Template talk:Wikipedia languages/Archive 5

Archive 1Archive 3Archive 4Archive 5Archive 6Archive 7Archive 8

ltwiki -> 100,000

The Lithuanian wiki is now up to 100,000 articles. Its depth is 19, and a random sampling of 25 articles gives:

  • 4 articles with relatively long copy
  • 10 articles with medium-length copy
  • 11 articles with 1 or 2 short paragraphs.

However, since the wiki is only just over the threshold, it might be better to wait a few days, to see if it settles above 100,000. 210.84.62.15 (talk) 06:34, 20 January 2010 (UTC)

Lithuanian Wikipedia now 100 000

Lithuanian wikipedia has reached 100 000 articles. Please update the template. --Vpovilaitis (talk) 07:44, 22 January 2010 (UTC)

Czech Wikipedia

The Czech Wikipedia has reached 150,000 articles. Could someone add it to the right section? --Yair rand (talk) 22:01, 17 February 2010 (UTC)

Done. —David Levy 01:05, 18 February 2010 (UTC)

About Wikipedia languages on Main Page

Why is there "More than 40,000 articles" and why not "More than 50,000 articles"? Also they are not up to date. Pelmeen10 (talk) 20:08, 15 March 2010 (UTC)

It is up to date. According to the template documentation "This is not a complete list of Wikipedias containing 40,000 or more articles; Wikipedias determined to consist primarily of stubs and placeholders are omitted." --Yair rand (talk) 20:18, 15 March 2010 (UTC)

Hindi Wikipedia Redux

Given the recent changes on Hindi Wikipedia, I would like to reopen discussion on including Hindi in this template. During the last one year, we have had several editors regularly contributing good-sized and good-quality articles. Random test of 50 articles today yielded: 18 placeholders and worse; 18 stubs providing usable definitions; 7 mid-length and 7 full-length articles.. In a wikipedia of size ~54k, this amounts to about 15k articles that are better than stubs. Is this good enough? If not, maybe we can start a discussion on establishing some more tangible criteria that give the editors something to strive for! -- Longhairandabeard (talk) 17:11, 31 March 2010 (UTC)

I would like to disagree with your assessment. I did a random test as well, and found that only 6 of the 50 random article had any content at all. Most of the other pages were (a) empty (b) had untranslated English language lists/tables (c) only 1 sentence (d) just a infobox template or a navigation template etc. The Hindi wikipedia may have 54k "pages", but only 2k/3k of those can be anything close to mid-length. From my random test, it seems that at least 30K-35K of the 54K pages are empty/oneliner/template-only pages. --Ragib (talk) 17:26, 31 March 2010 (UTC)


Just to ensure my assessment was not the result of a freak randomness, I checked out another 50 random pages. This time, I got even worse results, with only 5 pages having anything more than 1 sentence. I assume you are not claiming 1-sentence pages as stubs!! In my check, almost 25 of the pages had no content (either empty, template-only, or nav-tempate-only). --Ragib (talk) 17:32, 31 March 2010 (UTC)
I guess the random test is quite random.. I repeated the test twice with 50 articles each: Got 4 non-stub articles the first time, and 10 the second time. The truth, I guess, lies somewhere between those numbers.. Assuming that there are about 15% non-stub articles in Hindi Wikipedia, extrapolated to 7-8k non-stub articles, I repeat my question, is this good enough? -- Longhairandabeard (talk) 20:24, 31 March 2010 (UTC)

Macedonian Wikipedia

The Macedonian edition of Wikipedia now has over 40,000 articles.[1] --Local hero talk 00:57, 17 April 2010 (UTC)

Per #40,000 home page limit, the Macedonian Wikipedia is composed primarily of stubs and placeholders. We omit such Wikipedias. —David Levy 14:18, 18 April 2010 (UTC)
And you still believe in such accusation against Macedonian Wikipedia. This was 9 months ago, and as I can see, stubs which every edition of Wikipedia consists of are noted. Is it the only criteria for omitting? Why are you so sure, that this is the right view for Macedonian Wikipedia? It seems like, you want to overturn us, with only one assertment against Macedonian Wikipedia.--Kiril Simeonovski (talk) 14:37, 18 April 2010 (UTC)
I'm sorry, but I don't entirely understand the above (apparently due to a slight language barrier). I assure you that no one here has anything against the Macedonian Wikipedia or its editors. It simply is composed primarily of stubs and placeholders (which artificially inflate the article count). —David Levy 15:50, 18 April 2010 (UTC)
I'm not sure that you provide it. Noting a warn opinion, which actually seems to be an agitation to the weaknesses of Macedonian Wikipedia (but also weaknesses for many other Wikipedias) convinced me to think that you agreed with it. But, doesn't matter. In fact, I've asked for explanation of your decision, and about minimal requirements of listing random articles to be eligible of adding in the tier. Have you established this requirements for the time since I've noted it?--Kiril Simeonovski (talk) 16:36, 18 April 2010 (UTC)
My apologies, but I'm unable to fully comprehend the above (again because of the language barrier). I think that I understand some of it, but I've already responded on your talk page. —David Levy 16:54, 18 April 2010 (UTC)
Ok. I' ll try to simplify it:
  1. Your first statement above looks familiar with the statement titled #40,000 home page limit.
  2. Which are the requirements to be added in the tier, depending of the outcome of the test of entailing checking random articles?
I don't need an explanation about the first, because the second one is more important. Best regards.--Kiril Simeonovski (talk) 17:10, 18 April 2010 (UTC)
1. I don't understand. I cited that section in the comment.
2. The requirement is that the Wikipedia contain a reasonably low percentage of stubs and placeholders. I plainly stated on your talk page that "we haven't established a specific numerical cutoff" (and I explained why). —David Levy 01:06, 19 April 2010 (UTC)
The accusation that MK Wiki is based only on stubs is false. MK Wiki is one of the few Wikipedias of similar quantity that has featured article every week (most other have every month). That stubs that you can find, are the same stubs that EN Wiki has - we have just translated them. Most of those 40.000 are articles that provide enough information for the reader, you just need to search a bit. The limit is 40.000 and you should add mk wiki.--MacedonianBoy (talk) 17:16, 18 April 2010 (UTC)
Additionally, take a look at the Estonian Wiki and Greek Wiki, try random articles and you will find 5 out of 10 articles that have only one sentence.--MacedonianBoy (talk) 17:18, 18 April 2010 (UTC)
1. There is no "accusation that MK Wiki is based only on stubs."
2. A quantity of 40,000 articles is not an automatic qualifier.
3. The Estonian and Greek Wikipedias contain a significant number of stubs, but they don't contain page after page after page of placeholders with empty sections. Please explain how those articles "provide enough information for the reader." —David Levy 01:06, 19 April 2010 (UTC)

There are clear rules or not?

Before we take the matter further, let's see if we can resolve it here in a simple way.

  • Are there clear rules/gudelines about what is the maximum proportion of stubs allowed for inclusion on en.wiki Main Page? I also see the term 'depth' used. Any predetermined rules on that?
  • Any other criteria?

If there are clear ways to determine this, and if David Levy is right, then I'm afraid that the same will have warrant the removal of (at least) several other Wikipedias of the same group, because even now - long after their inclusion - they will inevitably fail that test. So no double standards please.

Some (longer) time ago, we've been told about the proportion of stubs on mk.wiki, and since then, we've been working very hard to improve the situation. So performing measurements before concluding is a must.

I expect these questions to be answered (i.e. whether there are clear predetermined rules or what), and measurements to be performed before we discuss it any further. And all that in a timely manner, or else I will have to take our issue elsewhere. Regards --B. Jankuloski (talk) 22:23, 18 April 2010 (UTC)

There is no codified numerical cutoff. Thus far, we've dealt with clear-cut cases that wouldn't approach such a hypothetical line. (The "depth" criterion was rendered irrelevant by the widespread use of multiple bots creating and performing non-meaningful edits to stubs and placeholders.)
The Macedonian Wikipedia is in better shape than many, and its proposed inclusion certainly is a valid topic of discussion. My opinion is far from sacrosanct.
I would sincerely appreciate a more civil and less confrontational tone. I'm answering everyone's questions to the best of my ability. —David Levy 01:06, 19 April 2010 (UTC)

My evaluation of Macedonian wikipedia

I did a random sampling of 50 pages from Macedonian wikipedia.

Here is the methodology and definition I used:

  • e = empty pages containing no sentences, only 1 sentence (e.g., saying "X is something at Y") along with an infobox, or empty year pages. example, example
  • s = stub articles containing 1 paragraph or a few paragraphs of text. example
  • m = midsize articles (larger than stubs, but has only a few paragraphs) example
  • L = large full size article example

Here is the result string for 50 pages: eeeessLseeeeseLsemesesseesssmeeseseesesLmeeeeeeese

Counting numbers:

  • e = 28 (56%)
  • s = 16 (32%)
  • m = 3 (6%)
  • L = 3 (6%)

I found that the Macedonian wikipedia has 90 completely blank pages for the years 2011-2100, which are clearly script generated. Most of the "empty" pages had no sentences of a single sentence accompanied by an infobox -- indicating that the page was probably generated by a bot script.

Getting only 3 respectable sized articles among 50 random ones shows a big problem with the wikipedia ... I assume David was referring to this above. Instead of bot-generated empty pages like this, you guys need to focus on the content, and that can hardly be done by bots copying templates from English wikipedia.

