Wikipedia:Articles for deletion/Geographic.org

Geographic.org

Geographic.org (edit | talk | history | protect | delete | links | watch | logs | views) – (View log)

(Find sources: Google (books · news · scholar · free images · WP refs) · FENS · JSTOR · TWL)

Note to admins/closers - per WP:IAR, although the nominated article has not been created, IMHO, this discussion should be allowed to run for a week, to get as full an input on the issues raised as possible. This discussion could affect up to 2,537 articles. Mjroots (talk)

Over the last few months, 2537 articles on villages (from Afghanistan, Kazachstan, Saudi Arabia, Oman and a few from Iran) are created nearly-identical, in rapid-fire batches of five or more articles per minute, based on (or at least referenced to) an outdated or incorrect source, geographic.org (but sadly not referenced to the specific page about the village, but to the main page of the site). Geographic.org is a copy of an older version (or a poor copy) of the geonames database of the National Geospatial-Intelligence Agency, in itself a reliable source, which I used to check these articles.

These articles are now listed in Category:Articles only referenced to geographic.org, and I propose to Delete all articles in this category.

A lot of articles are correct (my estimate would be around 1,000 of the 2,500), but the other 1,500 have major problems (which each time one or two of many possible examples):

• Subject doesn't seem to exist: Ab Kalak
• Subject is normally known with another name: Khvoshi, Ahsham-e `ali ahmad khiari
• Subject is situated in the wrong province: Pitigal, Al ahramiyah
• Article title has the wrong capitalization: majority of the Saudi Arabia and Iran articles, and many Oman articles: As suwayriqiyah, Al ahramiyah
• An article for the location already existed at a better name: Al `utaybiyah exists as Al Utaibiyyah, Bandar jissah exists as Bandar Jissah
• Multiple articles for one location are created: Kabanbay and Qabanbay, Komsomol'skoe, Kazakhstan and Komsomol'skoye, Kazakhstan
• Multiple actual locations correspond to the single article created: Komsomol'skoye, Kazakhstan
• The subject is not a village, but a neighborhood, a district or even a farm: Al `utaybiyah, or Tokhmari (described in the article as a "large village", actually a farm)

Many articles have more than one of these problems, e.g. Kariz Dashak should be at Karez-e Dashak and is not in Farah province but in Herat province. User talk:DGG#More Afghanistan has an analysis of twenty articles (the first and last ten of the category at that moment), and only 7 of the twenty had at first glance no major problems. 13 of the other ones needed moving, correct content, or simply deleting, even though these articles only have one line of text. Wikipedia:Reliable sources/Noticeboard#Geographic.org has some more discussion on this.

I started with an individual AfD, Wikipedia:Articles for deletion/Dara-I-Pech, for a single problematic article I found, but looking more into this has shown that this is such a huge group of problematic articles, that the loss of a number of correct short stubs doesn't outweigh the damage of having so many clearly and seriously incorrect articles. I do not dispute that many of these places deserve an article, but we have no deadline, and it is better that for the time being we don't have an article on them, than that we have more than a thousand unreliable, dubious ones.

The articles that tipped me over towards deletion instead of some other rescue mechanism (which would take much, much more time) were Bona see Buna (the ultimate evidence that no meaningful human check is being done when these articles are created), and Al madinah al munawwarah, which is not only badly capitalized, but is actually another name for the "village" of Medina, population 1,300,000. Fram (talk) 08:20, 11 May 2011 (UTC)

PLEASE NOTE: That Geographic.org was not the source used to create these articles, it was a confusion with Geonames from which the settlement database was downloaded from which is considered to be reliable for verifiying populated settlements. The nominator himself acknowledges that geonames is a reliable source and it was that site which was actually used to create most of the "problematic articles" so this nomination is completely redundant given that the vast majority were not even created using geographic.org but were obtained from a 2008 version of geonames. NO WAY are there 1500 "problematic articles". I've assessed 300 ish so far and I've found just one error. And those which do have "problems" the condition is extremely trivial and certainly not a cause for deletion. There is no major urgency to constitute this mass AFD. The articles are mostly correct other than being sourced to geographic.org rather than geonames or in regards to certain Afghan villages now being in a different province. Virtually all 2500 entries are valid populated settlements and can be viewed on google maps. Oh and all articles with non capitals have been moved to capitals and Tokhmari IS a sizeable village not a farm and duplicates can easily be spotted. Komsomolskye can be viewed on google earth here. A major town actually and obviously notable. As for lower casing or different transliterations, there is always dispute over Arabic and other Asian transliterations, and genames database does not always get it right, it referred to one place as Jabal something; well it is widely accepted that Jebel is the ideal transliteration, meaning mountain. So they are not errors they may be slightly different. By the nominator's own admission 1000 articles have no problems whatsoever, does this not strike anybody then as an extreme solution to blindly mass delete verifiable real world places in this way? Please think about this people approaching this AFD and actually look further into this than taking Fram's word for it. There is absolutely no cause for this. The articles yes need expansion and improvement but deleting the entire lot is wrong. ♦ Dr. Blofeld 13:27, 12 May 2011 (UTC)

