Slashdot Mirror

The Problem with Wikipedia: http://upload.wikimedia.org/wikipedia/en/8/89/Penn y_Arcade_comic-20051216h.jpg

Actually, the foundation that runs Wikipedia, Wikibooks, Wiktionary, and the like does currently have six staff members. By "free", Wikipedia means that its content can be freely distributed, and nothing else really, although Wikipedia does lean towards free, open source software (it runs on MediaWiki). Having several paid encyclopedia-writers could be troubling, though, because it promotes inequality and even envy in the community, possibly driving away or discouraging those that put their free time into improving Wikipedia, or promoting elitism. Check out this page for a perspective on that; and think of the further problems that would result if this forced elitism applied to article writers!

Often, it's easier to write a featured article about something like this than something like this. As far as I am aware, there have been no repeat Featured Articles on the Main Page, so that means that featured articles keep on coming... but some are also being defeatured due to quality concerns. There was a net gain of four featured articles this week—gained nine, lost five. Often vandalism gets in the way of constructive article writing, and people have to spend more time on that, rather than on content-producing.

Finally, one of the goals of featured articles is to get an article to a place where it is incorruptible... but not unimprovable. (Motivation is another goal.) So if someone helped bring an article to featured status, they might notice any factual errors that were introduced. Wikipedia certainly has dynamic... but it's losing some of that. With a team of vandal-fighters and no content-writers, Wikipedia will only be able to preserve integrity -- not improve it.

Actually, the foundation that runs Wikipedia, Wikibooks, Wiktionary, and the like does currently have six staff members. By "free", Wikipedia means that its content can be freely distributed, and nothing else really, although Wikipedia does lean towards free, open source software (it runs on MediaWiki). Having several paid encyclopedia-writers could be troubling, though, because it promotes inequality and even envy in the community, possibly driving away or discouraging those that put their free time into improving Wikipedia, or promoting elitism. Check out this page for a perspective on that; and think of the further problems that would result if this forced elitism applied to article writers!

Often, it's easier to write a featured article about something like this than something like this. As far as I am aware, there have been no repeat Featured Articles on the Main Page, so that means that featured articles keep on coming... but some are also being defeatured due to quality concerns. There was a net gain of four featured articles this week—gained nine, lost five. Often vandalism gets in the way of constructive article writing, and people have to spend more time on that, rather than on content-producing.

Finally, one of the goals of featured articles is to get an article to a place where it is incorruptible... but not unimprovable. (Motivation is another goal.) So if someone helped bring an article to featured status, they might notice any factual errors that were introduced. Wikipedia certainly has dynamic... but it's losing some of that. With a team of vandal-fighters and no content-writers, Wikipedia will only be able to preserve integrity -- not improve it.

Atomistic simulations of biomolecules. Chain a bunch of those together, and you begin to simulate systems on realistic time scales. Higher-resolution weather models, or faster and better processing of seismic data for exploration. Same reason that we perked up when the R8000 came out with its (for the time) aggressive FPU. 125 MFlops/proc@75MHz was nothing to sneeze at 15 years ago. If they can get this chip into production in usable quantities, and if it has the throughput, then they're on to something this time.

Of course, this could just be a single-chip CM2; blazingly fast but almost impossible to program.

http://upload.wikimedia.org/wikipedia/commons/4/42 /Carlb-sockpuppet-02.jpg

I got that you didn't like MediaWiki. Yeah, it is slow, but it is slow because of what it does.

As far as the actual studies involved, I would have to at the moment refer you to the Wikimedia development team directly, although I've seen some published values that do go into some details.

For general statistics of Wikimedia projects, I would have to refer you to http://stats.wikimedia.org/ that goes into some depth about individual projects and what the general demands on them are, including statistical summaries of leading contributors, growth of content, and edit counts that would certainly be of general interest in terms of trying to compare to other Wiki environments.