Finally, "or else I will have to take our issue elsewhere." -- this threat is quite unfortunate. David's criteria is clear enough, and a quick test confirms the content problem. I can re-do the evaluation with another 50 if you want, but given the large number of template-only blank pages, I doubt it will be any different. --Ragib (talk) 01:18, 19 April 2010 (UTC)

I have no idea why the results are such. They certainly do not reflect reality. In fact, the fellow admins at mk.wiki tell me that if that test were performed on en.wiki, it yields even worse results. No doubt they'll coment about that here soon. If this is the case, it means that we'll have to do some other kind of evaluation.
Regarding my statement. 'Take the matter further' was not meant to be any kind of threat at all, it was a mere statement. David himself rightly said that his opinion isn't sacrosanct, hence he did not react about it. And no, the criteria were not explained beforehand, and they certainly weren't clear enough. He informed me the vague guideline that exist only after I asked myself. And that's perfectly fine, because at least he assumed good faith and answered nicely. --B. Jankuloski (talk) 11:58, 19 April 2010 (UTC)
On 6 July 2009, I explained to Kiril Simeonovski (one of the fellow admins at mk.wiki and commenters here) why the test is not directly applicable to larger Wikipedias. Please see #Tagalog wikipedia.
To reiterate, we test a Wikipedia's percentage of stubs and placeholders as a means of gauging its actual quantity (not percentage) of other articles. Even if 99% of the English Wikipedia's articles were stubs and placeholders (which is not even close to accurate), the remaining 1% would constitute more than six times the number of non-stubs/placeholders contained in the Macedonian Wikipedia (based upon Ragib's figures). —David Levy 13:52, 19 April 2010 (UTC)

My evaluation of English wikipedia

I did a random sampling of 50 pages from English wikipedia. I use same methodologyyy as Ragib used for evaluation of Macedonian wikipedia, plus I added another one category: D = disambiguation page.

Counting numbers:

  • e = 10 (20%) ([2], [3], [4], ...)
  • s = 23 (46%) ([5], [6], ...)
  • m = 5 (10%) ([7], ...)
  • L = 7 (14%)
  • D = 5 (10%)

What I want to say, this method of evaluation is not valid, is stupid, and is intentional. And I want to explain that I'm not for pumping the number of articles, I prefer quality. This approach is a big problem for wikipedia. You guys need to focus yourself on the deeper analyses about the quality of some wikipedia edition and first may be clean your own house.

Of course we can perform same analyses taking random sampling of 50 pages from number of wikipedias with above 50K articles. I expect to see similar results.--Brest (talk) 13:22, 19 April 2010 (UTC)

Please see #Tagalog wikipedia and my above reply to B. Jankuloski.
I don't count disambiguation pages when performing the test. But even if we do, your figures indicate that the English Wikipedia contains 24% mid-size/large articles. That equates to more than 782,000 mid-size/large articles (approximately 19.5 times the total number of articles contained in the Macedonian Wikipedia, including stubs and placeholders). —David Levy 13:52, 19 April 2010 (UTC)
What do you mean?? We talk about proportion here, not numbers. En.wiki is vastly larger than mk.wiki (abour 81 times). How can you compare sheer numbers like that?? --B. Jankuloski (talk) 14:03, 19 April 2010 (UTC)
There is no reasonable expectation of the Macedonian Wikipedia containing anywhere near as many articles (good or otherwise) as the English Wikipedia. (I'm not the one who introduced the comparison.)
But as explained above, calculating the percentage of stubs/placeholders is merely a means of gauging the actual quantity of non-stubs/placeholders. That's why the test is not directly applicable to larger Wikipedias (which can contain a similar percentage of stubs/placeholders and still offer vastly more good articles — the measure that actually affects readers). —David Levy 14:37, 19 April 2010 (UTC)
Ok, I've calculated that, with Ragib's findings of 12% of articles on mk.wiki being non-stubs and non-placeholders and given that en.wiki is 81 times larger as a whole, mk.wiki's stiation would be 405,000 such articles in en.wiki terms. True. En.wiki has got 782,000. But is the difference so extremely large that it warrants our exclusion? Besides, you have to be relative in treating other smaller wikipedias for inclusion on interwiki, especially in the lowest stratum of listed (so-called '40,00 and over')! I am starting to find your opposition sincerely strange. Moreover, I've been informed that certain other wikipedias that are listed here are in comparably even worse-off position than mk.wiki by your criteria, but no one raises objections and they remain here. Not fair, isn't it? --B. Jankuloski (talk) 14:28, 19 April 2010 (UTC)
Again, we calculate the percentage of stubs/placeholders as a means of gauging the actual quantity of non-stubs/placeholders. That's why the test is not applicable to larger Wikipedias on the same terms.
In real life, readers don't care about the percentage of stubs/placeholders among 50 random articles. They simply want to be able to type (or follow a link to) a term and arrive at a good article. A large Wikipedia containing x percentage of stubs/placeholders offers vastly more good articles than a small Wikipedia containing x percentage of stubs/placeholders does. —David Levy 14:37, 19 April 2010 (UTC)
What kind of science is statistics? Exact sum of wrong numbers. Quite clear is that 24% of 3M and more articles of en.wiki is 19.5 times more then of total number of articles from mk.wiki (stubs and placeholders included). But I don't discuss about the size of en.wiki vs the size of mk.wiki, I discus about the used methodology as a tool to accept or refuse to include mk.wiki (and other) at the main page of en.wiki. And I think that you will agree with me that that methodology (proposed and described by Ragib) is not appropriate to measure the quality of the particular wikipedia. If you do not like to include at the main page, the wikipedias which has more then 40K, you can simple just rise the quota, let's say at 50K, it's better approach then tell us the dumb statistic story. Or you should work on better methodology. Or you can simple ignore us, and rise your nose.--Brest (talk) 14:20, 19 April 2010 (UTC)
I'm answering everyone's questions to the best of my abilities. It's frustrating and disheartening that you perceive this response as "rising up my nose." —David Levy 14:37, 19 April 2010 (UTC)
I just note as an option, different ways how to escape from this situation, depends on you, which one you will perceive as your option. The next possible option is to include mk.wiki, because per state at main page, "some of the largest are listed below", and the different wikipedia language editions are grouped by the NUMBER OF ARTICLES, more then 40K, 100K, 150K, 500K. Seems very logical to me to accept this option.--Brest (talk) 15:04, 19 April 2010 (UTC)
We state that "some of the largest are listed below" (emphasis added) specifically because we don't include every Wikipedia containing more than 40,000 articles.
Using the total number of articles as the sole criterion is highly illogical; it assigns the same value to empty pages as it does to good articles (resulting in the inclusion of Wikipedias with fewer useful articles than some that are omitted). It also encourages the creation of stubs and placeholders (usually via automated scripts) purely for the sake of appearing on our list. (This is not hypothetical; it became a common occurrence before we instituted qualitative criteria.) —David Levy 15:22, 19 April 2010 (UTC)
Yes, "some of", means not all above the limit, but I can't see anywhere on the main page related to this stuffs, that you state that you can personally (or in the groupe of two) judge, as I realized according to above mentioned methodology which we find out that is not quite appropriate, which one will be included or excluded from the grouped lists (grouped by total number of articles and alphabetically sorted).--Brest (talk) 15:41, 19 April 2010 (UTC)
I'm sorry, but I don't understand the above. —David Levy 15:43, 19 April 2010 (UTC)
Sorry about my bad English, if you try in Macedonian may be it will be easier for me :). Yes, "some of", means not all above the limit. I can't see anywhere on the main page any word related to this discussion, that say that there is other criteria which should be fulfilled by wikipedias to be included or excluded from the lists. As I realized you use the above mentioned methodology (analyzing the sample of 50 random articles), which we find out that is not quite appropriate for quality evaluation. But the quality is not stated as criteria which wikipedia will be included or excluded from the lists (at the bottom of the main page, grouped by total number of articles and alphabetically sorted).--Brest (talk) 15:57, 19 April 2010 (UTC)
I don't speak any Macedonian, so you're far ahead of me.  :)
The template's documentation (transcluded on the template page itself) contains the statement "This is not a complete list of Wikipedias containing 40,000 or more articles; Wikipedias determined to consist primarily of stubs and placeholders are omitted." This is an editor-focused message that isn't appropriate for the reader-focused main page. —David Levy 16:08, 19 April 2010 (UTC)
It's clear that are not listed all wikipedias containing 40K or more articles. We have an dispute about: are mk.wiki consist of "primarily" stubs and placeholders or not? I don't deny that mk.wiki has a stubs and placeholders, no one deny that. I want to change your mind that mk.wiki "primarily" consist of stubs. One of the proofs is that your methodology doesn't work, try to implement it on other wikipedias with above 40K number of articles (which already pass your quality criteria) and you will take very close results to mk.wiki. Next, the editors at mk.wiki work very hard to improve the quality of their wiki day by day, hour by hour. And they want to see mk.wiki visible here. Look at translate.wiki statistics about Macedonian language, 100% of messages are translated, 0% obsoleted, it is another proof that Macedonian editors work hard. Look at commons picture of the day, where you can see that besides biggest languages, all the picture descriptions are translated in Macedonian, there, I can't see most of the wikipedias listed in your now disputed template. So the editors of Macedonian wikipedia work very hard to improve the quality of their wiki, their primary goal is not to pump the numbers, and they want to see mk.wiki visible here.--Brest (talk) 16:44, 19 April 2010 (UTC)
You are getting a bit emotional here. No one is commenting on the sincerity of the Macedonian wikipedians. However, tests show clearly that almost 88% of the (non disambig) pages on Macedonian wikipedia are either empty, or one liners, or stubs. More than half of the pages are actually empty pages (no text, or one liners with just an infobox). While there are about 12% regular sized articles, this translates to about 4800 usable pages on the Mk wikipedia. David was making the point that the larger wikipedias have a large number of usable pages, so in a 3 million page wikipedia, even a 10% good sized article means 300,000 such articles will be in good shape. In case of Macedonian (as well as some others), the number and percentage of blank pages is simply too large. Anyway, I think we are walking in circles here ... reiterating the same point over and over. --Ragib (talk) 17:20, 19 April 2010 (UTC)
I think that we are not in the circle, but we are close to the solution. I don't agree with your numbers given above, because per privious discusions, your methodology is not reliable. You don't like to compare statistics (taken using your own methodology) with the larger wikipedias, like English one, may be for emotional reasons :). OK, I will agree with you, but let's to compare it with near mk wikipedias in size (let's say between 40K and 100K), which pass your quality criteria. Can you do comparative analyze please? --Brest (talk) 17:36, 19 April 2010 (UTC)
You mean, my *numbers* are not right? Please feel free to re-do the random test, and show a statistically significant different result. I am not a native English speaker, and my main work is in the Bengali wikipedia, which, if you look above in this page, suffers from the same problem of too many stub/empty pages. I had once argued for its inclusion ... but our community there has now focused on improving the number of full articles. Anyway, back to the issue, I think it is a reasonable request to David that some of the currently listed wikipedias in the 40k tier should be checked, and delisted if necessary. --Ragib (talk) 17:42, 19 April 2010 (UTC)
To be precise, I mean your numbers are less or more correct but your conclusion taken from that numbers is false, because it is taken from unrepresentative sample, just 50 random articles. If you like to talk with statistic language, for me more reliable statistic is presented on this [list of wikipedias]. And please mention the "depth" indicator, Edits/Articles × Non-Articles/Articles × Stub-ratio, which is more reliable quality indicator because it runs over all pages at some wiki (all article population in main space and other pages, calculated in the correlation with the number of edits as indicator of user activity). In this moment I don't like to delist any one, I just like mk.wiki to be listed.--Brest (talk) 18:23, 19 April 2010 (UTC)
I re-did the test not just over 50 articles at random, but over around 200 articles. Since you mention statistics, at 95% confidence level and a 7% confidence interval, 195 is the sample size you need for 40k pages. Finally, depth has exactly zero value in measuring quality. Let me show how to manipulate depth: you just create a bot that makes cosmetic edits to articles, and run it over all articles. That will increase "Edits" and thereby increase depth. Another trick is to write another bot to create talk pages with talkheader templates, and run that for all articles. By manipulating these, you can easily achieve a 100+ depth for a wikipedia that still has few real articles. So, don't wave "depth" as an useful measure of any sort. --Ragib (talk) 20:22, 19 April 2010 (UTC)
Over and over again, I've explained why the test is not directly applicable to large Wikipedias (because we care about the actual number of useful articles, not the proportion). Over and over again, you've ignored this explanation, and you now you suggest that it "may be for emotional reasons" (as though we have some sort of emotional bias against the Macedonian Wikipedia?).
You're welcome to continue ignoring our explanations and attributing the situation to whatever imaginary conspiracies you want, but it isn't helpful. —David Levy 19:43, 19 April 2010 (UTC)
Over and over again you stay at your position without logical reason. And the "emotional reasons" was the phrase used by your supporter (Ragib), I just replay to using his words. Sorry but I don't think that we can continue further civil discusion here if you choose option to "rise your nose". And I don't think that I can explain to you any thing if you continue to put your head like an "ostriche in sand". And probably is better to stop here, I will leave you to ask your mirror which wikipedia is the best.--Brest (talk) 20:05, 19 April 2010 (UTC)
No, Ragib did not use the phrase "emotional reasons." Ragib commented that your tone seemed "emotional" — a statement unrelated to our rationale for omitting the Macedonian Wikipedia.
I'm sorry that you don't see the logic behind my position. I've tried very hard to explain it you, and I'm quite taken aback by your extraordinary rudeness. —David Levy 20:24, 19 April 2010 (UTC)