• Comment. It is not clear that AfD is the appropriate way to address the problems cited by the nominator. In particular, it is obvious that WP:BEFORE has not been conducted on each article nominated for deletion. This is not to suggest that the bulk deletion of these articles (without prejudice to creation on a one-by-one basis) is inappropriate. And if the community decides that this is the appropriate venue for consideration of this proposal, then I will support it. ^Bongo_matic 08:39, 11 May 2011 (UTC)
  • How about this. Since this is Dr. Blofeld's horrid mess, Dr. Blowfeld should be the one to sort through all several thousand of these and pick out the ones that should be kept. If he's going to generate 2,537 one line articles, he should be the one to find the three or four out of that that warrant keeping. Looking at these, I doubt he can find more than that, if that at all. Sven Manguard Wha? 08:47, 11 May 2011 (UTC)

• (edit conflict) Delete All I'm sorry, but what the hell was Dr. Blofeld thinking? One line is not an article, one line isn't even a stub. Every one of the 30 random articles from that category that I checked had less than twenty words. It's not useful, especially if it's not guaranteed accurate. Sure, we should shoot for encyclopedic coverage of every city and province in the world, but just like every other article, multiple reliable sources are a requirement. Wikipedia is not an atlas, every single insignificant town does not warrant coverage. If sources can be found that indicate some semblance of notability at the WP:GNG threshold, then yes, like anything else it warrants an article, but if someone were to tell me that this is an acceptable article, I'd assume that they were a brand new user and direct them towards WP:GNG or the more blunt WP:42. TLDR This is unacceptable, these articles are all worthless. Sven Manguard Wha? 08:44, 11 May 2011 (UTC)
  • Regarding "Wikipedia is not an atlas, every single insignificant town does not warrant coverage", it has been the general consensus for years that if a town of any population can be verified to exist, it doesn't need to pass GNG (see WP:NPLACE). Jenks24 (talk) 08:56, 11 May 2011 (UTC)

• Keep articles and reference them properly and fix problems. I have just done two. Mass deletion, without even placing a template on the artciles is definitely not the way to go about this. Jezhotwells (talk) 08:47, 11 May 2011 (UTC)
  • I'm sorry, but you didn't. You edited one, not two articles, but you didn't fix the problems. Al hijar is after your edit still sourced to an unreliable source, and still incorrectly capitalized. Fram (talk) 08:54, 11 May 2011 (UTC)
    • In principle, though, it's a fine plan - though "keep" may not be the best summary. If you reference one of these articles properly then it should no longer be in the category, and it escapes this AfD. bobrayner (talk) 14:03, 11 May 2011 (UTC)

Note: all pages under discussion now have the AfD template (transcluded through the geographic refimprove template). Fram (talk) 09:00, 11 May 2011 (UTC)

• Delete all unencyclopedic entries, without prejudice to re-creation subject to WP:RS etc. Mjroots (talk) 09:23, 11 May 2011 (UTC)

• Keep This is a reckless approach to deletion and in my view sets a dangerous precedent. We don’t delete articles because a particular source is deemed un-reliable, we delete articles because there are no reliable sources to support notability of the article topic. In this case, even the nom admits that many of these geo-locations are valid and warrant articles. Nom didn’t get the results wanted in this discussion [1], but still choses to uni-laterally declare that geographic.org is unreliable. On an article by article basis, there may be instances where geographic.org is not a reliable source and/or contains significant content errors and no other sources exist to support notability of the article topic. However, if more recent or reliable data sources are available to support notability and content (must be available if errors in geographic.org can be validated), they should be used in the article. Unless WP:BEFORE is applied to each of these articles, we run the risk of creating a new and dangerous deletion tactic--find a way to discredit a source and then nominate all the articles that use that source for deletion. Categorical deletion as is being proposed here is a really poor approach to improving the encyclopedia. All that said, I do believe there should be more engagement with the creator of these articles to get the sourcing correct on an article by article basis.--Mike Cline (talk) 09:33, 11 May 2011 (UTC)

There have been extensive discussion of these articles, the sourcing, and whether the mass creation is appropriate in other fora. See, for example, the RS/N discussion, including the rough poll where everyone opining agreed that articles should not be created with single sourcing to geographic.org. As I mentioned above, I agree that AfD is not an ideal way to address these articles that were created in violation of what turns out to be (wasn't known at the time) community standards. ^Bongo_matic 09:44, 11 May 2011 (UTC)

• • (ec)The WP:RSN discussion you quote did produce the result I wanted, thank you. In the end, there was clear (but not unanimous) support that geographic.org is not a reliable source. Apart from that, this is hardly reckless (or dangerous), but well thought out. Article deletion should be as easy as article creation, easy come and easy gone. If WP:BEFORE doesn't apply to article creation (even for experienced editors), and it can be coupled with mass article creation (basically a violation of WP:BOTPOL, then the only way to check such flood of poor articles is deletion with recreation allowed on a case-by-case basis. Way should the only way of improving be "keep rubbish until someone cleans it"? Knowingly keeping around many, many clearly and fundamentally incorrect articles is a much more dangerous and reckless approach than mine, and a typical example of the way WP:BEFORE gets misused. Fram (talk) 09:48, 11 May 2011 (UTC)

Actually, it doesn't appear to be "basically a violation of WP:BOTPOL", but an actual, specific violation of it (see Wikipedia:Bot policy#Mass article creation. However, as mentioned above, that suggests that this may be the optimal forum to address the issue. ^Bongo_matic 11:21, 11 May 2011 (UTC)