I also would like to mention that Erik Zachte, the person who has written this statistical summary mentioned above, has also gone into depth regarding general usage data where he has been given direct access to the Apache server logs and has noted areas that were critical for Wikimedia projects. Brion Vibber is also actively involved with these reviews, and several of these statistical summaries were noted among the internal developers lists, with hints of these studies being mentioned from time to time on other foundation mailing lists.

There have also been formal requests for performing this sort of statistical analysis by several university research teams that have been eager to get such a statistical set, which also prompted the WMF to establish specific guidelines for obtaining this sort of raw data.

Is this specific enough? I don't know right off hand besides these direct studies, but I do know there are others that do exist as well. Wikipedia is a heavily studied topic in part because much of the data is open and available, which gives some interesting sociological interpretations as well if studied through the lens of a statistical review. And there is enough raw data to come to conclusions that may not fit the traditional orthodoxy, so you can also tweak some noses at the same time.

The reason I mention MediaWiki's feature set is that you are (I'm presuming here) claiming that one of the reasons why the Wikimedia Foundation is running out of money is due in part because they are foolishly spending money on server resources that could be better run had they only selected the proper Wiki software. I am offering a rebuttal that this is hardly the case, and that almost (because I can't claim absolute knowledge here) any other Wiki editing software package would die a horrible and nearly instant death if they had to deal with the same feature set and bandwidth issues that currently confront Wikipedia. Or that the other software packages are so lacking in the essential requirements needed to run Wikipedia that there is hardly room to even justify a valid comparison based off of only one single comparison.... content distribution bandwidth on the CPU.

Though Wikimedia claims to be an open group, all we get to see of their financial picture is the audited statement and whatever else they decide to dole out. We don't get to see how many thousands of dollars "unpaid volunteers" get to spend on travel (nearly 10% of Wikimedia Foundation's expenses in 2006) or even on Wikimania (is that part of the 8 percent of "other" expenses?).

Financial statements What I really liked about the supposedly GATT compliant financial statement was the way auditors picked and chose between "in-kind revenues". They don't list what were the in-kind revenues, but we can guess Yahoo's Asia server farm is a big part of it. What we do know is the value auditors placed on the in-kind donations of content that built Wikipedia. Zero. Notta. Nothing. So intangible as to be worthless.

According to the financial audit, in Summer 2006 they had $500,000 cash on hand and incoming donations of $30,000 per month. In January 2007 the last fundraiser finished with $1,000,000. In their own projection, they say they need $75,000 per month to keep the site running and to pay salaries. Assuming that that same cost was incurred already throughout the latter half of 2006, and that every month in 2006 they received $30,000 in donations but now after the fundraiser additional donations have completely dried up, Wikimedia has now enough cash on hand to operate through April 2009. Of course this assumes that they don't start hiring new people left and right (which unfortunately is in their plans).

Devouard's desperate outburst follows by two months Wikimedia Foundation's most successful fundraising month ever, if Wikimedia Foundation's Web pages can be trusted to report even the foundation's internal affairs accurately. Dec. 2006 produced $899,207 in donations -- more than three times the previous record of January, 2006.

Currently reported fiscal 2007 donations from July 06 through Feb. 07 are $1,317,438. That is 7 of 12 months, and if revenues continued at that pace for 12 months, total revenues for fiscal 2007 would be $2,258,466. As I estimated in the post at the root of this thread, based on previous year's growth rates, expenses this year would be $2,077,609. Estimated year-end revenues are still 108 percent of expenses.

Ms. Devouard's comments might be more desperate than the situation warrants, but her anxiety does inspire us to look at Wikimedia Foundation reports. An expanded review of recent Foudation reports reveals that growth in contributors' financial commitments is slowing down. Ms. Devouard might in fact realize that public fascination with Wikipedia has peaked, and the numbers show it. Donations in the first seven months of fiscal 2007 matched the total for fiscal 2006, but were on track to grow at half the pace of the previous two years.

