# Wikipedia:Village pump (proposals)/Archive 16

## A live (real time) messaging system

Out of curiousity, why hasn't wikimedia ever created a "live" messaging system? By that I mean one where a person's messages are sent and received in real-time, not requiring users to refresh the Wikipedia page. This would definitely make Wikipedia run far more smoothly and wouldn't be difficult to implement -- at least not more so than any revisions to Wikipedia software, period. It's particularly annoying when two editors are doing bold edits on the same page. The first one makes the edit. The second one sees "edit conflict" and then realizes that the several minutes of typing they did was in vain. A real-time messaging system would address this issue because editors could engage one another in real-time.

I know we could just use IRC, but it can be difficult to get all of the editors on all of the different articles we edit to all go on IRC and some of them may be confused by it. The above proposal would facillitate a stronger community regardless. Furthermore, having to login to both Wikipedia and IRC is tedious. It's already bad enough that you need separate accounts for every wiki run by Wikimedia. Zenwhat (talk) 04:15, 7 January 2008 (UTC)

It wouldn't be that hard to implement this. What would be hard is to then log such things properly. I mean, how do you even decide what page to put these conversations on? -Amarkov moo! 05:57, 7 January 2008 (UTC)
My proposal could be built on top of the existing messaging system where users edit talk pages, with the words "You have received a new message" being a widget on the left side of the screen that is updated in real-time, instead of something you have to refresh first in order to see. Zenwhat (talk) 08:45, 7 January 2008 (UTC)
This will probably not happen, simply because of the burden real-time messaging would create for the servers. Equazcion /C 11:02, 7 Jan 2008 (UTC)
If editors edit separate sections of articles, not entire articles, then edit conflicts are much rarer. For very active articles, taking several minutes to do an edit is generally a mistake. And when there is an edit conflict, it's not at all true that the editing was "done in vain"; pressing the "back" arrow gets you back to the edit box; if you were adding text, you can just open a new window, go back into the article, and add that text (it's unlikely that two editors are doing exactly the same thing to the same article at the same time, and, if so, then one of the two is editing in vain, real-time communications or not).
In short, real-time messaging is great if Wikipedia were to be a social networking site, but it is not. -- John Broughton (♫♫) 16:43, 7 January 2008 (UTC)
Wikipedia already has multiple IRC channels.[1]. --John Nagle (talk) 18:57, 7 January 2008 (UTC)
They aren't considered to be part of Wikipedia proper, and run independently -Halo (talk) 18:21, 12 January 2008 (UTC)

Equazcion: You seem to misunderstand how code works (server-side vs. client-side), so you exaggerate the burden it would have on the system by thinking such a program would involve something like "automatically-refreshing" Wikipedia over and over in order to provide live notification. That kind of system isn't necessary. As a specific example of how it could work: A flash module could be continually running client-side and, any time a message is sent, the server sends that module the notification, notifying the user in real-time with no additional burden on the server.

John Broughton: live notification is not intended to turn wikipedia into a social-networking site -- only to facilitate communication by giving users faster notification of messages while they're in the middle of editing. Gmail is not a "social networking site" but it still has live notification of messages simply because it's an efficient way of doing things.

John Nagle: The limitations of IRC have been pointed out. Zenwhat (talk) 23:50, 7 January 2008 (UTC)

I don't misunderstand at all. I know all about server-side vs client-side code and I wasn't suggesting an auto-refresh. I was referring to the messages the server would need to send to everyone upon receipt of a new talk page posting. Those would cause a burden on the server, and one that isn't worth the small benefit it provides. Also, reliance on Flash would be an accessibility issue. Equazcion /C 06:01, 8 Jan 2008 (UTC)
Why would be a greater burden? Let me see if I understand you correctly: As it stands now, whenever someone's browser requests a Wikipedia article, the server "checks for new messages." If there are, then you're given that message on screen, "New messages," when you receive the article content. This is true regardless of how many messages you have: If someone sends you ten messages in a row while you're AFK, then you come back to Wikipedia, the server only runs the algorithm once since it only runs when there's a page request and since you only refreshed once. Under a live messaging system, the server would send messages to the client every time a message is issued. Is that right?
I don't think it would take up that much resources. If Wikipedia was so concerned about that, we could help with the financial issue by deleting en masse all of the silly obscure geographical and musical directory entries, entered in by bots and cared about by no one other than inclusionists. From what I've heard, Wikimedia doesn't struggle financially (at least, not to support its website). Any additional costs seem to be outweighed by the benefits.
As for flash, you're right. JAVA? I'm not a programmer: Isn't it possible to program such things in XHTML, too? Zenwhat (talk) 11:25, 8 January 2008 (UTC)
Under the current system, when you receive a new posting to your talk page, a database flag is set. When the user loads a Wikipedia page, that flag is checked. It's an entirely passive system. The system you're suggesting would require the server to send out a real-time message at the time of each new posting, an active (not passive) action for the server. To keep it passive on the server side, the client could continually check the existing flag at an interval, but that too would put an additional burden on the server. Other than flash, the other possibility is Ajax, which because it makes extensive use of Javascript is still an accessibility issue, and the same would go for Java. Equazcion /C 16:50, 8 Jan 2008 (UTC)
Equazcion, you need to take one more thing into account, though: My proposal would also reduce the amount of edit conflicts, saving resources too. Yes, it uses more resources to send out messages actively than keep passive records that are only checked when the client makes a page-request. However, if editors can communicate in real-time, then they can more effectively collaborate in real-time. This is the justification for IRC, but as stated, the IRC has limitations. And again, I thought Jimbo has made it clear in the past that Wikipedia isn't struggling to support its website (since that's the least of Wikimedia's costs), so "we must save resources" doesn't seem like a valid argument. Zenwhat (talk) 22:35, 8 January 2008 (UTC)
I am a reasonably knowledgeable web developer. Flash is ugly and proprietary Java applets are just as bad - neither would create a good implementation. What you suggest could be made in JavaScript (see: Ajax (programming) or, preferably, Comet (programming)) combined with some server-side code but it'd be extremely technically complex to implement, a lot of JavaScript would need to be written (much of it fairly unpleasant), it would be a massive hit on servers (since there'd be thousands upon thousands of additional connections on the web server) and just be generally ugly. The disadvantages and complexity of such a thing would significantly outweigh any advantages - which is a key reason why sites like MySpace and Facebook don't do it either.
I can't see it ever happening but feel free to try to implement it into MediaWiki if you so wish to. Also, I doubt edit conflicts effect resources much, nor do pages that aren't frequently used - it's bandwidth and processing time that's expensive. Plus, there's no "justification" for IRC - it's not hosted on Wikimedia's servers and is ran independently from Wikipedia. -Halo (talk) 18:21, 12 January 2008 (UTC)
Oh, your Flash concept would not work due to firewalls and Flash being sandboxed. You'd need either the client to open a connection with the server and keep it open throughout (ala Comet) or to periodically connect to the server (ala Ajax), with the prior being preferable, so there would be a significant hit on the server either way. -Halo (talk) 18:29, 12 January 2008 (UTC)