Re: "we delete articles because there are no reliable sources to support notability of the article topic" - I would be happy to keep any one of these articles which gets updated with a reliable source in the next week, or however long this AfD runs. A week is plenty of time to find a source for a notable subject. Deletion is the best way to deal with the remainder - those which were produced en masse from an unreliable list and for which nobody has yet been able to find any alternative source (let alone a source which gives the in-depth discussion required by WP:V). If a bot operator finds it harder to automate the task of writing articles which meet wikipedia standards, rather than simply creating one-line articles from a list, they have my sympathy; but that's no reason to keep the bot's earlier indiscriminate output. bobrayner (talk) 12:05, 11 May 2011 (UTC)

Bob, my objection is the Categorial Deletion of 2500+ articles without subjecting each to the normal debate in AFD. If indeed as Bongo said above that the creation was a violation of WP:BOTPOL and such a violation is sanctioned with mass deletion, I have no issues with that. If a violation of BOTPOL is a valid reason for mass deletion, so be it. But mass categorial deletion, claiming no RS is wrong. And you are wrong when you say nobody has yet been able to find any alternative source because even nom has said that National Geospatial-Intelligence Agency, in itself a reliable source, which I used to check these articles. --Mike Cline (talk) 12:30, 11 May 2011 (UTC)

• Delete all. Nominator's argument is compelling. WP:RS fail and no other sources => delete. Tijfo098 (talk) 09:38, 11 May 2011 (UTC)
• Delete I can understand the argument that incorrect information in a stub can be fixed, but improperly titled pages can't even be found, so are very unlikely to get fixed. For example, say I am looking for Al Qal'i. Blofeld created an article at Al qal'i, but this does not show up in the search results. Trying to fix the issue will simply take way more time than massdeleting them and recreating from a proper source. Yoenit (talk) 10:03, 11 May 2011 (UTC)
  • Even worse, it seems the article is at Al qal`i, using a non standard accent. What a mess. Yoenit (talk) 10:05, 11 May 2011 (UTC)
• An unusual nomination, and I'm reminded of WP:KITTENS. I'm rather tempted by Mike Cline's view, but on balance I don't want to apply a simplistic one-size-fits-all measure to 2,500+ articles—irrespective of whether that measure is "keep" or "delete". In the circumstances, can we please incubate the whole lot of them so that each can be considered in the way that Mike Cline recommends?—S Marshall T/C 11:31, 11 May 2011 (UTC)
• Delete all, or incubate - and I dearly hope that future recurrences can be avoided. This project is not a race to create as many zero-quality placeholder articles as possible from an unreliable list, in the hope that somebody else will come along and clean up the mess later. Commendable research by Fram (talk · contribs) and others. The Bona see Buna example is really telling. If anybody is able to properly source a subset of these articles, I would be happy to keep that subset. bobrayner (talk) 11:57, 11 May 2011 (UTC) (later amended to add "or incubate").
• Delete all per nom, with mass incubation an acceptable alternative (with some special measure to identify the articles within the incubator, to avoid swamping everything else in it). This sort of unreliable content does not enhance the encyclopedia at all. These articles also appear the result of a mass creation effort of exactly the sort which Wikipedia:Bot policy#Mass article creation is supposed to require prior authorisation for, to prevent the need for exactly this sort of deletion discussion. Rd232 ^talk 11:58, 11 May 2011 (UTC)

Keep all and expand

First of all, lets please assume good faith people. Quite frankly the tone here is disgusting and good faith attempts to work towards a fuller coverage of what are generally all legitimate populated settlements should not be responded with "what the hell was Blofeld thinking" and"what a horrid mess". So please drop the personal attacks and anguish. This a wiki and content can be expanded, stubs are only useless until they are expanded. I agree that content is not best started in this way and should be started with sources beyond genames but given the sheer amount missing I felt the need to create them. As for "what the hell was Blofeld thinking" I did a google book search for some of the articles listed in geonames and I also looked them up on google maps. The VAST majority were legitimate settlements which can be expanded, even if these countries typically have very poor web coverage. It was intended in good faith as a basic start to build up coverage like Alishang which can be expanded at a later date. The idea is so eventually we have 2500 articles like Alishang. Not what people want? How sad. I expanded many of the Afghan villages I started and they are clearly notable. Nuke the whole lot because a few might not be perfect and may use different transliterations is a lousy solution to a problem blown into massive proportions by Fram who didn't get any support an ANI so has resorted to this. It is a fact that the vast majority of these articles are legitimate places and could be expanded. To delete them all would be contrary to the nature of wikipedia and a waste of time. People have friends who died in many of the villages I started in Afghanistan and could quite reasonably expand them. I'm very disappointed in you all that you a completely blind to the potential of these articles. What the nominator and others are missing here is that the vast majority are genuine populated settlements even if a few might now be in different districts or occasional duplicates (which could easily be spotted and eradicated. As for Ab Kulak not existing, bull shit, [http://maps.google.com/maps?ll=37.059722,66.043333&spn=0.1,0.1&t=h&q=37.059722,66.043333 zoom in here'. The apparent mass problems have been greatly exagerrated by the nominator , besides in regards to Afghanistan I have a UN directory downloaded to sort them into districts so the vast majority of problems would be fixed in the process of that. ♦ Dr. Blofeld 12:05, 11 May 2011 (UTC)

If you (or anybody else) find good sources for any of those articles, they would no longer be in the category, and could be kept. I'm unsure why you're angry with somebody wanting to delete minimally-sourced articles if you say they can get properly sourced; why aren't they better sourced, then? The best way to avoid deletion here is to add a source - any source, even a passing mention in Google Books - and move the article out of the category. The bar has already been set much lower than is usual for articles at AfD; the proposal here is merely to delete those which lack any proper source, whereas other AfDs generally require that notability is established, which would be considerably harder, even if we're really sure that Village X is visible if you zoom in on Google Maps.