If Devouard and the board (which Wales chaired at the time current expense levels were set in motion) naively anticipated that donations would continue to quadruple year after year, and budgeted expenses based on those unrealistic projections, we need to consider whether a slowdown in revenue increases or an irresponsible board of directors is to blame.

Revenue growth:
FY 2007/2006: 1.730238138 (projected)
FY 2006/2005: 4.461453801
FY 2005/2004: 4.15047311

http://fundraising.wikimedia.org/en/fundcore/brows e/2006
http://fundraising.wikimedia.org/legacy/ongoing/

Devouard's desperate outburst follows by two months Wikimedia Foundation's most successful fundraising month ever, if Wikimedia Foundation's Web pages can be trusted to report even the foundation's internal affairs accurately. Dec. 2006 produced $899,207 in donations -- more than three times the previous record of January, 2006.

Currently reported fiscal 2007 donations from July 06 through Feb. 07 are $1,317,438. That is 7 of 12 months, and if revenues continued at that pace for 12 months, total revenues for fiscal 2007 would be $2,258,466. As I estimated in the post at the root of this thread, based on previous year's growth rates, expenses this year would be $2,077,609. Estimated year-end revenues are still 108 percent of expenses.

Ms. Devouard's comments might be more desperate than the situation warrants, but her anxiety does inspire us to look at Wikimedia Foundation reports. An expanded review of recent Foudation reports reveals that growth in contributors' financial commitments is slowing down. Ms. Devouard might in fact realize that public fascination with Wikipedia has peaked, and the numbers show it. Donations in the first seven months of fiscal 2007 matched the total for fiscal 2006, but were on track to grow at half the pace of the previous two years.

If Devouard and the board (which Wales chaired at the time current expense levels were set in motion) naively anticipated that donations would continue to quadruple year after year, and budgeted expenses based on those unrealistic projections, we need to consider whether a slowdown in revenue increases or an irresponsible board of directors is to blame.

Revenue growth:
FY 2007/2006: 1.730238138 (projected)
FY 2006/2005: 4.461453801
FY 2005/2004: 4.15047311

http://fundraising.wikimedia.org/en/fundcore/brows e/2006
http://fundraising.wikimedia.org/legacy/ongoing/

Already being done.

http://meta.wikimedia.org/wiki/Wikimedia_servers#O verall_system_architecture

Total Expenses:
2006: $791,907
2005: $177,670
2004: $23,463

Total Revenues:
2006: $1,305,292
2005: $292,571
2004: $70,491

Revenue/Expenses:
2006: 1.648289509
2005: 1.646710193
2004: 3.00434727

And now Ms. Devouard claims the foundation needs 10 times as much revenue to meet expenses in 2007 as in 2006? It just doesn't fit the curve, even according to the audited financial statements the foundation eventually released (two months ago) about the foundation's first three years.

Do the math:
Devourd's statement implies Wikimedia Foundation needs $4 million for the first two quarters of 2007, or $8 million for the year. That would be about a 10-fold increase over 2006 expenses.

In fact, 2005 expenses were 7.57 times 2004 expenses. 2006 expenses were 4.45 times those of 2005. Expense growth (445%) from 2005 to 2006 was only about .59 (0.58861255) as much as growth in the prior year (757%). On this somewhat limited graph, if the downward trend in expense growth continued in a straight line, growth in 2007 expenses would be 59% of growth in 2006 expenses. This curve predicts 2007 expenses of $2,077,609 - not the $8 million Devourd implied the Foundation requires to preserve the millions of edits users have contributed since Wales and Sanger started the project in 2000.

Meanwhile, as increases in expenses have tapered, increases in revenues have been steady, at more than 400% annually. Total revenues were 300% of total expenses in 2004, and 165% of total expenses in both 2005 and 2006. How then does the president of Wikimedia Foundation support the claim that 2007 revenues ($1 million in January) are projected to be only 25 percent of 2007 expenses, which she implies will grow 10-fold this year, after growing less than three-fold last year. In my view, Devouard is either lying as part of a fear-mongering fundraising strategy, or suffering the same clouded reasoning that seems to infect many Wikipedia writers' appreciation of facts.