## To assess the accuracy of Wikipedia, collect and release stats.

To assess the accuracy of Wikipedia, particularly to help Wikipedia:WikiProject Countering systemic bias: Statistics on Wikipedia edits should be released. Since outside criticism is welcomed (in fact, beneficial), releasing these statistics would very much allow researchers to gain a greater understanding of how Wikipedia works. Though they could collect them themselves, they would first need to have their account established as a bot. No such request has been made, leading to the certainty that no in-depth look at Wikipedia edits has ever been done. Furthermore, releasing these statistics would create an incentive to study Wikipedia, using statistical analysis.

Some basic examples:

• Measuring the average number of edits made on articles relative to the average number of users, as a rough measure of the degree of collaboration as opposed to edit-warring
• Measuring the average edits on certain subjects
• Stuff listed on WP:Systemic bias relative to everything else
• Fringe theories overall, aside from what's listed on WP:System bias, relative to everything else
• General subjects overall
• Basic topics relative to advanced topics
• Whether most users focus on a broad array of articles, or focus on a narrow few
• The average amount of time and the amount of blocks it takes for a user to be banned for violating policy

Measuring trends could give empirical indications of the effects of certain Wikipedia policies, instead of just the common, "Well, I think policy should be..." based on personal opinion.

Any of the fine bot-programmers that work for Wikipedia could do this, and the folks over at the math reference desk could draw correlations. Myself, having taken Statistics 101, with the raw data I could provide correlations. Zenwhat (talk) 01:47, 12 January 2008 (UTC)

There has been some amount of wikipedia statistical work, often presented at wikimania, see if you can google some of the papers and presentations. :-)
I try to base as much as possible of what I say on a solid statistical basis, but I don't always have statistics on everything, and must rely on inference, extrapolation, or intuition instead. (But not typically on my personal opinion, as some people assume.)
It would be silly to use a bot for this work. Instead, download a wikipedia database dump, and use SQL to do the heavy lifting for you. After that, it's just a matter of applying your statistics 101 to the numbers. Using this method, your sample size is the entire population size ($n=N$). This means that you can use simpler formulas and can get very accurate outcomes, but also that you need to watch out, since many statistical formulas assume you are merely taking a partial sample out of a larger population.
An alternate approach some people use is to hand-pick representative samples and use statistics on those samples. A typical sample size might be 50-100 pages. This is a lot less accurate, and has a possibility of bias. But that's what statistics was invented for, right?
I look forward to reading your results! --Kim Bruning (talk) 02:57, 12 January 2008 (UTC)
How many gigabytes does Wikipedia take up and what are the server's bandwidth restrictions on such data dumps? Also, I'm no programmer, so I wouldn't be able to write a program to analyze it. I'd need the data to be in a spreadsheet. Zenwhat (talk) 05:41, 12 January 2008 (UTC)
The most recent dump of "All pages, current versions only" which includes userpages and talk pages, is 5.9 GB, compressed into bzip2 format. The "All pages with complete page edit history" dump is currently running with an ETA of 2008-02-26. You can download them at http://download.wikimedia.org/enwiki/20080103/. Mr.Z-man 05:55, 12 January 2008 (UTC)
It works out to about 35-40GB when imported into a MySQL database. --Carnildo (talk) 06:30, 12 January 2008 (UTC)

Actually, if you're mildly intelligent and aren't too horrible at logic, you can probably sit down with a book and a machine and learn enough SQL over a rainy weekend to be able to get nice numbers for your spreadsheet already. --Kim Bruning (talk) 07:27, 12 January 2008 (UTC) For comparison: I learned it on a sunny summer afternoon while sitting under a tree, enjoying the shade. :-)

I doubt anybody would have a problem with statistics being calculated and released, but I don't think you should expect, nor do I think you'll be able to, get others to do the technical work for you. I just don't think "I would like someone else to do the technical work for me" is something that gets things done here, especially as the technical work is usually the most boring and complex part. As an example, look at the long-proposed user survey which a minority have been trying to establish since 2005. Their lack of success is due to a combination of not wanting to do the technical work themselves, their inability to recruit developers to work for free, not being able to get enough people interested in the topic and, in my opinion, their arrogance and attitude towards certain people not doing them any favours. As previously suggested, the best way to go forward is to download the database dump, put it into MySQL and doing the SQL yourself. -Halo (talk) 16:08, 12 January 2008 (UTC)