Either we are getting lost in translation, or the discussion above is seeing very unfortunate choice of words from Brest and others. I urge the Macedonian wikipedian community to remain civil in the discussion. :( --Ragib (talk) 20:27, 19 April 2010 (UTC)

I don't see how anyone here is being rude, whether that be Brest, myself, or any other mk.wiki user. As admins and as users, we are all very polite and take our duties seriously. What is happening is argumentation and, yes, frustration over your policies that you defend so vehemently, in spite of the fact that there might be another solution. I am not saying that anyone here is lying or deliberately discriminating. The people are trying to show you why mk.wiki should be included in spite of your present guidelines (i.e. that they aren't the right guidelines in this case), and they rightly get upset when you still oppose it on the same old grounds, in spite of there being no fixed criteria (i.e. you have room for moving beyond them, but you're unwilling to do it). That causes the frustration and I see it as justified. Besides, there are other wikipedias on the list that would, according to the same criteria, be also unstatisfactory and should no be included, but they stay there (and I've no intention of kicking anyone out, just including mk.wiki). Let's face it - you are unprepared to deal with the issue on case-by-case basis. You said it yourselves above that you had no such case before (only clear-cut ones). That's not fair. Since you did not, how can you judge according to a non-existent precedent? The case is peculiar, and requires a 'peculiar' solution. --B. Jankuloski (talk) 04:28, 20 April 2010 (UTC)
If you regard Brest's above message as "very polite," you must have radically different civility standards at the Macedonian Wikipedia than we have here. —David Levy 04:52, 20 April 2010 (UTC)
No, trust me, they are no different. We've been faithfully following the same guidelines that are written here on en.wiki and other general projects. --B. Jankuloski (talk) 06:30, 20 April 2010 (UTC)
Brest stated that our "method of evaluation is not valid, is stupid, and is intentional." Brest then accused me of turning up my nose and burying my head in the sand, noting that he/she would "leave [me] to ask [my] mirror which wikipedia is the best."
You regard these remarks as "very polite"? —David Levy 11:18, 20 April 2010 (UTC)
Quite so. The remarks aren't polite at all. What I am saying is that these people are very good admins that have shown a great deal of wiki-professionalism throughout their adminship. This is the very first time that I see them talk in this manner. --B. Jankuloski (talk) 11:49, 20 April 2010 (UTC)
My impression is that you will be satisfied with no "solution" other than the the Macedonian Wikipedia's immediate inclusion. Rather than merely disagreeing with our rationale, you dismiss it as patently illogical and express astonishment at its application. —David Levy 04:52, 20 April 2010 (UTC)
There is nothing qualitatively bad about dismissal. It's simply a higher level of disagreement. --B. Jankuloski (talk) 06:30, 20 April 2010 (UTC)
One of the key elements of our civility policy is to "not ignore the positions and conclusions of others."
Should I dismiss your complaints? Is that what you believe I've done? —David Levy 11:18, 20 April 2010 (UTC)
No, neither have you dismissed them, nor should you do so, and you certainly have my respect for that. I do apologise if it sounded like that. If you see my comments as a whole, you'll see that I am geenrally arguing for a change of attitude/criteria of inclusion. Since I believe that they aren't right for us, I can safely say that they make no sense in our case, i.e. untenable in our case. Perhaps my mistake lies in my not stressing the 'in this case' bit. Otherwise, for all other cases, I don't feel like I can tell you what to do, I don't have the experience and expertise in that - therefore I am not generally saying that your opinions are worth nothing. They might be worth a treasure, and I don't challenge them per se, I challenge their application in our case. --B. Jankuloski (talk) 11:49, 20 April 2010 (UTC)
In my opinion, we are correctly applying our inclusion standards to a Wikipedia lacking a sufficient quantity of useful articles. While there certainly are worse Wikipedias, the Macedonian Wikipedia does not rise to a level that I perceive as borderline, nor does this situation strike me as "peculiar" (apart from your organized campaign).
There is nothing 'organized' going on. Other people have been informed of the issue and it's only natural that they should follow it and express opinions about it at their own accord. --B. Jankuloski (talk) 06:30, 20 April 2010 (UTC)
Statements along the lines of "...I've been informed that..." and "...the fellow admins at mk.wiki tell me..." reflect external discussion. That's fine, and I merely note that it's the standout circumstance. —David Levy 11:18, 20 April 2010 (UTC)
Yes. You are right. It's not disagreeable, but I have to find my info somewhere. That's not to say that we've been conspiring and thinking up methods of assaulting anyone. --B. Jankuloski (talk) 11:49, 20 April 2010 (UTC)
As previously stated, your disagreement is welcome. But I resent your insistence that my position is indefensible. —David Levy 04:52, 20 April 2010 (UTC)
I see no cause for resentment in what is simply a higher level of disagreement. Resentment is, after all, a very strong word (and emotion). I certainly don't hold any such attitudes against you. We are simply arguing a point. --B. Jankuloski (talk) 06:30, 20 April 2010 (UTC)
1. You (by which I refer to the group) are not merely arguing that the link omission is incorrect or ill-advised. You're arguing that we haven't even cited a logical reason for it. I have no objection to the expression of disagreement with our rationale (and the belief that we should do things differently), but it's extremely frustrating to continually provide explanations and repeatedly be told that I haven't.
2. Kiril Simeonovski's first reply to the discussion began with "And you still believe in such accusation against Macedonian Wikipedia." Brest stated that our "method of evaluation is not valid, is stupid, and is intentional." Brest then accused me of turning up my nose and burying my head in the sand, noting that he/she would "leave [me] to ask [my] mirror which wikipedia is the best." These comments appear indicative of resentment. —David Levy 11:18, 20 April 2010 (UTC)
You are right, they have not been polite to you and I agree. I was refering merely to myself and my argumentation, that it does not contain resentment and hence, does not deserve any. --B. Jankuloski (talk) 11:49, 20 April 2010 (UTC)
You ignored my question yet again. Why are other wikipedias who are in the same (or similar) position as mk.wiki still standing listed? If you ask me, they should indeed be still standing, but according to your criteria there are several that should not, on the same grounds as mk.wiki's non-inclusion. Either this is very much the case, or I've been badly misinformed, which is unlikely. Why shouldn't we feel discriminated and discontent if we are suffering such double standards? --B. Jankuloski (talk) 06:39, 20 April 2010 (UTC)
Please tell me which included Wikipedias you regard as the same or similar, and I shall evaluate them. Until then, I cannot address your assertion. —David Levy 11:18, 20 April 2010 (UTC)
I will ask and will inform you. But I repeat that it is not my intention to kick anyone out. I don't think that they should be removed, on the same grounds as I think that we should be included. --B. Jankuloski (talk) 11:49, 20 April 2010 (UTC)