Wikimedia Foundation financial statements:
http://upload.wikimedia.org/wikipedia/foundation/2 /28/Wikimedia_2006_fs.pdf

They paid more than $100,000 a year in salaries to their growing staff? And built up a $500,000 cash reserve? http://en.wikipedia.org/wiki/Wikipedia:Wikipedia_S ignpost/2006-12-11/Financial_audit http://upload.wikimedia.org/wikipedia/foundation/2 /28/Wikimedia_2006_fs.pdf

Have a look at Wikimedia servers page. It's more advanced and complicated than what you think.

I wonder if the time has come for a more distributed hosting model. In the current model, a site is hosted on a single machine.

What I have in mind is something like a simple daemon program that anyone can run, which caches part of the content of a site and serves it to site visitors. I am sure many people would be willing to run such a daemon for sites they like. I think the system could even be made transparent to browsers, so that anybody could be using it as soon as the system is implemented.

I wonder if the time has come for a more distributed hosting model. In the current model, a site is hosted on a single machine.

What I have in mind is something like a simple daemon program that anyone can run, which caches part of the content of a site and serves it to site visitors. I am sure many people would be willing to run such a daemon for sites they like. I think the system could even be made transparent to browsers, so that anybody could be using it as soon as the system is implemented.

Downloads of all the Wikimedia Projects. You need to do a lot of DB work (XML -> SQL conversion, importing, rebuilding tables, etc.)

The issue is simply that massive servers are not cheap. Wikimedia is already at 100+ servers, and they are barely getting by. They could spend half a million on servers and still have a wish-list. And bandwidth isn't cheap. They get a charity discount, and a bulk discount, but it's still gigabytes and gigabytes a day.

We all know what we need to do. Join with me!

EARTH!!
FIRE!!
WIND!!
WATER!!
HEART!!

GO PLANET!!

Obligatory Captain Planet Themesong Link

What're these "must-have" features in KDE? Any time I've used it, I've found a bunch of stupidly-named applications, and a big, bulky UI filled with toolbars. I'd rather use GNOME. Hell, I'd rather use Windows.

I like how you linked to an article on KDE 4, has anything of KDE 4 even been shown yet? All I recall seeing is some hand-drawn diagrams and some mock-ups. The Release Plan indicated they were aiming for a technical preview release in October, did that even happen?

I did find a screenshot, hardly looks like the revolutionary improvements shown in the mock-ups (linked site not in English, but if you scroll about half way down the page you can see the mock-ups).

http://upload.wikimedia.org/wikipedia/en/f/f0/Spli t_Screen_in_Pac_Man.gif

Every time I look into the peephole to get scanned, I'm relatively certain a large needle will shoot out from behind the glass and stab me in the eye

You want scared? Over 90% of everybody gets cataracts before they die. I'm 54, they did this to me last summer.

I was looking forward to having them poke my other eye with a needle, until I had a torn retina. The biggest risk factor for a torn retina is severe nearsightedness, followed by having had cataract surgery.

I thought it hurt when they welded the retina back together with a laser, but that was nothing! My IOL is a CrystaLens. as teh linked wiki article puts it, "The position of the lens can be changed by the ciliary muscles of the eye, allowing for natural focusing." After wearing glasses all my life, then contact lenses AND reading glasses, I don't have to wear glasses or contacts at all now! The downside is if your retina tears, it can get in the way of the eyeball welding laser and they have to freeze your eyeball with a probe cooled with liquid nitrogen.

If I'd been strapped to a chair at gitmo when they did that, I'd have confessed to anything!

I hope none of you are squeamish, sorry... but if you need cataract surgery, spend the extra money and get the good one.

I think Wikipedia is on several mysql servers simultaneously (not to mention well over a hundred http and squid servers). But I don't know the technical details.

details are here

Alternatively, the Sierpinski triangle. Zelda nuts recognize the first iteration as the Triforce.