I'm unsure about the relevance of the "people have died" comment. People have died in practically all settlements, but that doesn't guarantee notability, and I doubt Dr. Blofeld intended to imply that soldiers from the western world are the only people who really count. Media coverage of a combat death might be a good way to establish notability, though; feel free to add such coverage to articles. bobrayner (talk) 12:22, 11 May 2011 (UTC)

Guess what? Ab Kalak does indeed exist, but is called "Ab Kuluk" or "Ab Kolok". I just spent 5 minutes looking that up, which I could have used to recreate 20 of these stubs from a proper source. Also, what on earth possesed you to use grave accents for the titles? That is 500-1000 unnecessary page moves (and useless redirects) which could have been avoided if you had thought about it for a second. Yoenit (talk) 12:30, 11 May 2011 (UTC)

A post on another page sums this up well:

"Fram, absolutely acceptable (although doing so would be pointy if the sole purpose was to expose errors in the source). Given the existence of multiple reliable sources related to geographic names, geo-locations, elevations, et. al., any error that geographic.org data might contain would be easily rectified by data from another source. Especially considering the online mapping we have today, geo-locations are extraordinarily easy to validate. Geographic.org may not be a perfect (error-free) source for geographic data, but neither is GNIS or PCGN (UK). That said, a great many NY Times articles contain factual and other errors when viewed in the light of history and new evidence, but does that make it an unreliable source? No. "

I agree its not a good idea to create "sub stubs" using geonames and no other source but they have been created now and it would be far more constructive to try to expand/correct as many as possible or at least move them into Incubator until reliable sources can be added. @Yoenit Virtually every geographic place created exists There are almost entirely genuine places, this is what Fram has failed to identify even if a few have awkward transliterations or are now in a different province. At the time of creation the articles did exist in that given province... Additional sources can be found to update and filter into correct districts and where possible find google book sources.♦ Dr. Blofeld 12:33, 11 May 2011 (UTC)

Dr. Blofeld, if you want people to assume good faith, don't talk bullshit: "At the time of creation the articles did exist in that given province..." is pure and utter nonsense. All the articles nominated were created this year, in 2011, and those with an icorrect province were equally incorrect at the time of creation. "There are almost entirely genuine places, this is what Fram has failed to identify even if a few have awkward transliterations or are now in a different province." Have you read my nomination? Where do I claim that most are not "entirley genuine places"? However, the problem is not "a few awkward transliterations", the problem is many, many errors of all kinds, as explained in the nomination. Fram (talk) 12:49, 11 May 2011 (UTC)

At the time of creation I am referring to at the time that geographic database entry was created.♦ Dr. Blofeld 13:51, 11 May 2011 (UTC)

Really? Pitigal, which you locate in Kunar Province, is according to the Geographic.org entry, last edited in 2006, located in Nuristan[2], which is the same province suggested by Geonames. So even when geographic.org is correct, the article gets it wrong. Where did you the the info that it is a village in Kunar? Fram (talk) 13:59, 11 May 2011 (UTC)

BTW Anomie has the UN index downloaded so any mistransliterations can easily be sorted. Google search [PDF] Afghanistan Settlements Index and you'll find it. The plan is to sort into districts and built a comprehensive coverage on what we have. Also note that one of those Kazakh villages was AFD d before and it ended up being fully expanded. The best way forward would be to add a UN source to verify place name and district and a source from google books, just like Ab Kolok. Gradually more and more info would become available on the web about it but that would be far more productive. Sorry I do NOT think the existing minor errors are serious enough to validate this mass deletion. Any minor errors can be easily corrected and the articles improved a little.♦ Dr. Blofeld 12:44, 11 May 2011 (UTC)

"any mistransliterations can easily be sorted."? You used this to "correct" Ab Kalak, but how are people to know that this is the same as the "Ab Kuluk" from the UN index? Ab Kalak is not listed in the actual geonames database, only Ab Kuluk and Ab Kolok, so you still have an imaginary or at least very obscure form of the name inlcuded in the article, together with the unreliable source. Had you just created an article on Ab Kuluk, your time would have been better spent. The original article didn't have coordinates, so there was no way to match the two apart from a similarity in the names, which is a dubious method. Anyway, the argument is not that the articles can't be cleaned (apart from those that are really beyond help), merged, moved, improved, properly sourced, ..., but that the effort to do so would be better spent in starting from scratch, and that so little actual value is lost by deleting these articles, and so much incorrect info is removed at the same time, that deletion is the best option here. Fram (talk) 13:01, 11 May 2011 (UTC)

Sorry, but you seem to intentionally be making life difficult. You are being excruciatingly tough picking holes in content. Can we be certain about any village in Afghanistan? Its a nightmare especially when double names exist. Its a difficult topic to get right but that does not mean pussying out of it is the correct answer. I agree, I'd rather create them properly first time with some additional sources but because of the sheer amount missing I'd like give up. ♦ Dr. Blofeld 13:06, 11 May 2011 (UTC)