Questions

Dear all, I'm forced to say that this discussion leads nowhere, and we can keep up this endlessness discussion, which won't give us (admins at Macedonian Wikipedia) the right answer. If you're affronted of our relation during the whole discussion, I offer an apologize to those who got affronted. I appreciate giving advices like Ragib's "Instead of bot-generated empty pages like this, you guys need to focus on the content, and that can hardly be done by bots copying templates from English wikipedia." and I am disappointed, why user David Levy did not give us such advice. My point in this long discuss is not to test your ability to hold this template, or your effectiveness of maintaining it, but to get informations about the template (Don't worry! I'm not trying to pray for adding the link in the tier.). Watching the discussion, I've shortened it in few questions:

Q1 Does this template exist, just to add links in the tier, or to remove from the tier?
A1...
Q2 When raising the border to 40,000 articles, and removing the only criteria of Depth value 5, did you perform the test of random checking articles on all language editions of Wikipedia over 40,000 articles, or you've just removed the language editions with Depth less than 5?
A2...
Q3 If we expand the placeholders with reasonable, and useful content, how to get sure that the link will be added in the tier? Is there, just a small criteria that would incline you to add it?
A3...
Q4 What we numerically have to enlarge? (Of course to be mentioned in the tier. Ragib's test looks pretty good for it. Can you just change the numbers, and note them as criteria? This is also helpful in establishing criterias, which apparently would not lead to endlessness discussions like the given above.)
A4...

I hope that the holders of the template will assume good faith to answer the questions in the given order. Best regards.--Kiril Simeonovski (talk) 21:14, 19 April 2010 (UTC)

P.S. I would be more satisfied if you give an honest answer to Q2, and will not avoid direct answer on Q3. Best regards.--Kiril Simeonovski (talk) 21:17, 19 April 2010 (UTC)

I've plainly stated (numerous times) that the problem is that too few of the Macedonian Wikipedia's articles are not stubs/placeholders, and I've sincerely striven to answer every question asked (even as I've been accused of arrogance and ignorance), so I don't understand why you believe that I haven't attempted to provide helpful advice.
However, I accept your apology and will continue to do my best to answer your questions.
A1: I don't understand this question. The template exists to link to the Wikipedias of most value to readers.
A2: The "depth" criterion has been abandoned. At the time of the "more than 20,000 articles" tier's removal, the problem appeared to exist almost exclusively with the Wikipedias in that range (to which the Macedonian Wikipedia belonged until now), and I personally tested all of them. A small number (two, I believe) borderline Wikipedias were removed from the list because there simply were not enough of them to sustain the tier.
A3: You seem to seek a bright line that simply does not exist. If the Macedonian Wikipedia's quality continues to improve (which should occur for reasons entirely unrelated to this template), it eventually will reach a point at which there is consensus to list it. I'm sorry, but there is no magic numerical threshold to cross.
A4: Again, I don't understand the question. I'm sorry. —David Levy 22:32, 19 April 2010 (UTC)
The main point in Q4 is more comprehensive answer. It seems better when we have somethning like "e = 28 (56%) s = 16 (32%) m = 3 (6%) L = 3 (6%)" as a criteria. This was noted to confirm the accuracy of the statement that the articles of Macedonian Wikipedia primarly comprise stubs, and placeholders. Could you please give enlarged numbers which are required (i.e. e=15 s=15 m=10 L=10). Regards.--Kiril Simeonovski (talk) 22:59, 19 April 2010 (UTC)
As I have explained repeatedly, we have not codified any formal numerical thresholds of that nature. —David Levy 23:40, 19 April 2010 (UTC)

An evaluation of the Bulgarian Wikipedia

I took the trouble to do a random sampling of 50 pages of the Bulgarian Wikipedia, by analogy with what Ragib did about the Wikipedia discussed exuberantly above, and I used the same methodology and definitions as Ragib did. Here are the results. You can view a translation of the title of each article, and sometimes a description or opinion about it, by pausing the mouse pointer on any of the letters below, and you can view the articles themselves by following the links.

msmsemssDemsssssmsmsssesssmeLsseeDsmessssssemseesm

Counting numbers:

  • s = 27 (54%)
  • e = 10 (20%)
  • m = 10 (20%)
  • D = 2 (4%)
  • L = 1 (2%)

I only got one respectably sized article out of 50 randomly selected. The number of the completely or nearly completely blank pages shall be vastly larger than those 90 of the Wikipedia discussed above, because the Bulgarian Wikipedia has an article on every year since 1 AD, including such ones as 2, 18, 34, 1036, or 1306, though some of those have been given at best a couple of events to boast with, e.g. 55, 1065, or 1456. There are also many blank pages and stubs about villages in Bulgaria, FYR Macedonia and Northern Greece, or towns in Scandinavia, Germany, France and Italy, and there are also similar articles about football clubs from the UK and other countries, e.g. this one, this one, or this one. And there is this one, which was once proposed for deletion, but was voted to be kept.

Considering the above calculations and contentions, may I take the liberty of suggesting the removal of the link to the Bulgarian Wikipedia Main Page from this template, and, accordingly, the Main Page interwiki link to it? --Магьосник (talk) 22:48, 25 April 2010 (UTC)

In my opinion, the Bulgarian Wikipedia contains a sufficient number of medium-sized articles and reasonable stubs (i.e. not empty placeholders) to warrant its inclusion.
However, I base this assessment on the generous standards applied to date. The possibility of applying stricter standards is worth discussing. —David Levy 23:14, 25 April 2010 (UTC)

I've mentioned this on the main page's talk with a few supports, so I'll repeat it here. I think that it would be a lot more intuitive for non-English speakers looking for their native language if we put the link to the full list as a fifth bullet at the bottom of this template rather than keeping it where it currently is in the middle of an English sentence. —Arctic Gnome (talkcontribs) 20:03, 29 April 2010 (UTC)

This thread on its own is unlikely to generate much discussion, so I've gone ahead and reverted to the previous format (a centered link at the bottom). I relocated the link to the top in 2008 (upon removing the obsolete link with which it was paired), so this is a self-reversion on my part. Because it now stands alone, I've added "of Wikipedias" to the displayed text.
If anyone objects to the change, further discussion is welcome. —David Levy 14:55, 30 April 2010 (UTC)

bg at 100k

The Bulgarian Wikipedia has reached 100,000 articles. Please update this template. Thank you. --Yair rand (talk) 07:38, 24 May 2010 (UTC)

Done. —David Levy 13:22, 24 May 2010 (UTC)

Formatting/remove categorization

{{editprotected}} In order to remove the categorization of the Main Page, could you replace this template's code with the one from sandbox (probably good to protect {{Wikipedialang/core}} before this happens). This template does the following:

  • Adding "lang" and "xml:lang" tags to the link text, which makes this line better readable for browsers and machines to recognize that this is not an English text.
  • Eases formatting when resorting this template or adding new languages.
  • By using {{#language:}}, we make sure to always use the native language name that's used by MediaWiki, for example within the interwikis.