Thank you Wikipedia community for being so helpful. "If you're mildly intelligent and aren't too horrible at logic" - do I detect the presence of sarcasm there? I am not a programmer, but I intend to make this project one of my primary goals, since I care about the Wikipedia project and not just suppressing the views of editors I don't like. When you look at the number of bots on Wikipedia and see WP:Expert rebellion, it's not all that clear to me that Wikipedia has a drought of professionals who have offered to work. They have left because they have been treated with hostility the way you are treating me now. Zenwhat (talk) 00:31, 13 January 2008 (UTC)
Sarcasm doesn't carry online, so I try to avoid it. SQL really is fairly easy to learn, it was originally intended for use by end-users. It does help if you understand (boolean) logic and some basic set theory, but basically anyone with a decent brain can learn it pretty quickly. --Kim Bruning (talk) 02:41, 13 January 2008 (UTC)
Erm, excuse me? You seem to be very, very tetchy. I didn't say anything hostile at all - I was trying to be helpful and showing, with an example, that 'recruiting' programmers to do work for you will be extremely difficult and doing it yourself will be much much easier. If, as you say, "Wikipedia has a drought of professionals" then you're just proving my point - I'm sorry you don't like my answer but frankly it's the truth. -Halo (talk) 04:36, 13 January 2008 (UTC)

## New archiving proposal

I have an idea for a new way to archive pages here User:Mbisanz/ArchiveSpace and would like to test the water of what other users think, before I go ahead and create a bugzilla or project-space page. MBisanz talk 02:35, 12 January 2008 (UTC)

I don't like this suggestion - I see little significant advantage, would require significant technical work, the suggested link layout looks fairly ugly and I don't see the problem as big enough to warrant correcting. There will also be significant opposition against creating any new namespaces (i.e. the table namespace). A better way would simply be to improve the crappy search engine.
In case you aren't aware, creating a poll at this stage is considered bad form, and isn't how policies such as this are generally implemented. -Halo (talk) 16:21, 12 January 2008 (UTC)
Sorry, I was going off of what I've seen at other proposals. Yea, I know, I saw the table namespace and was disheartened at the opposition it faced. Another idea would be a way to tag archive pages for non-inclusion in normal searchs (ie a search button that says "include archived pages"). MBisanz talk 04:01, 13 January 2008 (UTC)

## Dealing with vandals

I am rather concerned about the user 206.82.17.190. This user, I noticed in December 2007, had vandalised the page on aging, by replacing the good work which lots of people had done on this article with a single sentence "Aging is what you get when you get freakin' old" and if you look at the other contributions this user makes, it seems that this user is keen to vandalise Wikipedia. Is there a way to stop users such as this editing - sorry, vandalising - Wikipedia? ACEOREVIVED (talk) 22:05, 12 January 2008 (UTC)

User:206.82.17.190 is currently blocked until the 14th. See WP:VANDAL for general anti-vandalism policy. 131.111.8.96 (talk) 23:24, 12 January 2008 (UTC)

## Number of hits for all articles

Hello, do you plan to show the number of hits for each article, like they do on youtube? Thank you, Adrian Comollo (talk) 20:49, 13 January 2008 (UTC)

Individual page statistics are turned off for performance reasons. There was Wikipedia:Popular pages, but it is inactive. And there is an outside page that lists the most visited Wikipedia pages of the month, but it is not exactly what you are looking for. -- ReyBrujo (talk) 21:17, 13 January 2008 (UTC)
http://stats.grok.se/ gives the number of hits per day for articles. Tra (Talk) 21:48, 13 January 2008 (UTC)

## New CSD proposal

Cross posting to inform the community about a new proposal. Discussion taking place at Wikipedia talk:Criteria for speedy deletion#New criteria proposal. - Rjd0060 (talk) 02:55, 14 January 2008 (UTC)

## Current events subportals

A discussion about most of the Current events subportals has been opened at Miscellany for deletion. You are invited to participate in discussing the fate of these subportals. I thought I should let some more people know about this, than currently are aware, and therefore figured the VP should be informed. --TheDJ (talkcontribs) 17:38, 13 January 2008 (UTC)

## Dollar versus Euro

A quote from the (English) Euro article:

It is the single currency for more than 320 million Europeans. Including areas using currencies pegged to the euro, the euro directly affects more than 480 million people worldwide. With more than €610 billion in circulation as of December 2006 (equivalent to US\$802 billion at the exchange rates at the time), the euro is the currency with the highest combined value of cash in circulation in the world, having surpassed the U.S. dollar.

There are a lot of articles which use the Dollar when speaking of monetary values (for instance, GDP articles), even the regional pages (non-English) use the Dollar.

Considering the above quotation, should that not be the Euro as it is the biggest currency in circulation?

If not, what about Countries using the Euro; should not their regional pages use the Euro? NitroX infinity (talk) 10:11, 27 December 2007 (UTC)

IMO countries that use the Euro should have values of for example GDP both in Euros and in Dollars. The fact is that using the dollar for such figures is global standard practice, also for countries in the Euro zone. I suppose this is because most raw products are still traded in Dollars, but for whatever reason it is, I think we should just keep to what is most commonly used in litterature and media. Martijn Hoekstra (talk) 10:16, 27 December 2007 (UTC)
I guess that this matter will be re-examined in a year or sooner. Events in global economy tend to unfold rapidly. Waltham, The Duke of 17:23, 27 December 2007 (UTC)

Thanks, Fuhghettaboutit, for pointing me here, and Teratorniss, I've read your notes on currency. I'd like to add that I support a move to Euro as the standard, given its position as the most powerful currency today, unless where the country's own currency is used. The fact that people have become accustomed to seeing things in dollar or it is the going practice to do so, is immaterial - the same was said when the metric system was suggested and today barring a few exceptions, it is the system of choice. Also, an Euro is an Euro - no confusion there - whereas a dollar is anything from American, Australian, Canadian to Zimbabwean. Finally the Euro represents 27 or so countries, whereas the USD represents but 1.