I think I have a few suggestions for that design. Here is an example of a one-person vehicle running (mostly) on sugar. Here's another.

The main issue with both of these is efficiency. It could be measured in terms of "miles per bushell of oatmeal", I guess...

I think I have a few suggestions for that design. Here is an example of a one-person vehicle running (mostly) on sugar. Here's another.

The main issue with both of these is efficiency. It could be measured in terms of "miles per bushell of oatmeal", I guess...

She looks like a stunt double.

I think you're missing the fact that people tend to attribute characteristics -- attractive people are good and conversely good people are attractive -- based on unrelated factors. Seeing a chick with a computer can be as compelling for a nerd as seeing an old man in a Ferarri might be for a model, where seeing the same people in different circumstances would likely result in indifference.

Beauty is in the eye of the beholder, but people don't always realize everything they're beholding.

No way.
I also haven't seen any fluorescent "safelight" bulbs. No reason to ban these as they usually just use 15watts.

Yes, Price-Anderson is the Act that makes commercial use of nuclear power possible. For Chornobyl the periodic control zone is about 5000 sq mi (http://upload.wikimedia.org/wikipedia/commons/0/0 7/Chornobyl_radiation_map.jpg) or 3.2 million acres. So, the limit of $10 billion liability comes to about $3000 per acre, not even what unimproved land sells for. With a $250,000 house per acre, we see the magnitude of the liability subsidy. You may feel that is fair, but I feel it is a market distortion. And, as you say, nuclear power is not possible without it. Nuclear power has not demonstrated itself to be clean even disregarding the waste problem as numerous accidents have shown.

What to do with the nuclear waste fund? Until we know that, we'd do best to stop making more waste.
----
Have a strange love for nukes? That's OK. You can still save money using solar. http://mdsolar.blogspot.com/2007/01/slashdot-users -selling-solar.html

Can you even imagine what a sight that would be? Suborbit to see infinite stars already sounds amazing, but dancing lights would make the millions worthwhile. Check out the view from the ISS http://upload.wikimedia.org/wikipedia/en/6/63/Auro ra_Borealis.jpg/ (640k link) That said, count me surprised if flying an extraordinarily complex spaceship through ionic storms, but maybe I'm overestimating what it exactly does, despite glancing through the wiki article on northern lights.

The situation on wikimedia commons is split into two. First we have a separate signup project called "flickrlickr" which is a separate program run by Eric Moller. He has run the bot on wikimedia acceptable licences (only one of the two at the moment) and then reviewers, such as myself, wade through the 1000 and 1 pictures of cats and dogs to find really HQ pictures, highlights:

http://commons.wikimedia.org/wiki/User:FlickrLickr /Highlights

which we then flag and they are uploaded to the commons. Eric knows his program has scanned all the correctly licensed photos - and it has came down to the situation that if a Flickr author asks wikimedia to remove them we may - but we don't have to - we have the program logs and we can prove that they were uploaded at one point under a cc licence, no matter what the Flickr page now says.

For pictures uploaded to wikimedia commons by individual users the situation is a little more blurred. There is now a situation in place where a bot checks most of the pictures and cetifies they are under the correct licence - but that is pretty recent and so there are quite a few older ones which have had licence changes and they just have to be removed as we have no way of proving they once were CC.

In short, at wikimedia commons there has been a major drive to cover our backs to be able to prove that something was uploaded under CC. If an author later changes licences and asks the commons to remove then in most cases we will oblige, but we are now in a position to refuse if we need to (such as a very useful picture etc). I think some folks on the commons got in touch with the folks at Flickr and have tried to persuade them to show a history of licences and there would be no further problems - but so far they haven't obliged.