All templates used within the sandbox are already protected, except Wikipedialang/core. Thanks, --The Evil IP address (talk) 11:00, 3 June 2010 (UTC)

I tried it but had to revert myself because it didn't work. can you check the diff and see what needs changing? HJ Mitchell | Penny for your thoughts? 16:37, 3 June 2010 (UTC)
It worked (and still works) for me. Do you still know what exactly has been the problem? --The Evil IP address (talk) 17:18, 3 June 2010 (UTC)
Got it! HJ Mitchell | Penny for your thoughts? 20:33, 3 June 2010 (UTC)

Vicipædia

The Latin language edition of Wikipedia is going to have its 40,000th article some time in the upcoming days. I have a presentiment that this milestone could result in lengthy discussions whether it should or should not get added to this list. Some will insist that Latin is a language with a remarkable history that has left a remarkable legacy for us, and has been used internationally as a scholarship language for centuries, and therefore the Wikipedia in that language should not be judged with the same criteria as Wikipedias in languages of some small countries are, and has to be included in the list irrespective of the length and quality of its articles. My personal humble opinion is quite different, but I shall express it later for I'm not sure it would add any weight to the discussion right now. I'm interested in the following: is the Latin Wikipedia originally intended to be included in the template on its passing the 40,000 threshold? Thanks in advance. --Магьосник (talk) 03:49, 5 June 2010 (UTC)

I should think so, provided that those 40,000 aren't predominantly made up of placeholders or stubs or other non-articles (like redirects a disambiguations), the grounds on which I believe several other editions of WP aren't on the template. HJ Mitchell | Penny for your thoughts? 03:55, 5 June 2010 (UTC)
I would really love to see the link to the Latin Wikipedia among the others on this template and on the Main Page. But I'm not quite convinced it is worth adding. I do not want to offend anyone working on la.wp - I have been making observations of their work for a few years now, and I am impressed favourably by it. Also, I don't insist on my present opinion very much as I'm too little experienced with editing en.wikipedia, and I even less wish to seek a quarrel with anyone. I'm simply offering that as an alternative viewpoint.
Latin is currently not, and is gigantically unlikely to ever in the future become, anyone's mother tongue. So every contributor to the Latin Wikipedia has to have studied Latin as a foreign language. That would often affect the level of correctness and intelligibility of the articles there. The articles themselves have been divided into several categories in accordance with a system of evaluating the level of their Latinity. As of now, the Latinity of 482 articles needs to be improved, 376 articles have doubtful Latinity, 251 are very doubtful, 116 are poor, 173 are very poor, and 1,993 have not yet been verified.
I recently talked to a professor of Latin at the Sofia University. Our discussion went into the topic about usage of Latin in modern contexts. She expressed her view that she was understanding that as disparaging Roman history. I challenged her to guess what the Wikipedia in the Latin language had called with the name pediludium - a word invented from the words for "foot" and "game" - and her instant suggestion was "some kind of dance". It's actually football. Wikipedia in Latin contains a considerable number of articles related to contemporary popular culture, music, cinema, sports, as well as computing, science, economics, as well as placenames, personalities and history of Eastern Asia, Africa and the Americas - all those topics require a large amount of invented terms, which would be nowhere near intelligible for those who used to speak Latin natively many centuries ago.
Actually, it's the other way round: the editors of the Latin wiki take pains to use attested terms, and the local policy is quite specifically not to "invent terms." Pediludium is in use in Latin-writing communities worldwide; it's not particular to the Latin wiki. Often when someone tries to invent a term in the Latin wiki, other users will pounce on it and demand an attestation. Jacob (talk) 14:19, 12 June 2010 (UTC)
And there could rise the question "Cui bono?" (which could equally apply for the Esperanto Wikipedia). Anyone would prefer to read an encyclopædia written in their native tongue (or in English) rather than in Latin. And still there are a lot of stubs.
I of course will not bother in any way if la.wikipedia.org gets included in the list. But what would you say about the above? --Магьосник (talk) 04:24, 9 June 2010 (UTC)
"Some kind of dance" seems pretty close to me :)
I wouldn't blame a professor of Latin for being unfamiliar with such terms, it all depends whether said professor gets involved with "living Latin" or not. Many don't. Just for the record, Vicipaedia doesn't invent terms (that is ruled out as "original research"). It takes terms already used by writers of modern Latin, and if none are available, it borrows terms in international use; this works in Latin just as it does in other languages. Andrew Dalby 11:22, 12 June 2010 (UTC)
Very true Andrew. We should also consider the fact that most professors associated with Latin and Greek are in fact classicists and not Latinists. This means that their opinion should be trusted and valued for matters concerning literary theory, intertextuality, history, etc. The overwhelming majority of these however don't actively work with the language at all.
As for this professor's fear that using Latin to talk about modern concepts is "disparaging Roman history", surely she must be aware that Latin was a native tongue for many people for a good 1000 years after the fall of the Roman empire. I would bet there were a good number of them who weren't even aware that there was a Roman history to disparage. No, they were just using a natural language to discuss natural topics.
None of this is meant to defend the unfortunately high amount of crap that's on the Latin wikipedia; we have tons of it. The notion/vision of Vicipaedia however is impeccable.
--Ioscius 08:38, 13 June 2010 (UTC)
That's kind of outside the scope of this talk page, which is for discussion of the associated template, not the validity of the latin Wikipedia. If it gets to 10,000 valid articles, it will probably get a spot on this template, because deciding on the worthiness of having a Wikipedia in a certain language is not what we do. HJ Mitchell | Penny for your thoughts? 04:33, 9 June 2010 (UTC)
You must mean 40,000 articles. I was thinking of the template rather than of the validity of other Wikipedias, but let's really abandon this discussion. Cheers! :) --Магьосник (talk) 06:52, 9 June 2010 (UTC)

Four Wikipedias with 40,000+ articles

{{editprotected}}

The editions of Wikipedia in all Latin, Georgian, Cebuano and Aromanian have recently got their 40,000th article. Links to them need to be supplied. --Магьосник (talk) 18:05, 10 June 2010 (UTC)

Having 40,000 entries is not the *only* criteria for inclusion. Please see discussion above. --Ragib (talk) 20:14, 10 June 2010 (UTC)
If other criteria are in play, then it would be helpful for them to appear on the page. To most readers, "More than 40,000 articles" will surely mean "More than 40,000 articles" (and nothing more). 71.191.114.242 (talk) 11:46, 12 June 2010 (UTC)
The template (and therefore the main page) states: "Many other Wikipedias are available; some of the largest are listed below." [emphasis added]. It's true that the specific criteria are not noted there (though they are on the template's documentation page), but I know of no way of doing so in a manner that is both concise and relevant to most readers.
And we certainly don't want a lengthy, distracting disclaimer on the main page, especially for the sake of conveying a distinction of little consequence to readers (for whom a Wikipedia composed of 1,000 decent articles and 39,000 stubs/placeholders is not meaningfully different from a Wikipedia containing 1,000 decent articles alone). It's only because of an arbitrary, editor-centric technicality (MediaWiki's "non-redirect in a content namespace with a minimum of one internal link" standard) that this is even counted as an "article" in any context. —David Levy 12:53, 12 June 2010 (UTC)
In the Latin wiki, it would be marked "not [even] a stub" (non stipula) and deleted seven days thereafter. The Latin wiki has developed criteria for achieving stubness, and it marks many of its articles according to the quality of their grammar & style (1 is good, −6 is awful, and −7 is so bad the article, even if it's more than a stub, may be deleted in seven days). ¶ The quality of most wikis is already being assessed according to their handling of The 1000 articles every wikipedia should have. By that assessment, the Latin wiki ranks 46th among the wikis, immediately below Simple English, four places above Euskara, and six places above Eesti, which your methods include among your "some." Jacob (talk) 13:53, 12 June 2010 (UTC)
Few would contest the claim that the Latin Wikipedia is in significantly better shape than the other three Wikipedias discussed in this section are.
If our criteria were based upon the lengths of articles on those 1,000 topics, perhaps the Latin Wikipedia would be included (though we currently list only 41 Wikipedias.) —David Levy 14:29, 12 June 2010 (UTC)
Why not raise the cutoff, say, to 50K, and lessen the impact of the POV-ness inherent in any attempt to estimate quality? Jacob (talk) 14:34, 12 June 2010 (UTC)
We occasionally adjust the numerical thresholds, and this is a reasonable idea to consider. However, it wouldn't eliminate the need to perform qualitative assessments. (We already omit Wikipedias comprising more than 50,000 articles, and others composed primarily of stubs and placeholders will continue to reach that quantity.) —David Levy 19:05, 12 June 2010 (UTC)
I'm going to disable the template for now, because I don't know how to determine if these editions of WP "qualify" for inclusion on the template. I believe this page is watched by several admins, including myself, but if nobody else comments in the next 24-48 hours, raise it on T:MP, which attracts a lot more attention. HJ Mitchell | Penny for your thoughts? 22:14, 10 June 2010 (UTC)
I just visited the four Wikipedias and performed our standard check of 50 random articles (with disambiguation pages excluded from consideration). The results were as follows:
Latin Wikipedia
  • 48 stubs
  • 1 article of moderate length
  • 1 short article
Georgian Wikipedia
  • 46 stubs/placeholders (mostly placeholders)
  • 4 short articles
Cebuano Wikipedia
  • 50 stubs (including 48 bot-imported, two-sentence descriptions of French communes)
Aromanian Wikipedia
  • 50 stubs/placeholders (mostly placeholders, with all stubs consisting of one sentence each)
None of these Wikipedias meet our inclusion criteria. —David Levy 00:32, 11 June 2010 (UTC)
I just visited the Latin and got one that would probably rate as start-class on our scale and 9 stubs consisting of one or two sentences. I'll check the other as well, mostly for my own curiosity. HJ Mitchell | Penny for your thoughts? 00:40, 11 June 2010 (UTC)
I checked all the above projects and I concur that none of them meet the criteria for inclusion on the template. HJ Mitchell | Penny for your thoughts? 01:24, 11 June 2010 (UTC)
The results of visiting each project and clicking "random article" were:
  • La:
    • 0 placeholders
    • 36 stubs
    • 11 that would rate as "start" class on our scale
    • 1 that looked complete and useful
  • Ka:
    • I had to click "random article" 4 extra times because it returned the same article three times and a different article twice
    • 8 placeholders (containing no prose whatsoever)
    • 20 stubs
    • 18 starts
    • 3 of moderate length
    • 1 that looked complete
  • Ceb
    • 0 placeholders (no prose)
    • 49 stubs, of which 45 were bot-translated, 2 sentence articles on French communes
    • 1 "start"
  • Roa-rup:
    • 45 placeholders (no prose at all)
    • 4 stubs
    • 1 start
--HJ Mitchell | Penny for your thoughts? 01:24, 11 June 2010 (UTC)
While I wouldn't want to disparage the work being done on those other three contenders, you people might want to glance again at Latin from time to time. As you have already noted, it has a different pattern from some others: no placeholders; many of the pages you would count as stubs have useful external links and bibliographies. Andrew Dalby 09:15, 12 June 2010 (UTC)
And no one wishes to disparage the work being done there. But it doesn't meet our criteria for inclusion here. —David Levy 12:53, 12 June 2010 (UTC)
Of course, nobody wants to disparage any of the people working to build any of those 4 or any other edition of Wikipedia, but unfortunately they don't yet have sufficient good quality articles to be included here, because space is somewhat limited. Maybe they will in a few months, and then they can be added. HJ Mitchell | Penny for your thoughts? 14:56, 12 June 2010 (UTC)


