:43Z")
> [@Martin\_Hyde](#): > > Odd, since no one in this thread has done that Sure, as you say. ---
**Author:** ![mecaenas](https://avatars.discourse-cdn.com/v4/letter/m/91b2a8/32.png) [@mecaenas](https://boards.straightdope.com/u/mecaenas)\ **Post date:** [January 1, 2010, 9:49am UTC](https://boards.straightdope.com/t/could-wikipedia-get-too-big-to-fail/523130/46 "2010-01-01T09:49:18Z")
> [@Kobal2](#): > > That would most likely be the pro-active decision of one (or a few) of the aforementionned tight-assed editors. > > Yes and no. > The cost of hosting space is pretty negligible - the only necessary costs are one outdated box (even a 486 would do), a bunch of hard drives and some CAT cable to daisy-chain the new server to the rest of the bunch. How many wikipedia pages can you cram on a 20 bucks, 500 Gig drive ? > > The real money sucker is the serving part, I’ll grant you that. Phat pipes don’t grow on trees, and renting the bandwidth adds up. However, if the information added by “unimportant” pages really is trivial, then it won’t be consulted much, will it ? No significant increase in costs there. And if it does drive up the costs, then by definition it’s not trivial - information that lots of people want to know is important information. Even if it’s Captain Picard’s bio. Wow, you think you could host Wikipedia on a 486 because the problem is only internet bandwidth? Hahahahahahahaha! Sure. Whatever you say. ---
**Author:** ![Kobal2](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/kobal2/32/20_2.png) [@Kobal2](https://boards.straightdope.com/u/Kobal2)\ **Post date:** [January 1, 2010, 11:59am UTC](https://boards.straightdope.com/t/could-wikipedia-get-too-big-to-fail/523130/47 "2010-01-01T11:59:04Z")
> [@mecaenas](#): > > Wow, you think you could host Wikipedia on a 486 because the problem is only internet bandwidth? Hahahahahahahaha! Sure. Whatever you say. Exaggeration to make a point. It’s a distributed database. A handful of heavy hitting machines to handle the DBMS, and the rest are run of the mill boxes worth 2-4k a pop to house the files the DB points to, cache the most common requests etc… You don’t need a Cray2 to handle get requests. I just [checked](http://meta.wikimedia.org/wiki/Wikimedia_servers), and the grand total worth of the wiki’s hardware since its creation is 1.1 million dollars. OTOH, the foundation’s net income for [July to November 2007](http://upload.wikimedia.org/wikipedia/foundation/8/8c/WMF_Financial_Statements_July_1_to_November_30_2007-1.pdf) alone (period picked at random) was 700k. That’s right, the Wiki’s volume could increase by its current total size every year, and still it’d turn a profit. So, yeah, I don’t think hardware is a significant bottleneck, even on the current panhandling dependant financial system. And on a government budget ? Pffft, fuggedaboudit. That’s chump change. ---
**Author:** ![annelions](https://avatars.discourse-cdn.com/v4/letter/a/ed8c4c/32.png) [@annelions](https://boards.straightdope.com/u/annelions)\ **Post date:** [January 1, 2010, 12:27pm UTC](https://boards.straightdope.com/t/could-wikipedia-get-too-big-to-fail/523130/48 "2010-01-01T12:27:49Z")
> [@Curtis LeMay](#): > > No they ain’t at least for Star Wars. Wikipedia doesn’t have articles for major Star Wars characters like San Hill! We have Wookiepedia or Memory Alpha for that. Memory Alpha? Sorry to be picky, but why are you putting Star WARS info on a Star TREK wiki? 😛 Back on topic, though, I think it’s obvious that Wikipedia is mainly a “shortcut” when it comes to finding out most info. I usually start at Wikipedia and work outward, when I want to find something. Indeed, I came to Straight Dope tonight because I was curious about some specifics of scurvy; Wikipedia didn’t have what I wanted and I remembered a certain article about eating bread here. Never did really find what I wanted, but I did waste a few hours here! ---
**Author:** ![Ximenean](https://avatars.discourse-cdn.com/v4/letter/x/aca169/32.png) [@Ximenean](https://boards.straightdope.com/u/Ximenean)\ **Post date:** [January 1, 2010, 1:05pm UTC](https://boards.straightdope.com/t/could-wikipedia-get-too-big-to-fail/523130/49 "2010-01-01T13:05:47Z")
> [@Martini\_Enfield](#): > > It’s a big issue for the military history side of things, for example- a particular editor might know that Event X happened at Battle Y in place Z because their grandfather was actually there, but since “My Grandfather was there and he said…” isn’t “Verifiable”, it can’t be included. I’m afraid I would have to side with the Wikipedia editors on this one. People’s grandfathers tell them all sorts of stories. The information may be worth recording, but not in an encyclopedia. ---