• Comment Create Geographic.org from [3] promptly. Then all of us can know about the source of these articles. Just a brief glance at this website tells me that it is a very valuable resource.--DThomsen8 (talk) 13:09, 11 May 2011 (UTC)
  • This source has been discussed at Wikipedia:Reliable sources/Noticeboard#Geographic.org: basically, it is a commercial, outdated and poor copy of a reliable source, containing things like A Sherton, which is actually the place Asherton, Texas. It is a site that should not be used on any articles, since it is a bad mirror of a better, more reliable source (I have no idea if the errors were in the older version of the Geonames database, or if geographic.org took those from somewhere else or fabricated them out of thin air). Fram (talk) 13:20, 11 May 2011 (UTC)
    • This is rather tangential (it shouldn't have any effect on how we treat articles which cite only a geographic.org entry), but we do have articles on sources which are themselves notable even if we don't put much trust on content from that source. Consider Youtube or Twitter, for example. However, those two are a lot more notable than geographic.org (regardless of their relative reliability). bobrayner (talk) 13:29, 11 May 2011 (UTC)

Your two responses have already told me more than I knew about geographic.org before today, reinforcing my proposal to have someone who knows the subject create an article, which of course should have the appropriate warnings about its unreliability in various ways. I do not have time to devote to this effort until later in the month, if then, and besides I do not know enough to create a reasonable article. --DThomsen8 (talk) 13:54, 11 May 2011 (UTC)

The thing is, even beyond geograph names compiling information is incredibly difficult. FOr instance you identified Khvoshi, which is actually now Bala Deh, we've established that. However, the UN directory listed five or six villages named Baladeh and none of them appear to be in Logar Province. The information about the village of Bala Deh with 1000 inhabitants and suffering from drought in 2005 is likely Baladeh, Kabul. Its incredibly tough to know which source refers to which village and then it also becomes increasingly complicated when villages names Deh Bala pick up. Are they the same or different or not. Very demanding, and I really don't think eliminating all stubs on them is the answer as problems with accuracy and confusion of place names will continue even after they are expanded.. The UN directory at least is the best up to date guideline we have. Deleting all stubs certainly won't do anything to improve wikipedia and i believe wikipedia would be better of as a resource if they could be expanded and sorted out properly. Some day they will be recreated by somebody else and the tough taks of compiling accurate sources still stands regardless of geographic.org.♦ Dr. Blofeld 13:26, 11 May 2011 (UTC)

Who is the "we" that established that Khvoshi is Bala Deh? Khvoshi is an error from geographic.org, Khoshi is (part of) Bala Deh. And that it may be hard to find good sources is no excuse to create poor articles from a poor source in a bot-like manner, but should be an indication that we should have proceeded extra careful instead. Fram (talk) 13:32, 11 May 2011 (UTC)

I would agree that getting accurate information on these things is difficult. This is an encyclopædia; if there's a high probability that a page is wrong, due to confusion over names &c, then why create the entry? If all we have is the name and coordinates &c, and we're not even sure of those, I cannot fathom how keeping it is a step forward for encyclopædic quality bobrayner (talk) 13:35, 11 May 2011 (UTC)

I'll admit Fram has a strong argument about the stubs being in the wrong provinces and districts. Checking the up to date UN source [Khvoshi]] does not exist but the exact coordinates are that of Bala Deh which is not in Logar Province but in Gardez District in Paktia Province and if you view on a google map and UN you'll see this is accurate. I'm changing my vote to Delete all of Afghanistan entries which haven't been further expanded and sourced as many of them do appear to be in the wrong provinces and under different names and I certainly haven;'t the time to correct 2500 entries and research them myself. I think the best thing would be to recreate using the UN list as guidance and which detailed what district it is in. I have the district villages downloaded, they can be put into districts and templates and started again using more up to date source and a google book source if possible. As far as I can see Saudi Arabia, Oman, Yemen and Kazakhstan villages are in the correct provinces and haven't changed, even if transliterations may be awkward so I don't see a problem in keeping them, they just need expansion. I really haven't the time or energy to bother faffing around with a bunch of villages. I really don't care enough about them to argue against this. They are not important to my life to worth bothering with and wasting my time arguing here. Yes I think wikipedia would be better off expanding them all and fixing them all but nobody is likely going to do so given the scale of the task so it would be better in regards to Afghanistan to delete those unattended for and recreate at a later date using the UN source. If you do the rearch you'll find geographic names is most certainly a rleiable source for coordinates and for places which do actually exist, howeve,r in regards to some places like Afghanistan the provincial listing ar enow out of date. Its not that its an unreliable source. Its a dated source. Given that we now have the UN list for direction it would be better to recreate under the up to date sources and nicely organized by district templates. ♦ Dr. Blofeld 13:58, 11 May 2011 (UTC)

Thanks for this post. However, let me point out that not only Afghanistan articles are in the wrong province, I also gave the example of Al ahramiyah, which is given as being in Ad Dhahirah, but which is actually in Al Buraimi Governorate, which used to be a part of Ad Dhahirah. So there as well, the province or region information is outdated (and the article has to be moved to the correct capitalization as well). Fram (talk) 14:07, 11 May 2011 (UTC)

OK, please assess the other countries and those which are in the wrong province delete. What's ironic about this nomination is that I myself have tried very hard to get fallingrain blacklisted for containing false population and altitude data.. I believed geonames database and geogrpahic.org were at least valid for coordinates and provinces, although it seems that they have not updated it for some countries and have made some errors with actual provinces. Perhaps there ought to be a ban on using any geographic database solely to create articles. If Fram you can keen to get a ban on this I will hope you would extend it to Fallingrain and any other database created years ago. They are fine for coordinates but little else.. The problem with banning is that we are missing a massive amount of content which are real places and their cooridnates are accurately given in these databases even if they botch other things up from time to time. Maybe something should be written into the guidelines that people creating ge articles must provide a solid reliable source other than a database entry? I would have no problems with that but Fram I do wish you had addressed this subject with less belligerence and attitude towards me when I mostly happen to share your views on creating geo articles. ♦ Dr. Blofeld 14:10, 11 May 2011 (UTC)