Regards, --Rui ''Gabriel'' Correia (talk) 07:51, 3 January 2008 (UTC)

Wikipedia's goal is to describe the world, not to change it. We'll switch over to Euros when Euros become the major currency for describing prices, and not before. --Carnildo (talk) 08:32, 3 January 2008 (UTC)
Maybe it is the best to let it depend on the source for the numbers. If that source speaks of (US) dollars, use the (US) dollar, if it speaks of euro, use euro, and if it speaks of South African Rand, use the South African Rand (and give a link to the article of that currency, so that people can estimate how much it is worth). Calculating euro's or dollars here is dangerous, as these rates tend to change fast these days. effeietsanders 10:10, 3 January 2008 (UTC)
I think that this statement makes the best sense, but if we end up with information from multiple sources in a section, we need a mechanism for continuity. --Kevin Murray (talk) 11:38, 3 January 2008 (UTC)
There's a reason the amounts have defaulted to American dollars up until now. Wikipedia is fundamentally an American institution. It was started in America by an American and is now run by an American organization. We're governed by American laws, our documentation uses American language and spellings. Agreed that articles should use the monetary unit quoted in the source, but when in doubt, I don't see the logic in in going with whichever monetary unit happens to be the most powerful at the time. Sure, an argument could be made that the EN Wikipedia should be considered just as much British as it is American, but in the end I just don't feel that's the case. Equazcion /C 12:04, 3 Jan 2008 (UTC)
Just bad luck that most of the versions of Wikipedia are in languages not spoken by Americans hae? :P And that currently the majority of the Wikimedia Foundation Board is from Europe :). Although the Wikimedia Foundation is for historical (and probably also other) reasons located in the US, it is a world wide foundation. Every editor has to obide to it's own laws, so there is not one single law Wikipedia falls under. (note by the way the difference between Wikipedia and the English language Wikipedia). I think en.wikipedia.org should be considered neutral and not British, American, Canadian, Australian, Chinese, Dutch or Antarctican. Especially concerning the content, it should be neutral. I think we all agreed on that with Wikipedia:Neutral Point of View? However, please bear in mind that our location influences how we see Wikipedia. An American like you might more easily see Wikipedia as an American project then a Dutch like me. effeietsanders 12:41, 3 January 2008 (UTC)
Oh I absolutely agree that it should be neutral wherever possible, but neutrality isn't an option in this case. A choice does need to be made, unless we invent a new monetary unit (eurodollars? wikipediollars?). Wikipedia's database servers are located in America (to my knowledge) and are subject to American laws -- and besides which, again, the EN version is an American entity by other evidence, such as, again, the American style used in documentation and in the interface, in its beginnings, and in the way the articles have, again, defaulted up until now. If wikipedia were originally European we'd be seeing something very different (yes, in the EN version) than what we're seeing now, and I wouldn't be against keeping Euros as the default unit, as it would have been from the beginning. But as it stands, I'm not sure what argument could be made for making a change to Euros (as the default), but if the fact that it's the more powerful monetary unit at the moment is the only reason, I say that's not a good reason. Equazcion /C 12:55, 3 Jan 2008 (UTC)
But why defaulting the currency at all? If only the exchange rates were stable... Of course, when we're speaking of South-American countries which rate to the US dollar is fixed (if that is the case) then mentioning the US dollar is no problem to me. Or if the currency is highly stable against US dollar (I expect this for instance for the Canadian dollar?). But a similar situation comes into mind when we go to certain European countries, such as slovakia, czechia, latvia etc. I can imagine their currencies are highly stable against the euro, so mentioning also the euro would be a logical approach. But if it is about a currency that is stable against neither of the euro or US dollar, just leave it, and do not mention either. If you mention a value right now, it is worthless next week, because the rates have changed much. A tool that automatically calculates the current value in dollars and euro's would be useful here by the way. But that is besides the point. I personally prefer correctness before ease, so if there is doubt, do not use dollar, as you suggest, but to use neither.
If you are comparing values from different countries with each other, yes I agree there is a need for a common currency. A neutral would be preferred yes, but you are right, there is none. I would suggest in that case to just use both Euro and US Dollar. Most of the times this will be in tables anyways, so it would be not so much trouble to add just another column to the table. effeietsanders 13:10, 3 January 2008 (UTC)
"Default" wouldn't mean changing the currency where one clear currency is used in the source. It would just be for instances where there are multiple currencies to choose from. Since this is the English Wikipedia, I don't see any question of what that default should be. The other language versions can use their native currency as the default, and English articles would use our (native-English-speakers) native currency. In fact I think a better argument could be made for using the GBP than the Euro. Equazcion /C 13:23, 3 Jan 2008 (UTC)

Putting both USD and Euros into articles would be rather silly, as the constantly fluctuating exchange rates would render one or the other horribly inaccurate within days. Best to just use the same units used in the source material--and if said source material lists multiple currencies, list them all, in the same order they're listed in the source. Rdfox 76 (talk) 15:11, 3 January 2008 (UTC)