Oh, the Freespace series was great, the best of the now extinct space shooter genre. They didn't have very much new material, but synthetized the best from all the other ones. You finally get to see properly done enormous beam weapons, too, and one of the only ancient mysterious alien races that actually STAY mysterious. The story's light on characters, but let me put this way... the first part is A New Hope, which features an intriguing self-contained tale. The second one is The Empire Strikes Back, which expanded the universe, started a number of new plot points and left no answers, only further questions.

Then Return of the Jedi was cancelled for the next Descent.

... is here; they seem to be concerned about a "search engine optimization world championship".

Personally I think we can all do our bit and stop linking to Wikipedia so much, because Google is starting to give the impression that Wikipedia is the fount of all knowledge - to the detriment of pages which contain better information but which don't happen to have WP's massive net presence.

The Indian space program is now working on a recovery vehicle with four 'arms' to retrieve satellites. Here is a an artist's rendition.

how many lego combinations are possible

To simplify the question, we could consider just these classic bricks. By different combinations we'll understand fully connected arrangements, with no regard to combinations of colour, rotations, or symmetries. I suppose that Legos can connect with a single corner, correct me if I am wrong.

Le(1) = 1

Le(2) = 17

Then, for one of the combinations in Le(2), there are 18 ways to add the third piece. The problem seems to be barely tractable now without the aid of at least lego pieces and a piece of paper, but I'll make bold assumptions. If Le(n) grows at least as fast as 10^n (and my gut tells me that it grows much faster), then measly 100 pieces will give you a quantity that dwarfs the number of particles in the known universe.

Meta has been very helpful. Of course, I hadn't referenced it in how to set up a proper server...

I have some wireless power for you, right here!

But seriously, I don't see something like this working through somebody's skull and brain very safely or well in the near future.

Microsoft's (international) revenues are less than a third of a percent of US GDP. Check, that's true.

Adding Microsoft's revenues to those of two other companies can total *almost* one percent of GDP. Check, I can believe that too. Not sure why you picked 3M and P&G -- too lazy to search for a relationship. Therefore I'm simply going to assume that you picked two other decent sized but not huge companies (Microsoft is only 48th on the Fortune 500; Exxon Mobil is first with profits of about $36 billion -- i.e. almost the size of Microsoft's revenues; at #13, Berkshire Hathaway is more than twice as big as Microsoft). Combined, your three companies are smaller than Citigroup (8th on the Fortune 500). Not sure what your point was. Why combine those three companies? Is there some reason that breaking up the Microsoft monopoly would automatically affect the other two?

See http://money.cnn.com/magazines/fortune/fortune500/ full_list/index.html for Fortune 500 data.

If Microsoft's revenues went to zero, it would significantly harm the US economy. Basis? I don't believe that. Consider that defense use to make up about 6% (6.2% in 1986) of US GDP. It dropped from 4.8% to 3.7% between 1992 and 1995. In general, those were considered to be good economic years. From 1995 to 1999, it dropped a further .7% to 3%. Yet somehow, despite this, those were considered to be great economic years. The 1992-5 era is especially interesting, as spending dropped from 297 billion to 259 billion. That's about 38 billion dollars. I.e. roughly the same magnitude as Microsoft (albeit in more valuable 90s dollars rather than the relatively depreciated 2005 dollars). In other words, the defense shrinkage from 1992 to 1995 was actually larger in magnitude than Microsoft's revenues. Yet somehow the economy not only survived but prospered.

Of the twenty-nine agencies and departments listed in the 2005 federal budget, thirteen are larger than Microsoft's revenues.

Defense % of GDP from http://www.truthandpolitics.org/military-relative- size.php
Historical budget numbers from http://www.whitehouse.gov/omb/budget/fy2007/ (in particular, http://www.whitehouse.gov/omb/budget/fy2007/pdf/hi st.pdf ).