I have no objection against blacklisting geographic.org, fallingrain.com, tageo.com, maplandia.com, and so on, as the incorrect use of these sources as seemingly reliable but in fact unreliable sources far outweighs their benefits. But I would like to stick to the articles at hand here, I believe that this is quite sufficient for one AfD... Fram (talk) 14:24, 11 May 2011 (UTC)

These websites are all generally accurate for coordinates, but little else. Its fine to extract coordinates from them and check on a google map but I agree it is not really a good idea to rely on them as a single source. Of course given the lack of editors working on desolate areas of Asia are mimimal anyway so the "sub stubs" would take a long time to all be expanded. OK fine, extend this to the Tibetan villages which have not been expanded to the delete pile, they are a much worse mess than these because they contain falling rain data and should have been nuked long ago. Its best to work with what we have in google books and try to build the best coverage we can piece by piece. What resorted me to generate sub stubs was frustration at the sheer amount of missing articles and a task which seems mammoth.♦ Dr. Blofeld 14:29, 11 May 2011 (UTC)

I'd agree that fallingrain-based articles suffer from broadly the same kind of problem - I had to do a lot of cleanup in the past after somebody used fallingrain inappropriately (but in good faith; such a big database is very tempting to people editing articles on settlements).

Mapping is difficult in a lot of developing countries; it's an inescapable problem. Last time I planned a long drive in Africa, the only thing you could truly trust were your eyeballs, looking at a satellite image - the overlay of settlements & roads from geographic databases would often be wildly inaccurate (even the hallowed Michelin maps have flaws). Even if a settlement is visible on a satellite image, you can't always be sure that the overlaid name, pulled from some database, is accurate. I would love to fill these gaps - wikipedia is systematically biased against these parts of the world - but we can't really fill the gap until we can get better sources. It would also help to have editors with more local knowledge / language skills, who would recognise dubious demonyms. bobrayner (talk) 14:46, 11 May 2011 (UTC)

Exactly, and this is the dilemna I am always faced with and the reason why I have ever created stubs on such places enmasse. The ignorance of many countries on here where content missing is in the vast majority like Iran and it sickens me that the amount missing is so vast. And then theres the problem in finding good sources even for major Iranian towns let alone villages and you'll find it ends up being a desperate sub stub attempt to at least get a mention of them. Google books is gradually improving, I've proved villages in rural Afghanistan and Burma can be expanded using snippets in google books but it is very uneven at present, very many settlements are not picked up in google books and contain nothing on the web but a computer generated database from 1995. But probably best to go where the information flows, mention in historical gazeteers and history books is a start.♦ Dr. Blofeld 14:51, 11 May 2011 (UTC)

OK, heres the plan. See User:The Anome/Villages in Afghanistan index. This was drawn up using the most up to date Afghanistan settlement directory by the UN and which are featured on their official maps. We use this list to create nav boxes by district and list the settlements in the articles and in the templates. We apply them and only start those articles which one can find at least one good source for and not in sub stub fashion. We put the existing unattended articles into incubation until they can be corrected and sources found. Other than this we find a bot which can read off of this UN index and correct any article given in the wrong province/district and build upon what we have started gradually...♦ Dr. Blofeld 15:21, 11 May 2011 (UTC)

• Delete all per nom. The proliferation of unsourced and ridiculously "sourced" mass-created substubs is a blight on Wikipedia that must be halted. The creation of an article should be a thoughtful exercise by a human being who is guided and informed by reliable sources, exercising human judgment and intelligence, not a database dump or a mindless robotic task (even if no bot is actually involved). WP:BEFORE is often used as a stick with which to beat folk nominating articles for deletion—where's the stick that's needed for people who create basically useless articles without sufficient forethought? Deor (talk) 16:24, 11 May 2011 (UTC)

What about a stick to beat moaning minnies who do nothing for wikipedia but turn up at AFDs and whine about sub stubs and the lousiness of their fellow editors instrad of doing something constructive to expand content they deem unworthy? Maybe there is no stick because of WP:AGF, which clearly you are not. ♦ Dr. Blofeld 16:37, 11 May 2011 (UTC)

Cool out, relax. Things like this work out. Trust me. AfD often provokes strong feelings on both sides, so some harsh words are inevitable sooner or later; but I'd invite both of you to stick to the subject at hand. Generalisations about editors, rather than articles, are likely to cause more problems than they solve. If you want to have a fruitless argument with somebody on the internet, take it to my talkpage instead, and let folk here try to agree on the best way to deal with these articles. bobrayner (talk) 17:02, 11 May 2011 (UTC)

Assume Good Faith still applies, and large-scale efforts to at least mention parts of the world which are poorly covered and try to work towards uneven bias should be commended not attacked, even if the method use is far from ideal/problematic. All "useless sub stubs" can be expanded or corrected very easily. If everybody assumed good faith and that we all have the same goal here conversations would not get side tract or heated. ♦ Dr. Blofeld 17:15, 11 May 2011 (UTC)

Incubation

Would incubation be an acceptable compromise, if we can find some way to avoid overwhelming any existing incubated articles? (Which is surely technically possible; just a subcategory of Category:Articles in the Article Incubator could work).