Generally, I think sources use one basic currency to make their calculations in. I think that one should be used if there are multiple options (as in: 200 options, yeah, you'd have to pick one or two). The other occasion I could imagine is if a country has a fixed exchange rate against a currency, then it could make sense to mention that one too. (i.e. south american countries having a fixed rate against US dollar, or european countries about to enter the euro zone, having a fixed rate against the euro). Another occasion I could imagine is if a country would switch currency at some point. Yes, then you mention euro too for france before 2002... (to compare). When multiple countries are compared, the currency should be used that was the basis in the research used as source. I expect that this will for historical reasons mostly be US dollars. But if for instance it would be derived from a spanish institute pre 2002, yes, it could be that it would be compared all in peso's! (and against the exchange rate of THAT time, not of the current one of course) effeietsanders 20:18, 3 January 2008 (UTC)
Why not use a template whereby you add, for example, {{dollars|33}} and it results in \$33 (~€22) and is updated semi-automatically as the exchange rate changes? You could even sort out inflation using clever syntax. -Halo (talk) 00:35, 11 January 2008 (UTC)

We use USD for one simple reason. It is what the world uses for economic comparison. When currencies are compared for their strength or weakness, they are typically compared against the US dollar. This is because <insert enlightened macroeconomic reasoning>, and thus the dollar is the best candidate for the task (I remember hearing it in economics class once, but I can't remember, and it doesn't matter here, because so long as the world uses it, we use it). Also, it is important to note, that in the manual of style, at WP:\$ (note the dollar-biased shortcut, good for a chuckle):

• In country-specific articles, such as Economy of Australia, use the currency of the country.
• In non-country-specific articles such as Wealth, use US dollars (US\$123).

Thus, I believe this is solved.--Vox Rationis (Talk | contribs) 23:43, 7 January 2008 (UTC)

USD is the defacto world reference currency - most countries that tie their currency to another tie it to the dollar, for example - so stick with it for now. BUT, to keep it all square, why not include the exchange rate of several other major currencies, or make them easily available? Wikimedia could set up a page of exchange rates for (say) each quarter or month in the reasonable past, and retain weekly or daily values going forward. Mark each date entry with an anchor fragment, and articles would be able link into the applicable date(s) in the exchange rate page(s). Will 15 January 2008 / 2008/01/15 —Preceding unsigned comment added by 210.8.170.141 (talk) 21:57, 14 January 2008 (UTC)

## Suggestion: article-naming wizard

Is it possible to add a wizard to Wikipedia to guide editors in following naming conventions for new articles and moved (renamed) articles, as well as articles resulting from merged articles and split articles? I have in mind a series of menus beginning with options, such as the following.

a person?
a non-human living organism?
an abstract idea?
a field of study?
a part of a human?
a part of a non-human organism?
an event?
a procedure?

Subsequent menus would fine-tune the type of article even further, and then the editor would be presented with a checklist of things to consider, such as:

capitalization, hyphenation, American and British English spelling differences, diacritics, numerals, and so forth. Someone can decide on what specific guidelines should be applied to specific types of articles.

At this time, we have article names such as the following.

Please notice the difference in capitalization. Recently, I moved Bali Communique to Bali Communiqué, and 100,000 year problem to 100,000-year problem. (Please see list of English words with diacritics and hyphen#Compound modifiers, paragraph 5: "Hyphens are used to connect numbers and words in forming adjectival phrases". There is also Wikipedia:Manual of Style#Hyphens, section 3, point 7.)

This is just a preliminary plan for what such a program might do. More refinement of the plan would require more time and reflection than what I have spent on this. -- Wavelength (talk) 07:48, 14 January 2008 (UTC)

## Silence as consensus

Made a new essay at Wikipedia:Silence as consensus AzaToth 19:21, 14 January 2008 (UTC)

Would make a good guidline. Many times, espeically on page moves, no people may comment. MBisanz talk 19:24, 14 January 2008 (UTC)
Done this on some page moves. Leave a comment, nobody responds for a week, move anyway. Works quite well. 20:29, 14 January 2008 (UTC)
Isn't this policy already? How could the reverse be true: wait forever until a quorum arises? A long standing policy is to be bold. Even on page moves, you do not need to wait unless the move is contentious. Deafening silence has to be interpreted as no contention. Cheers! Wassupwestcoast (talk) 21:06, 14 January 2008 (UTC)
Is this meant to go beyond WP:CON? ("In essence silence implies consent if there is adequate exposure to the community. In the case of policy pages a higher standard of participation and consensus is expected.") And is this proposal self-referential? --Boson (talk) 21:10, 14 January 2008 (UTC)
Did you take that quote from there? As I can't find it. AzaToth 22:04, 14 January 2008 (UTC)
Second paragraph, third/fourth sentences. Algebraist 22:15, 14 January 2008 (UTC)
Ah, there it was, had trouble finding it. AzaToth 22:21, 14 January 2008 (UTC)

## Suggestion: researching correlation of article size with number of incoming links

I have noticed what appears to be a high correlation between (a) the number of lines in a typical article and (b) the number of links listed when I click on "What links here". I do not know to what extent, if any, "a" affects "b", or "b" affects "a", or "c" (another variable) affects "a" and/or "b". Also, there may be some significance to different types of lines in the article (for example, main text, outgoing internal links, and external links) as well as different types of links listed in "What links here" (for example, article pages, article talk pages, and user talk pages). (My interest in this correlation began with my interest in adding incoming internal links to short articles in order to help in their expansion.) Do the benefits of researching this correlation outweigh the costs? Also, please consider correlations between members of other pairs of variables in Wikipedia statistics. -- Wavelength (talk) 21:49, 14 January 2008 (UTC) -- Wavelength (talk) 19:06, 18 January 2008 (UTC)