**Author:** ![samclem](https://avatars.discourse-cdn.com/v4/letter/s/a9a28c/32.png) [@samclem](https://boards.straightdope.com/u/samclem)\ **Post date:** [January 1, 2010, 1:58pm UTC](https://boards.straightdope.com/t/could-wikipedia-get-too-big-to-fail/523130/50 "2010-01-01T13:58:44Z")
> [@annelions](#): > > Indeed, I came to Straight Dope tonight because I was curious about some specifics of scurvy; Wikipedia didn’t have what I wanted and I remembered a certain article about eating bread here. Never did really find what I wanted, but I did waste a few hours here! [Can man live by bread alone?](http://www.straightdope.com/columns/read/191/can-man-live-by-bread-alone) Cecil wrote that 32 years ago. ---
**Author:** ![mswas](https://avatars.discourse-cdn.com/v4/letter/m/e99b99/32.png) [@mswas](https://boards.straightdope.com/u/mswas)\ **Post date:** [January 1, 2010, 2:53pm UTC](https://boards.straightdope.com/t/could-wikipedia-get-too-big-to-fail/523130/51 "2010-01-01T14:53:20Z")
> [@Condescending\_Robot](#): > > It does so indirectly. If he were to chase off all the pseudoscientists, cranks, people who get off by publishing pictures of their fetishes, and people who think they are saving African children by documenting every appearance of sandwiches in Firefly, then he would decimate the Wikipedia user base. The most enthusiastic cheerleaders and donators would turn away from Wikipedia, and the free ad for Wikia would dry up. The Wiki business model is _fundamentally dependent_ on producing a poor-quality encyclopedia about nonsense, which is where it differs from respectable reference sources that need to pass muster with librarians and experts in various fields in order to continue making sales to educational institutions. Except that studies have shown that the quality of wikipedia is not significantly different from professional encyclopedias. So you are expressing a faith-based opinion. ---
**Author:** ![amanset](https://avatars.discourse-cdn.com/v4/letter/a/8dc957/32.png) [@amanset](https://boards.straightdope.com/u/amanset)\ **Post date:** [January 1, 2010, 2:55pm UTC](https://boards.straightdope.com/t/could-wikipedia-get-too-big-to-fail/523130/52 "2010-01-01T14:55:14Z")
> [@Kobal2](#): > > Exaggeration to make a point. It’s a distributed database. A handful of heavy hitting machines to handle the DBMS, and the rest are run of the mill boxes worth 2-4k a pop to house the files the DB points to, cache the most common requests etc… You don’t need a Cray2 to handle get requests. > > I just [checked](http://meta.wikimedia.org/wiki/Wikimedia_servers), and the grand total worth of the wiki’s hardware since its creation is 1.1 million dollars. OTOH, the foundation’s net income for [July to November 2007](http://upload.wikimedia.org/wikipedia/foundation/8/8c/WMF_Financial_Statements_July_1_to_November_30_2007-1.pdf) alone (period picked at random) was 700k. That’s right, the Wiki’s volume could increase by its current total size every year, and still it’d turn a profit. > > So, yeah, I don’t think hardware is a significant bottleneck, even on the current panhandling dependant financial system. And on a government budget ? Pffft, fuggedaboudit. That’s chump change. Hardware costs are only a minor part of why hosting costs. Where are you going to locate those servers? How are you going to cool them? What about the (massive) electricity costs? How do you pay the sysadmins? ---
**Author:** ![mswas](https://avatars.discourse-cdn.com/v4/letter/m/e99b99/32.png) [@mswas](https://boards.straightdope.com/u/mswas)\ **Post date:** [January 1, 2010, 2:55pm UTC](https://boards.straightdope.com/t/could-wikipedia-get-too-big-to-fail/523130/53 "2010-01-01T14:55:31Z")
> [@Martini\_Enfield](#): > > The biggest challenge that Wikipedia faces isn’t funding, IMHO, but internal politics. > > Anyone who’s spent a lot of time editing Wiki will have been told at least once that “Wiki isn’t about Truth, it’s about Verifiability”, which means a lot of information in niche or special-interest areas that’s well known in those fields but hasn’t been published in any “Wiki-approved” source. > > It’s a big issue for the military history side of things, for example- a