It's also worth noting that no one is talking about sending Microsoft's revenue to zero. In fact, because of the way monopolies work, the normal result would be to *increase* revenues while decreasing profits. A monopoly only produces up to the point where marginal revenue (from sales) exceeds marginal cost (of production). See http://en.wikipedia.org/wiki/Monopoly -- in particular, http://upload.wikimedia.org/wikipedia/en/thumb/e/e f/Monopoly-surpluses.svg/250px-Monopoly-surpluses. svg.png

The green line represents marginal revenue. The red line represents consumer demand. The blue line represents marginal cost of supply. Note that marginal revenue is positive for at least part of the distance between the monopoly quantity produced and the competitive price. Also note the yellow region. This is the area where the economy *gains* as a result of switching from a monopoly to a competitive market. It comprises the benefits of increased production minus the costs. The blue rectangle (i.e. the part above the Pc line) is gain shifted from producer (Microsoft) to consumers.

To reiterate:

1. Microsoft is not really that big a part of the US economy in terms of revenue.
2. Even if it were, no on is seriously argui

Take a look at the following:

http://upload.wikimedia.org/wikipedia/en/7/7a/1112 FIC326x550.jpg
http://images.apple.com/home/2007/images/iphoneloc kscreen20070109.jpg

Did Apple really create something that has such a unique look??

What excellent news! I'm sure the world's IP supercreators won't mind. But what's this I see on their webcam?
Sealand Webcam

Apparently the same evildoers responsible for "Western Australia" painted some swans black and set them loose. That's a darn good coat of paint.

Maybe the AI is working from a local copy of the Wikipedia database that isn't vulnerable to live vandalism or anything silly like that. And maybe Wikipedia spammers are more interested in a) putting links to their sites at the bottom of articles to boost PageRank and to capture the attention of random viewers or b) putting in biased promotional material and and other advertisements in a relevant page. And maybe this is likely to be far more attractive of an option than spamming Wikipedia in irrelevant places in the vague hope to poison a Bayesian filter which may or may not exist and probably is unlikely to ever see a revision of the article with this irrelevant information before someone reverts it. (Remember, it's obvious, systematic vandalism that attracts the most attention).

(for december)
1. Main Page (duh!)
2. Wikipedia
3. Wiki
4. Wii
5. Sex
6. World War II
7. United States
8. Christmas
9. Edvard Munch
10. Deaths in 2006
11. Naruto
12. The Holocaust
13. Pornography
14. Sexual intercourse
15. List of sex positions
16. List of big-bust models and performers
17. Adolf Hitler
18. YouTube
19. Attack on Pearl Harbor
20. American Civil War

From http://hemlock.knams.wikimedia.org/~leon/stats/wik icharts/index.php?lang=en&wiki=enwiki&ns=articles& limit=20&month=12%2F2006&mode=view

Wii
Sex
World War II
United States
Christmas
Deaths in 2006
Naruto
Sexual intercourse
Pornography
The Holocaust
List of big-bust models and performers
List of sex positions

Sad.

I don't think the article ever implied this was malicious or trying to censor Qatar. However, this is not simply an innocent mistake either. This administrator likely made a mistake, but there has been an asserted effort to spread disinformation about internet technologies among administrators so that such a mistake is possible. For example you can see Dmcdevit spreading disinformation about open proxy servers in this thread. Most administrators talk about everything in terms of "open proxy servers" and shared IPs without understanding that a single shared IP can be the public IP for an entire private IP network of millions of IPs. There's no sense among administrators that the scale of a shared IP can be different and that the difference matters when deciding to block. Even before the Qatar mishap Wikipedia editors were trying to raise the issue on the Wikimedia site in relation to T-Mobile's large RFC 1918 private network. Instead of simply accepting the advice a bizarre edit war ensued (from the last week of 2006 through the first week of 2007). These things suggest that something strange is going on at Wikipedia/Wikimedia that cannot be explained by simply saying: "woops, we made a mistake."

In reference to the bandwidth requirements for displaying rather than editing a web page via wiki as opposed to a simple ordinary webserver, I think Wikipedia does a pretty good job for those who just casually browse (and hit hyperlinks) to the content.