Persian/Farsi Wikipedia now 100,000

Farsi wikipedia has reached 100000 articles. Please update the template. Dreamfall (talk) 06:28, 25 August 2010 (UTC)   DoneJames (T C) 06:38, 25 August 2010 (UTC)

Edit request from Lintu, 16 September 2010

{{edit protected}}

Hi there,

seems like the Romanian wikipedia is finally over 150.000 articles. Would like to move it to the right category.

Thanks, Andrei.

Lintu (talk) 14:16, 16 September 2010 (UTC)

  Done — Martin (MSGJ · talk) 16:12, 16 September 2010 (UTC)

Move request

{{movereq|Template:Wikipedia languages}}

Template:WikipedialangTemplate:Wikipedia languages — Should be changed for better readability. This template is rarely used, speaking of quantity, but used on the Main Page, so I think that it requires a discusssion. The template was created in 2006, when there have been thousands of different writing styles in template names. One used CamelCase, the other one wrote all in one word and so on, making things harder to remember. Nowadays, pretty much all uses the spelling how it's proper English, just like in the article titles. In this case, I'd propose a rename to "Wikipedia languages": This is the section name how it's used on the Main Page, thus making it easy to identify. I generally also feel that these templates of that kind on the Main Page should best always reflect the corresponding section name, so should it ever be changed, this name should so, too. Of course, I could understand possible doubts about it being a Main Page template, but that shouldn't be any problem. We'd just have to move the template, just as we move any article. The created redirect will automatically be protected as long as this template is protected, so no space for a vandal to harm Wikipedia. We might then correct the redirect on the Main Page, but that wouldn't be necessary, since redirects are cheap. --The Evil IP address (talk) 16:07, 22 September 2010 (UTC)

Given the lack of complaints about this, I wouldn't expect these similar changes to be controversial. As noted above, there really isn't any harm at all, so I'll go ahead and move the templates. —David Levy 17:26, 22 September 2010 (UTC)

Update Values

{{editprotected}}

Could an administrator copy over the code at Template:Wikipedia languages/sandbox to Template:Wikipedia languages? Since both the French and German Wikipedias now have more than one million articles, as seen here. Thanks. Set Sail For The Seven Seas 273° 31' 0" NET 18:14, 25 September 2010 (UTC)

  Done —WWoods (talk) 16:30, 26 September 2010 (UTC)
I've reverted. The addition of such a top-level tier (containing one or few Wikipedias) has been proposed on numerous occasions and consistently rejected by the community. (This includes more than one recent proposal for a "more than 1,000,000 articles" tier at Talk:Main Page.)
We didn't reach consensus to add the "more than 500,000 articles" tier until eight Wikipedias qualified. —David Levy 16:57, 26 September 2010 (UTC)

Proposal

See Talk:Main Page#Proposal: Add million-article level to Wikipedia Languages section

The Transhumanist 03:38, 12 October 2010 (UTC)

Vietnamese at 150,000+

{{editprotected}} The Vietnamese Wikipedia is over 150,000 articles. Please move the link to the correct section. --Yair rand (talk) 19:27, 12 January 2011 (UTC)

  Done [stwalkerster|talk] 20:29, 12 January 2011 (UTC)

Azerbaijani wikipedia

Hi. Please add Azerbaijani wikipedia ([[az:]]) to "More than 50,000 articles" section. So last week we have reached 50.000 articles: 10 000+ articles--Wertuose (talk) 21:30, 15 January 2011 (UTC)

I'm somewhat hesitant on this. I went through the Azerbaijani Special:Random and a lot of the articles it took me to were very small, some of them even completely empty. We generally have criteria beyond the number of pages in the mainspace. I'd like to hear other opinions. HJ Mitchell | Penny for your thoughts? 22:04, 15 January 2011 (UTC)
Yes we have empty articles, but most of them were created by one user and all of his articles about Azerbaijan cinematography. We are working on this problem. May be we will delete some of this empty articles. But our principle about small articles is that small article must contain main necessary knowledge about subject of article. Ofcourse in future we will enlarge them. Wertuose (talk) 10:44, 16 January 2011 (UTC)
Indeed, I'm seeing mostly stubs and placeholders (on both cinematic and geographic subjects). This Wikipedia doesn't appear to meet our qualitative criteria. —David Levy 12:18, 16 January 2011 (UTC)
According to your methods of defining quality of Wikipedia I also went through the Estonian and Croatian wikis, which are in the "More than 50,000 articles" section. And I found that a lot of the articles are very small, some of them even completely empty, there were a lot of stubs and placeholders. As an administrator of Azerbaijani Wikipedia I must say that our wiki is much better than this two wikis. So what can you say about it? Wertuose (talk) 22:16, 16 January 2011 (UTC)
I'm waiting for your respond. Wertuose (talk) 08:15, 18 January 2011 (UTC)
In my review it doesn't seem to compare that favorably against those two Wikipedias yet. There just too many more placeholders in the Azerbaijani Wikipedia. It looks like the project is on the right track though, and in any case, listing on the English Wikipedia page isn't really that central to improving the Azerbaijani Wiki. It's also not that bad though, so I wouldn't really oppose very strenuously, and not at all if the placeholders were improved. - Taxman Talk 15:58, 21 January 2011 (UTC)

Latin Wikipedia

Since the Latin Wikipedia now has more than 50,000 articles, will we be adding it to this list? I took a look at some random pages there, and it appears to me that the majority are fleshed out to some degree, not all one sentence stubs or anything, so it seems like it meets the qualitative criteria as well.--Danaman5 (talk) 09:10, 16 February 2011 (UTC)

A check of fifty random articles yielded fifty stubs/placeholders (mostly stubs). —David Levy 06:15, 23 February 2011 (UTC)

fa.wikipedia

Hi! The fa.Wikipedia is over 150,000 articles. Please move the link to the correct list. thanks--sahim (talk) 20:40, 22 May 2011 (UTC)

This was done by User:David Levy but he forgot to disable the request here. — Martin (MSGJ · talk) 12:08, 23 May 2011 (UTC)
I didn't see this request until now. I responded to one posted at Talk:Main Page. —David Levy 03:57, 26 May 2011 (UTC)

Please add Tagalog Wikipedia

Please add Tagalog Wikipedia on the 50,000+ list. Currently, it has over 51,000 articles and depth of 19. --Jojit (talk) 10:47, 31 May 2011 (UTC)

Esperanto Wiki

The Esperanto wikipedia just hit its 150 000th article. If you could move it into the correct category, we would be much obliged.

Hiyayaywhopee (talk) 16:17, 7 August 2011 (UTC)

Done. —David Levy 17:40, 7 August 2011 (UTC)

650,000 -> 700,000

In the next two weeks, pt: will reach 700,000 milestone, and the "More than 650,000 articles" cell may be change to a more round number. emijrp (talk) 21:26, 24 September 2011 (UTC)

I have made the change to 700,000.[8] PrimeHunter (talk) 22:51, 5 November 2011 (UTC)
Thanks. emijrp (talk) 23:13, 5 November 2011 (UTC)

Hindi Wikipedia

Hindi WP is more than 100,000 Articles.:-(

59.182.149.76 (talk) 08:32, 1 October 2011 (UTC)

Serbo-Croatian Wikipedia

has more than 50.000 articles and should be included in the appropriate list on the main page.--Antidiskriminator (talk) 08:11, 9 November 2011 (UTC)

This message was left at Talk:Main Page. As I know that reaching the 50,000 mark does not automatically guarantee inclusion on the list, I have copied the request here. BencherliteTalk 08:33, 9 November 2011 (UTC)
I just now saw the discussion about Macedonian Wikipedia and details about 50-article random pages test. I performed two 50-article random pages tests on Serbo-Croatian Wikipedia with exactly the same result: 26 articles were stubs or disambiguation pages and 24 articles were larger than stubs. --Antidiskriminator (talk) 10:37, 9 November 2011 (UTC)
Not related to the Serbo-Croatian Wikipedia, but is the 50-article random pages test really fair, considering that performing that test on the English Wikipedia often turns up a lot of stubs? I just tried it, and got 20 stubs, 18 non-stubs, and 12 pages that didn't have stub templates but needed them. AdamSommerton (talk) 16:42, 9 November 2011 (UTC)
This has been addressed in previous discussions. The relative quantity of non-stub articles isn't directly relevant; the absolute quantity is. The latter determines the likelihood that a reader typing "x" will find a useful article about x. (He/she probably doesn't care whether a non-useful article about y exists.)
For smaller Wikipedias, we gauge the relative quantity of non-stub articles purely as a means of estimating the absolute quantity.
For larger Wikipedias, this is unnecessary. Your English Wikipedia sample contained 36% non-stub articles, which translates to more than 1.3 million. Even if we cut that percentage in half, it amounts to about 682,000 non-stub articles. —David Levy 19:25, 10 November 2011 (UTC)