particular editor might know that Event X happened at Battle Y in place Z because their grandfather was actually there, but since “My Grandfather was there and he said…” isn’t “Verifiable”, it can’t be included. > > The actual foundation behind Wiki- a free Encyclopaedia- isn’t going anywhere. But there are calls for the “Editable by anyone” aspect of it to be dropped, or at least severely curtailed. It’ll be interesting to see how Wiki evolves in the next 5-10 years or so… Which is interesting considering ‘official’ admins and such have been some of the biggest source of controversy, treating particular areas like fiefs. Whether it be Israelis trying to take over the Israel/Palestine conflict or the Pro-MMGW guy deleting any post and even banneing people that didn’t suit his particular narrative. Wikipedia’s quality would likely go DOWN and not up if it did whta you suggest. ---
**Author:** ![Kobal2](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/kobal2/32/20_2.png) [@Kobal2](https://boards.straightdope.com/u/Kobal2)\ **Post date:** [January 1, 2010, 3:50pm UTC](https://boards.straightdope.com/t/could-wikipedia-get-too-big-to-fail/523130/54 "2010-01-01T15:50:32Z")
> [@amanset](#): > > Hardware costs are only a minor part of why hosting costs. Where are you going to locate those servers? How are you going to cool them? What about the (massive) electricity costs? How do you pay the sysadmins? - location : they already have one. Well, two. With space to spare, [apparently](http://meta.wikimedia.org/wiki/File:Floridaserversfront1.jpg). - cool them : non-issue. More computers in the ventilated server room do not increase the cost of the ventilated server room. It’s already built, and the AC is already blowing. - massive electricity costs ? For 300 boxes ? Do we operate on the same definition of “massive” ? Again, we’re not talking about NASA supercomputers, here. Any IT company draws that kind of power. A high school probably draws more. They generally manage to pay the bills without having to sell nuclear secrets and can Medicaid. - more filehosting computers does not equate linearly more sysadmins. And, at the risk of repeating myself, I never said running the entire wikipedia was free gratis. I said, and I say again, that hosting more content is not a significant financial issue once the basic infrastructure is in place, which it is. That’s the whole _point_ of a distributed database. The overhead is paid for, so are the salaries and maintenance. The hardware costs are demonstrably negligible. My seminal point remains : deletion on grounds of notability or “importance” can kiss my [citation needed]. ---
**Author:** ![Aji\_de\_Gallina](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/aji_de_gallina/32/260_2.png) [@Aji\_de\_Gallina](https://boards.straightdope.com/u/Aji_de_Gallina)\ **Post date:** [January 1, 2010, 4:15pm UTC](https://boards.straightdope.com/t/could-wikipedia-get-too-big-to-fail/523130/55 "2010-01-01T16:15:36Z")
> [@mswas](#): > > Except that studies have shown that the quality of wikipedia is not significantly different from professional encyclopedias. > > So you are expressing a faith-based opinion. The study was severly flawed, particularly in the way it tallied mistakes. All mistakes were considere equal so that saying that Napoleon was born on a Tuesaday and it was a Thursday counted as one mistake and Wiki saying that Napoleon was the winner in my school’s 100m dash in 2008 counted as one mistake. ---
**Author:** ![annelions](https://avatars.discourse-cdn.com/v4/letter/a/ed8c4c/32.png) [@annelions](https://boards.straightdope.com/u/annelions)\ **Post date:** [January 1, 2010, 5:08pm UTC](https://boards.straightdope.com/t/could-wikipedia-get-too-big-to-fail/523130/56 "2010-01-01T17:08:15Z")
> [@samclem](#): > > [Can man live by bread alone?](http://www.straightdope.com/columns/read/191/can-man-live-by-bread-alone) Cecil wrote that 32 years ago. Yeah, found that one, but I haven’t been able to find anything about the minimum vitamin C is needed to stop/prevent scurvy. A friend was arguing that it was just the amount in an orange or two a year, while about the only conclusive thing I found is that it “can be reversed at any time”. Wikipedia was no help and even Cecil’s article wasn’t as helpful as I hoped. 😉 Ah well! ---