• For those who have so far argued delete, incubation means that the contentious articles are removed from article-space (and noindexed).
• For those who have so far argued keep, incubation means that the contentious articles still exist, and any editor can improve / expand them, add other sources, tweak names &c and they can freely return to article-space after improvement.

Comments / complaints / criticism? bobrayner (talk) 13:24, 11 May 2011 (UTC)

No objection, as long as they are independently checked before returning to mainspace. Fram (talk) 13:27, 11 May 2011 (UTC)

The problem is that even after they are independently checked it is a minefield in regards to settlements of the same name and google books/web having a bit of info about the village but not knowing which village it is. I think the problem with Afghan settlements would continue beyond checking them.♦ Dr. Blofeld 13:32, 11 May 2011 (UTC)

Doctor B, that's something most here would agree with but doesn't read (to me, anyway) as an argument for keeping them. ^Bongo_matic 13:58, 11 May 2011 (UTC)

Incubation would be an acceptable outcome for me. --DThomsen8 (talk) 15:01, 11 May 2011 (UTC)

Incubation is also acceptable to me, subject to a time limit of a maximum 6 months. Mjroots (talk) 19:13, 11 May 2011 (UTC)

• Delete all - per WP:RS. To verify that these settlements exist, and thus that the articles are appropriate, each needs to be sourced to a reliable source. I'm not entirely opposed to incubation, but I think it would be better if we wrote articles on each village when sources appear. Anthem of joy (talk) 15:24, 11 May 2011 (UTC)

The thing is stubs like Bashanabad even if "useless" are clearly verifiable and visible and marked on google maps with a clearly visible settlement which looks notable and can at least be verified as existing but lack web coverage at present... Lack of book coverage or web sources for such villages most certainly is not a guideline for notability. Uneven development and bias is to blame. I stand by my belief that virtually all populated human settlements are encyclopedic. We have a real problem though with developing world countries and our encyclopedia is inherently biased towards anglo countries which have more sources♦ Dr. Blofeld 15:29, 11 May 2011 (UTC)

Isn't it original research to verify the existence of a village by looking it up on google maps ? Even if it isn't, how can you tell that google maps has correct information ? Search for Badom-Dara, and Google Maps bring up a settlement near to Bashanabad which doesn't seem to exist from the photographs. It seems pretty tenuous at best. Anthem of joy (talk) 15:46, 11 May 2011 (UTC)

Satellites don't lie (except in major political areas) and the coordinates given are generally accurate even if they cock up the provinces from time to time. Google earth does not lie about settlements existing down river from another settlement and it being framed by steep cliffs. Reliable sources back up my claim anyway that that basin contains steep cliffs, so not really OR. Observations from google earth backed up by general facts confirming them are fine. I use it all the time when creating articles like Pedreña, an invaluable tool. Besides up to date maps like this confirm what I have said. How can you tell than any source is accurate and has the correct information? Jeez, this is the problem with building an encyclopedia, sources often contradict each other.. A satellite map if anything is more useful that any source for verifying a place exists and you can see it with your plain eyes. ♦ Dr. Blofeld 16:02, 11 May 2011 (UTC)

Interestingly I was perusing an old map of the area and spotted Kotgaz Glacier. A google search picks up reliable sources and also this entry in geographic I think the website is good for coordinates and finding missing articles , but as Fram says they need to be verified and more info added from reliable sources..♦ Dr. Blofeld 16:16, 11 May 2011 (UTC)

(replying to Dr Blofeld...) I'd agree that satellites don't lie; we know that a village exists at those coordinates. However, we can't even be 100% certain that "Bashanabad" is the correct name for it (unless a better source says so), let alone that it passes the WP:GNG because, as far as I can tell, no source has been provided that covers it in any depth at all, only an entry from an indiscriminate listing. bobrayner (talk) 16:56, 11 May 2011 (UTC)

• Just to get back to incubation for a moment, I realise I was the person who proposed mass incubation of this material, but I do want to explain in more detail why I think it's the best outcome. As encyclopaedia editors, it's a sin for us to publish known error, and Fram presents convincing evidence that this material must be considered unreliable. Normally this would be dealt with by editing the material in question during the course of the AfD, but here, we're looking at 2,500 articles:- a Herculean task to edit and fix. We can't accomplish that in the time available. The last time we had a really huge mass nomination, we insisted on WP:BEFORE and tried to deal with each one individually—remember the X-Y relations articles? That experience was bruising and we need to learn from it. This time, let's be smarter.

  However, policy says mass deletion of material that we haven't individually considered is to be avoided.