Any opposition to this being added to the template? — Martin (MSGJ · talk) 11:56, 10 November 2011 (UTC)

The active editors / admins count is lower than any of the current ~50,000s (both around a third of Simple and 70% of Slovenian; the only user stat I could find lower than Serbo-Croatian is that Galician has one less administrator). I've pinged user:David Levy, who has done most of the recent additions, to see what he thinks. Chris Cunningham (user:thumperward) (talk) 12:55, 10 November 2011 (UTC)
In my opinion, there's enough to tip the scales in favor of inclusion.
Specifically, the proportion of non-stub articles is decent, many of the stubs appear human-created/edited and ready for expansion (as opposed to a massive pile of bot-generated year/place articles that no one touches), and I haven't seen any placeholder articles (those comprising empty sections).
My impression is that a small group of editors has built this Wikipedia slowly but surely, with no apparent efforts to artificially inflate the article count or employment of methods resulting in such a side effect. —David Levy 19:25, 10 November 2011 (UTC)

  Added — Martin (MSGJ · talk) 18:59, 11 November 2011 (UTC)

Should this template be updated?--В и к и T 22:16, 11 November 2011 (UTC)

Done. —David Levy 06:06, 16 November 2011 (UTC)

Language identification on the menu bar

Dkepi (talk) 18:16, 14 November 2011 (UTC) November 14, 2011 For users who wish to check how a word is spelled in another language but do not have any knowledge of this language, it would be very useful if each language name listed in the menu bar was spelled out in English or in the name of the prevailing language at one's location. Please, forgive me if this topic has been discussed before. Thank you for considering. Best regards, Daniel Keppler Dkepi (talk) 18:16, 14 November 2011 (UTC)

The use of this template is only to present the links to other language Wikipedias in the bottom of the main page. Your suggestion might be done, but another problem is the use of different writing systems that warrants not only IPA, but a correspondent writing in English as well. I doubt someone will like to see the spelling of a word written in no comprehensive alphabet.--Kiril Simeonovski (talk) 19:42, 14 November 2011 (UTC)
One way of checking this is to open the page for editing and look at the language tags at the very end: these identify what the article is called in different languages, and have language codes next to them. Chris Cunningham (user:thumperward) (talk) 09:25, 18 November 2011 (UTC)
Someone has removed this facility recently, and I miss it very much. I want to see which way a topic is translated in other languages, by just moving my cursor over them. JMK (talk) 07:36, 29 November 2011 (UTC)

Macedonian wikipedia

Macedonian Wikipedia (mk) now has 50,000 articles, so it needs updating here. Thank you! --B. Jankuloski (talk) 13:45, 2 October 2011 (UTC)

  Added — Martin (MSGJ · talk) 16:09, 2 October 2011 (UTC)
Undone. As noted in the template's documentation, "this is not a complete list of Wikipedias containing 50,000 or more articles; Wikipedias determined to consist primarily of stubs and placeholders are omitted." Apart from quantity, the Macedonian Wikipedia's condition appears little different from when this was discussed previously. —David Levy 16:24, 2 October 2011 (UTC)
David, the situation with the Macedonian Wikipedia is far better now and you should not rely on the previous discussion (we deleted many empty articles). Why are you so against the mk Wiki? Regards, --MacedonianBoy (talk) 19:23, 2 October 2011 (UTC)
I'm not relying on the previous discussion. I performed our standard 50-article sample. Of the 50 random articles that I viewed, 46 were stubs or placeholders (some of them completely empty, apart from issue tags). —David Levy 15:32, 4 October 2011 (UTC)
Yes, they were. Make new analysis maybe you'll get different results.--MacedonianBoy (talk) 18:14, 4 October 2011 (UTC)
You seem to have misunderstood. I'm not referring to the 50-article sample that I performed upon the previous request. I reran the test in response to this request. —David Levy 14:28, 5 October 2011 (UTC)
Sorry that I enter this discussion lately, but David the outcome of your test really induces me to think that either you've not performed the test properly and your outcome is nonsense, or the 50-article test is not the right measure. Just have a look at the special page that lists articles ordered by length on the Macedonian Wikipedia, and you can note that more than 5,000 articles are larger than 10 KB (sample article with length around 10 KB). Since the special page does not allow to view more than 5,000 articles, the next step is to use statistics to fit the values for the next thousands of articles. The conclusion is likely that more than 20,000 of the articles are larger than 5 KB (sample article). I admit that Macedonian Wikipedia contains about 5,000 placeholders (10% of total), but the staled number of 7,000 (then roughly 18%) that used to be the number of placeholders when surpassing 40,000 is vehemently greater in the terms of both, its absolute value and proportion.--Kiril Simeonovski (talk) 21:02, 4 October 2011 (UTC)
P.S.Most of the placeholders are articles about years (~3,000), ~1,500 of them are about villages, and the others are miscellaneous.--Kiril Simeonovski (talk) 21:15, 4 October 2011 (UTC)
I'm sorry that you dislike our criteria. —David Levy 14:28, 5 October 2011 (UTC)
This answer really shows that you're ignorant on the things that I've mentioned above. But the figures in the special page are self-evident and represent something that is contradictory to your conclusions. Again I sincerely appreciate to see a more friendly answer, where you can explain the outcome of your solely performed test. Best.--Kiril Simeonovski (talk) 17:38, 5 October 2011 (UTC)
In my 50-article sample, 8% of the articles were neither stubs nor placeholders. Given the margin of error, it's entirely possible that the actual percentage is slightly higher, as suggested by the special page. That doesn't affect the resultant conclusion. —David Levy 18:30, 5 October 2011 (UTC)
Wait. You need to explain your definition for a stub? We cannot say that it is any article with length bellow 10 KB.--Kiril Simeonovski (talk) 18:49, 5 October 2011 (UTC)
I've never seen a precise definition. I'm counting articles comprising no more than a few paragraphs. Some are tagged as stubs, while others are not. I didn't count this article as a stub, despite the fact that it's tagged as one. —David Levy 19:14, 5 October 2011 (UTC)
I must concur here, but we usually template articles not to mark them as stubs, but to note that the article needs more content. So far I've seen articles with very long copy marked as stubs, but others with almost sufficient content not marked. Normally, the definition of the template's use is not identical anywhere. Another thing that you don't gauge is the frequency of long-sized to middle-sized articles with length around 6 KB, that in the order of the special page would be classified within the interval of 15,000-17,000 (ex. 1, 2, 3 or 4). So, you don't need to perform the 50-article test to get to a proportion of about 30% of the articles that are larger than 6 KB, i.e. apparently being classified as non-stubs. Even if you claim that the small error is possible, which is absolutely true, the special page shows the first 5,000 articles ordered by length, but not the non-stub articles only. Best.--Kiril Simeonovski (talk) 19:55, 5 October 2011 (UTC)
As noted above, I didn't rely on the "stub" templates (and counted one "stub"-tagged article as a non-stub).
Forgive me, but I don't understand the methodology under which you deduce that about 30% of the articles are larger than 6KB.
I just performed the 50-article test again. This time, I viewed five articles that I didn't classify as stubs or placeholders. Two were rather long and three were similar in length to those linked above. —David Levy 20:30, 5 October 2011 (UTC)
No, you don't need to perform it again and again. The methodology used to deduce that roughly 30% are articles larger than 6 KB coincides with the distribution of the 5,000 articles listed on the special page. Use of a mathematical model yields a simple conclusion that the frequency of the articles is greater as the length decreases. The equation leads to an interval of about 2,000 articles with 6-7 KB, ordered within first 15,000-17,000 articles. Best.--Kiril Simeonovski (talk) 20:49, 5 October 2011 (UTC)
I'm no mathematician, but that doesn't strike me as a reliable method. —David Levy 20:56, 5 October 2011 (UTC)

David, I believe that your interpretation of the template's documentation is flawed. I assume the intent of the criteria is to exclude artificially inflated encyclopedias (thousands of articles in a manner of days), such as the Cebuano and the Haitian Wikipedia, and thus, discourage other Wikipedias to rely on bots. I've been contributing to the encyclopedia for more than 6 years, and AFAIK, the Macedonian wiki-community, with all its flaws and misgivings, is composed of real enthusiasts that have worked really hard to achieve this milestone. I see that emotions were involved in the previous discussion, and you do have some points, but I don't see that your strict criteria are applied to other Wikipedias, as well. I will assume good faith, but I'm not sure how this helps the English Wikipedia, the Macedonian Wikipedia, or the Wikimedia movement, in general. Do you care to elaborate? --FlavrSavr (talk) 20:20, 2 October 2011 (UTC)

Maybe I wasn't clear enough with my question, so I didn't receive a reply. So let me rephrase it again: How does excluding a LanguageX from the main page interwiki links help the quality of the English Wikipedia? I noticed that, during the last request for inclusion, a lot of time was spent on explaining why a Wikipedia shouldn't be included, but why so strict in the first place? I mean, it's not that there is some en.wiki policy or guideline that should be implemented here - even if there was, policies or guidelines that do more harm than good to the project should be ignored. I mean, come on, it's a template documentation note, or to be more precise, it's an interpretation of a template documentation note. Seriously, what's the big deal about it? --FlavrSavr (talk) 00:34, 4 October 2011 (UTC)