## A Proposal to Hide Popular Culture References and thus Improve This Encyclopedia

It is my proposal that the 'popular culture' segment of each article (especially articles dealing with classical themes, literature, medicine, &c.) should be hidden from the reader's purview. Although I cannot squander much time expounding my reasoning as to why, I shall, nonetheless, display several examples in order to sate my reader's curiosity :

• 1. In an article dealing with the ancient Roman poet, Catullus, I was compelled to read the following line within the popular culture section : ' According to the webcomic, Achewood, Catullus is the first poet who ever got his bone on. ' — Alas, my mind is now irrevocably fraught with useless and undesirable knowledge.
• 2. In an article on 'webbed toes', I am apprised of the dubious fact that Stalin may have had webbed toes and Marge Simpson does, indeed, have webbed feet! — What a delight! Truly!
• 3. The article on Crime and Punishment : I am now afflicted with the useless knowledge that 'The Sopranos episode Another_Toothpick by Terence Winter shares aspects of plot and character.'

The list is endless! Please, for my sake! Let us come to an agreement you pop-culture-loving cretins! The following design would be a consummate solution : in this article on the French wikipedia, there is a tab that reads 'dérouler' on the bottom of the page (to unwind, unroll). If we could do this with the popular culture segment of each article (in the very least, all classical, literary themes), wikipedia would actually be a quasi-respectable encyclopedia. -- Grammaticus VII (talk) 18:42, 11 January 2008 (UTC)

See WP:TRIVIA for more information on actual policy regarding these sections. The implementation I prefer is where these sections only cover references which are significant to the understanding of their the subject they are on, or the material they refer to, and can be referenced properly. The Marge simpson reference on webbed toes to me sounds useful, as it is a major appearance in the media of a character with webbed toes. The Catullus mention on the other hand adds little depth for either subject matter, and the crime and punishment relation sounds like it should not be there at all, being original research. LinaMishima (talk) 19:08, 11 January 2008 (UTC)
In this particular case I agree with LinaMishima. The problem is that the boundary is a matter of opinion. I suggest either (a) we set up a committee to decide whether particular instances are useful or worthless, leading to "case law" and eventually clearer guidelines; (b) or we ask the relevant Wikiprojects to do the same but post links on a central Talk page, again leading to "case law" and eventually clearer guidelines. Philcha (talk) 19:34, 11 January 2008 (UTC)
Like above. In some cases you need pop culture references for complete understanding, like Paradise Lost->His Dark Materials. Will (talk) 20:21, 11 January 2008 (UTC)
LinaMishima, the fact of the matter is: I do not wish to read about Marge Simpson -- seeing as a personage of this nature would not be displayed in your standard encyclopedia. In AD 2200 (if all goes well), do you believe that Marge Simpson will even be known? It is highly doubtful. Will Catullus or Dostoevsky? It is certainly more likely.
I believe that timeless subject matter warrants more attention and boasts more merit than dense, hoi polloi subject matter that will incontrovertibly perish in 50-60 years (if that). Unfortunately, it is impossible for the eye to evade the text which reads 'Marge Simpson', 'Danny Glover', &c. — and thus, I have presented this most wonderful compromise as a remedy to this flaw!
For I am not stating that the trivia should be deleted! No, no! -- merely hidden. (For this will also impress the pedants and the pundits, I assure you).
Unfortunately, the illustrious wikipedia is plagued with popular culture ad nauseam. Though why must he who fetters himself to classical themes also incur the horror of this plague?
(Moreover, to stress this facet, Bill -- I do NOT wish to delete the popular culture notes -- only to hide them). -- Grammaticus VII (talk) 20:29, 11 January 2008 (UTC)
Wikipedia is not a paper encyclopedia. Will there be a Wikipedia in 50 years? Maybe. Is there one now? Yes. If in 15 years something is no longer a popular culture reference, it will be removed. If its unsourced, irrelevant, minor, and/or incredibly stupid, you can remove it now :) You may not want to read about Marge Simpson, but other people do. Mr.Z-man 01:25, 12 January 2008 (UTC)
Cultural references can be encyclopedic. I.E. "Where's the Beef?" It's very difficult to define a policy to establish whether something is relevant. In practice, articles on pop culture have gotten out of hand because of a loose enforcement of WP:RS and WP:V, though it also has worked in the other way, with relevant pop culture stuff like Wikiality being removed. Zenwhat (talk) 01:51, 12 January 2008 (UTC)
Did you not read my initial proposal? Did you not observe the French wikipedia link and subsequently mark my earnest intention (which was to place the said information in a drop-down menu)? Nevertheless, I shan't endeavour to convince you inattentive laggards any further. Yes, yes... this is an argumentum ad hominem. —Preceding unsigned comment added by Grammaticus VII (talkcontribs) 03:10, 12 January 2008 (UTC)
Except that French Wikipedia article uses the box to hide sexually explicit content, not trivia. However, it is generally accepted that the English Wikipedia is not censored and that such boxes are not used for actual article content. Mr.Z-man 04:12, 12 January 2008 (UTC)

What we really need is a new Wikimedia project, a "WikiAllusion", comparable to the The Oxford Dictionary of Allusions or Merriam-Webster's Dictionary of Allusions, but with an expanded scope to cover all cultural topics. Of course, the most relevant allusions would remain in the article, but the complete collection would be on the new project. This is similar to the development of Wikiquote, which collects many quotations that would be excessive to all put in the Wikipedia article.--Pharos (talk) 03:46, 12 January 2008 (UTC)

Oooh! I like that idea. Where can I vote? -- Fullstop (talk) 15:59, 12 January 2008 (UTC)
http://meta.wikimedia.org/wiki/Proposals_for_new_projects. —Random832 15:21, 15 January 2008 (UTC)

## Experiment for contentitious articles

It appears to me that there is a wide disparity in opinions between editors and admins who deal mainly with controversial articles, and those who do not, in how to deal with POV pushing and meat puppets and sock puppets etc on controversial articles. This is particularly true in topics in which there is a scientific component such as medicine, and creationism.