**Author:** ![mswas](https://avatars.discourse-cdn.com/v4/letter/m/e99b99/32.png) [@mswas](https://boards.straightdope.com/u/mswas)\ **Post date:** [January 1, 2010, 5:11pm UTC](https://boards.straightdope.com/t/could-wikipedia-get-too-big-to-fail/523130/57 "2010-01-01T17:11:04Z")
> [@Aji\_de\_Gallina](#): > > The study was severly flawed, particularly in the way it tallied mistakes. > All mistakes were considere equal so that saying that Napoleon was born on a Tuesaday and it was a Thursday counted as one mistake and Wiki saying that Napoleon was the winner in my school’s 100m dash in 2008 counted as one mistake. Well the point of Wikipedia is not to be the end-all be-all of knowledge but to be a good starting point for further research. Like if I look up Napoleon and find out he had an adviser named Talleyrand I can go read a book about Talleyrand and find out all kinds of interesting info about Napoleon. ---
**Author:** ![amanset](https://avatars.discourse-cdn.com/v4/letter/a/8dc957/32.png) [@amanset](https://boards.straightdope.com/u/amanset)\ **Post date:** [January 1, 2010, 6:56pm UTC](https://boards.straightdope.com/t/could-wikipedia-get-too-big-to-fail/523130/58 "2010-01-01T18:56:43Z")
> [@Kobal2](#): > > - location : they already have one. Well, two. With space to spare, [apparently](http://meta.wikimedia.org/wiki/File:Floridaserversfront1.jpg). > > - cool them : non-issue. More computers in the ventilated server room do not increase the cost of the ventilated server room. It’s already built, and the AC is already blowing. > > - massive electricity costs ? For 300 boxes ? Do we operate on the same definition of “massive” ? Again, we’re not talking about NASA supercomputers, here. Any IT company draws that kind of power. A high school probably draws more. They generally manage to pay the bills without having to sell nuclear secrets and can Medicaid. > > - more filehosting computers does not equate linearly more sysadmins. > > And, at the risk of repeating myself, I never said running the entire wikipedia was free gratis. I said, and I say again, that hosting more content is not a significant financial issue once the basic infrastructure is in place, which it is. That’s the whole _point_ of a distributed database. > > The overhead is paid for, so are the salaries and maintenance. The hardware costs are demonstrably negligible. My seminal point remains : deletion on grounds of notability or “importance” can kiss my [citation needed]. More computers = more heat. More heat = more AC. More AC = more costs. Both in terms of electricity and the requiring of upgrades when heat output exceeds AC capabilities. 300 boxes can draw a lot of energy. What sort of boxes are they? 1U little crappy things or 4U monsters like the Dell R900 with 4 six core processors running at full hilt? There’s a huge difference. Also what sort of storage arrays are they using? Those things use of electricity like crazy due to the moving parts. I never implied that you thought it was free. I just think that you are vastly underestimating the running costs of server farms. Oh and btw that image may look like it shows “space to spare” but when you are talking about 300 servers what you showed isn’t actually very much at all. ---
**Author:** ![Kobal2](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/kobal2/32/20_2.png) [@Kobal2](https://boards.straightdope.com/u/Kobal2)\ **Post date:** [January 1, 2010, 7:23pm UTC](https://boards.straightdope.com/t/could-wikipedia-get-too-big-to-fail/523130/59 "2010-01-01T19:23:03Z")
> [@amanset](#): > > 300 boxes can draw a lot of energy. What sort of boxes are they? 1U little crappy things or 4U monsters like the Dell R900 with 4 six core processors running at full hilt? There’s a huge difference. Also what sort of storage arrays are they using? Those things use of electricity like crazy due to the moving parts. The specs are detailed [here](http://meta.wikimedia.org/wiki/Wikimedia_servers/hardware_orders), by year of acquisition. Knock yourself out 🙂 ---
**Author:** ![Tom\_Tildrum](https://avatars.discourse-cdn.com/v4/letter/t/e95f7d/32.png) [@Tom\_Tildrum](https://boards.straightdope.com/u/Tom_Tildrum)\ **Post date:** [January 1, 2010, 7:49pm UTC](https://boards.straightdope.com/t/could-wikipedia-get-too-big-to-fail/523130/60 "2010-01-01T19:49:16Z")
If Wikipedia got into serious money trouble, couldn’t it just sell ad space? [Previous page](https://boards.straightdope.com/t/could-wikipedia-get-too-big-to-fail/523130.md?page=2) [Next page](https://boards.straightdope.com/t/could-wikipedia-get-too-big-to-fail/523130.md?page=4)