1. I wrote the text in question, so it's unlikely that I'm misinterpreting it.
2. If you believe that the list contains Wikipedias of quality comparable to the Macedonian Wikipedia's, please name them.
3. We can't link directly to every Wikipedia, so we seek to include the ones of greatest utility to readers. If we were to abandon our qualitative assessment, numerous Wikipedias would be pushed off the list by others of higher quantity but lower quality.
Even now, some editors believe that the list is of little value and wish to remove it entirely. The insertion of Wikipedias primarily composed of stubs and placeholders would fuel that sentiment.
4. My apologies for the delay, which reflected no lack of clarity on your part. I was very busy and forgot to check back here. In the future, please feel free to leave a reminder on my talk page. —David Levy 15:32, 4 October 2011 (UTC)
Mr David Levy there is simple solution for this dispute, just move the line above 100k :) as you do before (from 40k to 50k) because you have unreasonable explanation why not Macedonian.--Brest (talk) 09:34, 5 October 2011 (UTC)
I'm sorry that you dislike our criteria. —David Levy 14:28, 5 October 2011 (UTC)
I'm so sorry that you dislike Macedonia, tx :D. --Brest (talk) 20:25, 6 October 2011 (UTC)
Now you're accusing me of bigotry. Do you honestly believe such a thing? Does this strike you as a constructive approach? —David Levy 22:05, 6 October 2011 (UTC)
Macedonian will be there, and you will continue dislike it, that's your problem.--Brest (talk) 20:55, 7 October 2011 (UTC)

Revisited

I've just read the discussion above and it appears that one editor is blocking a change to this template which is supported by a great number of editors, especially if you consider all the other past discussions on this issue. I am not accusing anyone of bias, but the problem with this "rule" is that it is arbitrary and subjective, and so upholding it can be perceived as being unfair. I also note that DL is the only editor to be opposed to the addition. Therefore I intend to re-add it shortly. David, if you believe that the community will feel as strongly as you about this, you are welcome to advertise this discussion in a more prominent place. — Martin (MSGJ · talk) 11:55, 10 November 2011 (UTC)

I just noticed the above message.
I'm confused as to how the change is "supported by a great number of editors", as very few have commented at all. Most are users from the Macedonian Wikipedia (who came here specifically to demand that it be linked). Nothing else sets it apart from the numerous other Wikipedias omitted from the list.
Surely, you don't believe that it makes sense to selectively disregard the criteria when editors from a particular Wikipedia happen to object to their normal application (which some of them attribute to bigotry against Macedonian people).
I certainly don't assert that my opinion is sacrosanct, but in the event of good-faith disagreement, consensus is required for an addition. Persons wishing to make the change are welcome to advertise this discussion in a more prominent place. —David Levy 09:30, 27 March 2012 (UTC)

Serbian Wikipedia

Serbian Wikipedia passed 150.000 articles. Please, someone should move .sr to appropriate section of the template. Thanks in advance. --WhiteWriter speaks 21:16, 20 November 2011 (UTC)

Done. —David Levy 22:54, 20 November 2011 (UTC)

Tagalog Wikipedia

I was wondering if the Tagalog Wikipedia can be included in the list. It has 54,452 with a depth of 21. --Scorpion prinz (Talk | contribs) 13:16, 11 January 2012 (UTC)

I just sampled of 50 random articles, of which 46 were stubs and placeholders (articles that are either extremely short or essentially empty). This does not meet our standards for inclusion on the main page list. Sorry. —David Levy 23:59, 12 January 2012 (UTC)

Georgian Wikipedia

Please, do add here Georgian Wikipedia as well. It's more than 54.000 articles written in Georgian wiki. http://ka.wikipedia.org/wiki/ RSVP. Georgianჯორჯაძე 21:43, 10 February 2012 (UTC)

There is discussion about this at Talk:Main Page#Put Georgian language wikipedia into More than 50,000 articles section! PrimeHunter (talk) 20:19, 12 February 2012 (UTC)
Indeed, and the same user initiated and participated in that discussion. Please refrain from forum shopping, GeorgianJorjadze. —David Levy 21:25, 12 February 2012 (UTC)

Telugu Wiki has more than 50000 articles

I would just like to tell that the Telugu Wikipedia crossed 50,000 articles.So I think it should be added to the mainpage.Thanks Srikar Kashyap Talk 14:27, 1 May 2012 (UTC)

You heard the man. --MZMcBride (talk) 14:30, 1 May 2012 (UTC)

What's the language code? They usually have two or three letters, as in de for the German Wikipedia http://de.wikipedia.org/ --Redrose64 (talk) 19:27, 1 May 2012 (UTC)
te (cf. List of Wikipedias ;-) --MZMcBride (talk) 20:38, 1 May 2012 (UTC)
As explained in the template's documentation, "this is not a complete list of Wikipedias containing 50,000 or more articles; Wikipedias determined to consist primarily of stubs and placeholders are omitted."
Per our standard practice, I just sampled 50 random articles, of which 47 were stubs (example) or placeholders (example). Of the other three, two were short articles containing multiple placeholder sections (example) and one was written in English (link). —David Levy 20:57, 1 May 2012 (UTC)
Added, then undone again. BTW: per WP:LISTGAP, please do not insert blank lines into lists. --Redrose64 (talk) 21:03, 1 May 2012 (UTC)
Pardon? To what edit(s) are you referring? —David Levy 21:18, 1 May 2012 (UTC)
This one. No other edit after my first post inserted blank lines. --Redrose64 (talk) 21:30, 1 May 2012 (UTC)
Your message was indented below mine, so I thought that you were addressing me.
It's been explained in the past that this technical issue pertains specifically to bulleted lists. Someone can correct me if that's inaccurate or has changed. —David Levy 21:41, 1 May 2012 (UTC)
Sorry... at the time that I began editing this thread the last post was MZMcBride's second (20:38). I got an e/c due to your 20:57 post, so copied my text, backed out, went for [edit] again, pasted it in and added another colon.
Anyway, if you follow WP:LISTGAP it does state "even when using unordered or definition lists". When we indent with colons in the wiki markup, the final HTML uses <dd>...</dd> elements enclosed in <dl>...</dl> elements - the building blocks of a definition list. The existence of a blank line will close all <dl> tags, and then open a fresh set. --Redrose64 (talk) 22:11, 1 May 2012 (UTC)
Our documentation needs to be clarified, as the situation apparently was unclear even to users fairly knowledgeable on such subjects (those who explained it previously). Most editors would have no idea that an unbulleted talk page reply is coded as a "definition list". —David Levy 23:53, 1 May 2012 (UTC)
David, regarding article counts, the standard seems completely arbitrary, though possibly a bit biased against smaller projects. At some point, whatever value is at Special:Statistics is used, whether it's for moving a project between tiers or showing the total number of articles on this page. That is, this template uses and advertises {{NUMBEROFARTICLES}}, but that figure doesn't take into account the thousands of stubs, disambiguation pages, and other "non-articles" throughout the English Wikipedia. Similarly when articles are moved between tiers, it seems to be based on whatever value is at that project's Special:Statistics page, not on what we think the value might be if we excluded stubs and placeholder pages and other ill-defined "non-articles." --MZMcBride (talk) 04:08, 2 May 2012 (UTC)
Indeed, the tiers rely on the raw quantity of articles (with no consideration of the quality thereof). All contain significant quantities of stubs, so this is more or less a wash. (The only other practical approach would be to eliminate the tiers entirely, which might not be a bad idea.)
For smaller Wikipedias, we gauge the relative quantity of non-stub/placeholder articles purely as a means of estimating the absolute quantity. The latter determines the likelihood that a reader typing "x" will find a useful article about x. (He/she probably doesn't care whether a non-useful article about y exists.)
This is necessary both to make the best use of our limited space (i.e. to prevent tier quantity adjustments resulting in Wikipedias being pushed off the list by others of higher quantity but lower quality) and to avoid encouraging Wikipedias to deliberately create stubs and placeholders (thereby reducing the sites' overall quality) as a means of making our list. (This was a problem before we began checking.)
For larger Wikipedias, this isn't an issue. Someone recently estimated that 70% of the English Wikipedia's articles are stubs, but that leaves more than 1.18 million articles that aren't. —David Levy 20:50, 2 May 2012 (UTC)
Regarding the order of Wikipedias in each tier, are they ordered by something in particular? I don't think the current order correlates with article count, or at least I don't think it does so with much accuracy, if that's the intention. I'd have to pull the numbers to be sure, but then I'm back at trying to figure out what number to pull.... --MZMcBride (talk) 04:08, 2 May 2012 (UTC)
They're ordered alphabetically (by the language's native name when feasible). —David Levy 20:50, 2 May 2012 (UTC)

The : talk page indentation syntax isn't creating a list, so any concerns regarding continuity of a list are baseless, as far as I can tell. And, tangentially, no screen reader is going to have a chance in hell of interpreting talk page messages in any kind of logical order, so list syntax is even less relevant. Spacing out comments makes them vastly easier to distinguish in the edit window, so unless there are actual reasons for pointing to and following WP:LISTGAP, I think spacing out replies makes much more sense.

I find it a little funny that there's now ongoing discussion about proper (semantic) list syntax when this template is currently using {{·}} instead of {{flat list}}. :-) --MZMcBride (talk) 03:55, 2 May 2012 (UTC)