I propose that we try to devise some sort of test where two articles in the same highly controversial area, of roughly the same quality, be managed differently for an extended period, such as 6 months or more. On one article, "regular" editors and NPOV defenders of various sorts would be discouraged from editing, and the article edited at will by those with specific agendas. Arguable NPOV edits would never be reverted in this article. Regular editors would be discouraged from responding to queries about content on the talk page. Basically, a laissez faire policy would be adopted.

The other article would be managed in a normal fashion with NPOV reverts etc and patrolling by regular editors and admins.

At the end of 6 months, the two articles, and their histories over the 6 months, would be submitted to some outside neutral group for examination. The quality of the articles over the 6 months could be evaluated.

This sort of test might be useful data in determing whether we are wasting our time and in fact giving Wikipedia a bad name by discouraging POV edits, and hurting Wikipedia's reputation in the process.

Comments?--Filll (talk) 00:26, 13 January 2008 (UTC)

we've done it at various times in various ways. an unprotected unwatched contentious article ends up with zero content. DGG (talk) 00:56, 13 January 2008 (UTC)

True, but those were not really controlled experiments and not carefully evaluated. And I have heard repeatedly, over and over and over, from assorted editors that protecting NPOV is a bad thing, and that it gives Wikipedia a bad name and creates bad will. I will grant you that this is possible. And I am willing to entertain that possibility.

At a minimum, it is claimed that editors should be exceedingly gentle when dealing with NPOV edits on controversial articles even those articles that are regularly attacked by dozens if not hundreds of trolls, sock puppets, meat puppets, POV pushers, fringe elements etc. Well I will admit that maybe our "storm trooper" tactics are more trouble than they are worth; they waste the time of experienced editors and administrators, they cause dissension and, as can readily be seen by visiting Wikipedia Review or The Register, this behavior causes bad will and gives Wikipedia a bad name.

So lets test these assumptions, at least in a small way. I get tired of being told that trying to have NPOV articles is a bad thing, and that people who believe that the scientific POV is a bad thing and evidence of how biased and ugly a place Wikipedia is. So lets test this a bit. Why not? --Filll (talk) 01:14, 13 January 2008 (UTC)

I would be most interested to see this experiment run, if it is possible to do so. The amount of time and energy expended by editors on contentious articles must be huge. A different approach might allow a lot of this effort to be used more productively. Wanderer57 (talk) 17:03, 15 January 2008 (UTC)

I have no idea where to start with contentious articles, since I've been trying to just keep all the others working for now, and haven't really had time to even start looking at them. Starting some experiments would be nice. :-) --Kim Bruning (talk) 22:32, 15 January 2008 (UTC)

## WikiGlobe: A Proposal for an Alternative Interface

The following is an outline for what I believe to be a unique approach to information consolidation, organization, and most importantly, access. It would NOT exist independently of Wikipedia. I have done my best to describe a scenario that would work alongside the existing structure of Wikipedia, one perhaps best thought of as an alternative, intuitive search engine.

A Globe With the cooperation (or at least the licensing) of something akin to Google Earth, the key element to this “search engine” would be an interactive model of our planet. Not only would users be able to interface with this representation in all of the spatial aspects to which we have become accustomed, there would be an additional dynamic element:

A Time Scale Differentiating this globe from the ones we typically use for online study or navigation, an adjustable time scale in the form of some sort of slider would allow the user to “view” the globe as it was at the particular time period selected.

The Combination of these two elements would allow the user to interface with history, geography, and a wealth of other information in an intuitive fashion never previously possible. A given location on the globe, at a given time interval, would be hyperlinked to the relevant categories and articles on Wikipedia. In this way, users would be able to view events in history in the appropriate geographical and relational context. One could compare, at a glance, the states of various parts of the world at any given time interval, thus providing an unprecedented amount of perspective into multiple areas of study.

All of this would simply be a means of intuitively navigating Wikipedia’s increasingly complex network of cross-references. Articles can already be broken down and sorted by factors such as date and location. A person, state, or event containing these two discrete pieces of information could easily be incorporated into this “meta”-search (i.e. the Battle of Gettysburg would be accessible through the Gettysburg, Pennsylvania area of the global map from July 1st to 3rd, 1863 on the time scale).

Without too much close inspection, there are clearly many issues to resolve before such a project could be implemented. For example:

Time intervals The “time slider” that carries the globe through its various epochs will necessarily be divided into discrete intervals. For much of the last few centuries, it could be broken down all the way to individual days (or further, where such information is available). The farther back we travel in history, however, the more ambiguities arise, and the farther apart our time intervals would have to be. Fortunately, this would be completely dependent on the amount of information available in the articles themselves, and would perhaps prove to be most troublesome only as far as in the designing of a manageable “time slider.”

Borders National boundaries, as well as other pieces of meta-information (all of which a user would be able to toggle on or off, as layers) would understandably be a point of contention, especially as they become blurrier (or nonexistent) farther back in the past. This may in fact prove to be the most challenging aspect of this project, partly due to the fact that we would be unable to get this information exclusively from existing articles (i.e. articles do not come encoded with a border-stamp). However, the establishment of such borders would be crucial to this representation, because in effect the national, regional, or municipal boundaries in question will be the boundaries of the hyperlinks to the corresponding articles.