There are two huge issues here, somewhat related:

1) Bandwidth demands even for those who are doing read-only access to Wikipedia are absolutely huge. The Wikimedia Foundation is buying network bandwidth at industrial rates, being essentially a secondary peer for the most part with much of what they offer. It is almost impossible to comprehend the sheer bandwidth it takes to serve everything, especially considering Wikipedia is now in the top 10 servers in the world by several standards.

2) The number of pages being served by Wikipedia alone are absolutely huge. The raw HTML for just English Wikipedia is close to 7 GB of data. That does include talk pages and user pages, so that number can be modified somewhat for raw content.

I'm not convinced that a mere mailing list would work with the volume of changes that happen to Wikipedia every day, even if you confine it to completed work-overs of articles by legitimate scholars interested in updating the content. And even then you would have to come up with a way to redistribute the changes to everybody interested in the "current revision" (presumably just the people who have made changes recently or want to make some changes).

There have been some alternative systems proposed, including some sort of P2P content distribution system that would allow you to access Wikipedia content and send updated articles to a central server. While technically possible, the software do get something like that going is hardly trivial. Although admittedly it may be possible to spend some of the current $1 million+ budget on perhaps some software developers to work on some of those alternative distribution systems.

There is an effort going right now among the major volunteer software developers with Wikipedia to develop a more static version that would only get updated once those contributing feel like the article is ready to be "published". The reasons for doing this are more for quality purposes than monitary cost, as the network bandwidth savings doesn't appear to be all that significant. I could be mistaken on this point, however.

Most of the CPU bandwidth is mainly conducting information traffic flow rather than performing actions on the behalf of editor/contributors. There is some "page assembly" stuff that happens (converting the Wiki-markup text to HTML, for example), it doesn't appear as though currently the CPU bandwidth is a killer issue.

The reason why I suggested about $5 million per year is in part due to potential growth of other Wikimedia sister projects, and the growth of non-english Wikipedias. Chinese Wikipedia, for example, could potentially grow substantially over its current usage patterns, as can several other major languages including Indonesian and languages of the Indian Sub-continent. I've watched both bandwidth usage and page growth, and both seem to show exponential curves even now, with no "flatening" in sight. It would be interesting to perform some regression analysis on the data, which you can find at http://stats.wikimedia.org/ (regularly updated too!)

In reply to ta bu shi da yu:
"Out of interest, have you checked your facts?"

Yes I have definitely checked my facts.

"There are certainly a number of paid employees (I think it's about 4 or 5)."

You're confusing this with the WikiMedia foundation that has only a few employees. However, another agency employs dozens of paid administrators to work shifts around the clock on the Wikipedia and the Wikimedia site.

"There are no paid administrators, and there are over 500 of them on the site last I checked."

No, there are over 1,000 administrators, though many are inactive. Many may indeed be volunteers, however most active ones are paid. If you're a volunteer administrator you may have been unaware that others are paid, but they are.

"I should point out that not all administrators know (or indeed should know) about subnetting or CIDR, and mistakes can be made."

It's worst than that. Many administrators learn a little bit about networking technologies that just gets passed amongst themselves. If someone in these threads is talking about open proxies in referring to what's clearly Qatar's RFC 1918 private network, you can bet they are an administrator from Wikipedia simply regurgitating what they've heard from fellow administrators. Some of these administrators actively try to keep information about network technology from other administrators. Again check out the bizarre edit war (around the last week of 2006 first week of 2007) at the foundation site that could have prevented this Qatar mistake. Also see the disinformation put out by Dmcdevit on the Administrators noticeboard. I have no idea what's behind that behavior, but that's what's going on.

Oh, I'm sure that's not the *only* way...
http://bugzilla.wikimedia.org/show_bug.cgi?id=1218
http://bugzilla.wikimedia.org/show_bug.cgi?id=4947
http://bugzilla.wikimedia.org/show_bug.cgi?id=8419
Couldn't find a bug on it, but using mod_deflate would probably help too, and the deflate work could be pushed on to intermediate servers.