Geological processes Should the project be expanded to go sufficiently far back in history as to necessitate the representation of continental drift or other such things, the model that worked for (most of) our human history would break down. However, there is no reason that the interface couldn’t extend into these distant periods, albeit at exponentially larger time intervals than are used for modern history.

Method of Integration Wikipedia’s existing methods of categorization by time should make the integration of key historical articles relatively simply in that aspect. As far as the location data, I propose that relevant articles be stamped with one additional piece of information that ties it to the proper area on the global interface. This could become a standard part of new article creation, and existing articles could simply be tagged much as they are, daily, for myriad other purposes (in need of review, etc.).

Representation of Vagaries At what time intervals, and in what locations, should inexact information be placed (if at all)?

In all I'm sure this will be dismissed as Utopian or impractical, but I believe that at heart such an interface could be relatively simple, and within the grasp of a few software developers competent in Flash or whatever it is the kids use these days. Obviously the cross-referencing and tagging would be a never-ending task of refinement. However, given the astonishing growth of WikiMedia in only the last few years, I think that organizational tools such as this are inevitable. The only question is, when do we start? (If I had the slightest clue what I was doing I'd be all over it).

Pstanley (talk) 04:56, 13 January 2008 (UTC)

An interesting idea, but it would be both technically and functionally difficult to implement as suggested, least of all due to free-content issues that collide with Wikimedia's free content policy (i.e. Flash and Google Earth). If you are interested in developing it, starting small and using free content such as Visible Earth to create a basic 2D interface to keep things simple would be a good idea and developing it onwards from there -Halo (talk) 05:23, 13 January 2008 (UTC)
We do have m:WikiMiniAtlas which gives a 2D interface for plotting Wikipedia articles on a map. To implement your suggestion, a filter could perhaps be added to restrict articles by the time period they are applicable in. The problem is that there are a lot of different ways of recording dates in articles - infobox parameters, categories and dates written inline. Therefore, the tool used to extract the relevant dates in articles would need to be programmed with all the different ways of specifying dates, which may be difficult. Tra (Talk) 15:11, 13 January 2008 (UTC)
I think (from a quick poke around http://www.mediawiki.org/wiki/Manual:Hooks and http://upload.wikimedia.org/wikipedia/commons/4/41/Mediawiki-database-schema.png) that a hook could be added to http://www.mediawiki.org/wiki/Manual:Hooks/ArticleSave or http://www.mediawiki.org/wiki/Manual:Hooks/ArticleSaveComplete to parse the contents of the article for dates as or after it has been saved. The dates that are discovered could then be used to add links to the article in whatever data set is used to back the new interface. To catch all the dates you need a set of regular expressions (easy in PHP) to match the dates, and a way for users to report date formats that aren't being caught. I'll help with the regular expressions, but this needs input from someone who is familiar with mediawiki's internals to get the design right. Willllll —Preceding comment was added at 22:31, 14 January 2008 (UTC)
The problem is that just because a date is written in an article doesn't necessarily mean that that is the date when the event took place. for example, whilst a human would be able to read Battle of Hastings and see that it took place in 1066, a computer would see other dates written in that article, and would not know whether they referred to the battle, or to another historical event that's relevant to the article. Tra (Talk) 23:47, 14 January 2008 (UTC)

Wikipedia is free content. Feel free to download a database and implement your proposal! :-) --Kim Bruning (talk) 22:33, 15 January 2008 (UTC)

## WP:CENT

In order to continue to ensure that policy discussions get adequate attention, I would like to propose putting Template:Cent (or, if there are concerns about security, subst'ing it each time it is updated) on the watchlist and recent changes). —Random832 16:08, 15 January 2008 (UTC)

It doesn't get updated that often and updates are not time-critical, it would be easier to full protect it and have non-admins just request an addition on WP:RFPP (which will likely get a faster response than {{editprotected}}) for new additions rather than having to request re-subst'ing whenever a change is made. Mr.Z-man 03:00, 16 January 2008 (UTC)
I strongly objected to a similar proposal here. --MZMcBride (talk) 03:04, 16 January 2008 (UTC)

It wuold need to be subst'ed, testing reveals that noinclude and includeonly don't work normally on the watchlist without it. —Random832 04:16, 16 January 2008 (UTC)

## Suggestion: searching for contributions by a specified editor to a specified article

I am interested in there being a program which would facilitate a search for contributions by a specified editor to a specified article. Many articles have long histories of contributions, and also many editors have long histories of contributions, so, at the present time, finding such information can sometimes require visually scanning many contributions on many pages.

This program could be designed with the possibility of choosing one of two methods to perform such a search. (An editor can be either a registered user or an IP address.)

With one method, a searcher first selects an article and then clicks on a button ("Editors"), which displays the number of editors and a list of all editors which have contributed to that article (and the number of contributions to that article by each editor). Clicking on that number displays a list of all contributions by that editor to that article. Clicking on each contribution displays a page showing a comparison of different versions before and after that contribution.

With the other method, a searcher first selects an editor and then clicks on a button ("Articles"), which displays the number of articles and a list of all articles which have been edited by that editor (and the number of contributions by that editor to each article). Clicking on that number displays a list of all contributions to that article by that editor. Clicking on each contribution displays a page showing a comparison of different versions before and after that contribution.

This program for a two-dimensional search might be extended to other namespaces besides the article namespace. -- Wavelength (talk) 16:51, 15 January 2008 (UTC)