  Thus we have material that ethically, we can't keep in the mainspace, and procedurally, we can't delete. We need a compromise that takes them out of the mainspace but leaves the content available to repair and reinsert, merge, or delete as appropriate, all in our own time. I think that incubation is the option that best fits the bill.—S Marshall T/C 16:45, 11 May 2011 (UTC)

  If our existing procedures cannot cope well with a new situation, leading us to a result which conflicts with encyclopædic ideals, that's what IAR is for. However, nobody's going to get consensus for "Ignore all rules" as an outcome at a much-commented AfD. Pragmatically, I'd agree that incubation may be the best compromise. bobrayner (talk) 16:53, 11 May 2011 (UTC)

I agree, but the fact is though that the settlements listed in geographic.org and indeed other landmarks like Kotgaz Glacier do exist and the coordinates are 99.9% correct or at least very close, so in that respect it is not an unreliable source. Fram has agreed that most of these settlements do exist and the coordinates do reveal settlements but rather has identified serious flaws in the background for them such as provincial location, place names and transliterations and confusing duplicates which make it a poor and often flawed source to create articles.♦ Dr. Blofeld 16:49, 11 May 2011 (UTC)

• Delete all per the nominator's introduction: there are too many problems with these articles. In general, Wikipedia is not cleanup; however, there are too many potential problems with these articles to justify keeping. Like spammy articles about notable topics, these articles (which I'm not saying are spammy) would need complete rewrites to be valid. We need to get rid of errors, even though they were added in good faith: we can start over again on these from scratch, since that's the work we'd need to do to fix them even if we kept them. Nyttend (talk) 17:30, 11 May 2011 (UTC)

This means we would lose articles like Alishang, Sunaynah, Al Qabil and Jebel Ghawil because nobody can bother to attend to them . Very sad. Deleting them is probbably the worst thing we can do to ensure they woun't be recreated with errors again. Incubating them and chiecking them and moving into mainsapce once corrected and expanded would. I am happy for these to be incubated, deleted absolutrly not, that is more destructive than the actual minor errors of province in many of them. It would also eradicate many articles which are indeed notable and can easily be expanded but if you treat them all as problematic then you are hiding this fact. Each place is different and to nuke them all under a pretence that they are all non notable or problematic is just wrong. None of them have serious enough flaws to constitue nuking entirely. Each should be assessed individually over time in the work space, please.♦ Dr. Blofeld 17:53, 11 May 2011 (UTC)

Wouldn't wikipedia be more valuable if people take the time to work on these and try to add a bit like Sunaynah, Al Qabil, Mazim and As Subaykhi. Not every article has errors and those that do can be swiftly corrected. Look at Mazim on google earth its a great looking village and the sort of content I would like to read about on wikipedia.. Incubate them and they can be expanded and corrected if necessary and more sources added. Delete them all without any effort without even assessing them is a crazy idea.. ♦ Dr. Blofeld 18:16, 11 May 2011 (UTC)

No, it does not mean that we'd lose those articles. Please read the nominator's statement: only pages in Category:Articles only referenced to geographic.org are up for deletion, and articles such as Alishang are not in that category because they've been completely rewritten. The only articles that should be deleted are ones that haven't been reworked, since they have no useful information from reliable sources. Nyttend (talk) 02:41, 12 May 2011 (UTC)

• Incubate (or Delete) WP:NPLACE has possibly the lowest requirement for a subset of Wikipedia:Notability, "Cities and villages are generally kept, regardless of size, as long as their existence is verified through a reliable source." To allow articles that fall below that standard is potentially damaging to the project, and should not be allowed to be able to be viewed. If Geographic.org is deemed unreliable and is the only source for a subject, then it does not satisfy the criteria. I recognise that there are issues of Wikipedia:Systemic bias relating to the inbalance between the visibility of Western places of inhabitation and those of the rest of the world, so rather than delete I would prefer that articles based on that unreliable source should be placed in incubation and only moved to the mainspace when a better or confirming source is found. LessHeard vanU (talk) 20:15, 11 May 2011 (UTC)
• Delete all; stubs that have an error rate this high are considerably more harmful than not having a one-liner "article" on the fraction that has no immediately obvious error. — Coren ^(talk) 21:45, 11 May 2011 (UTC)

Sorry but I'd have to say that the error rate has been exaggerated considerably. The vast majority of Oman articles I've looked at so far are correct and most certainly have the correct coordinates obtained. One region though had split in 2006 so some just needed fixing. I'm currently ploughing through Dhofar Governoate and so far it seems there is absolutely nothing wrong with the named, province and coordinates. I think we should assess those which actually are in the wrong region first before incubating them, the unproblematic ones should not be removed and should be expanded. It is just wrong to label every single one of them as incorrect.♦ Dr. Blofeld 22:14, 11 May 2011 (UTC)

• Incubate so these can be handled on a case-by-case basis. Some of the articles simply need a reference to the more reliable GEOnet Names Server (the source Geographic.org supposedly draws from), quite a few of them need to be moved from variant names or names with the wrong capitalization/diacritics but are otherwise correct, some have the right name but the wrong province/coordinates, and some need to be deleted outright, but there's no good way to tell these apart without checking all of them. A reliable source exists for verifying the existence or lack thereof of all of these, so we should use it instead of deleting reparable articles. TheCatalyst31 ^{Reaction•Creation} 21:52, 11 May 2011 (UTC)
• Wrong forum We should not be assessing several thousand articles at AfD. Instead, each article should be individually reviewed. If a group of articles is verified to not exist or something like that, then batch-AfD those. But not all 2,500 at once, please. I suggest a speedy close of this AfD and a project page be set up so users can review each of the 2,500+ articles individually, and separate the good from the bad. /ƒETCHCOMMS/ 02:23, 12 May 2011 (UTC)

You may be correct that this is the wrong forum, but you are completely wrong that there is no reason to batch delete or incubate these articles, when their creation was a violation of policy. What forum do you suggest? ^Bongo_matic 03:20, 12 May 2011 (UTC)

I never said I opposed batch incubation—which is essentially what I suggested—but I don't think deletion is really anything except a "quick-and-dirty" solution. We have plenty of time, there's no deadline, we can sort through all these articles. I'm not sure which forum is best for this because it's not a common issue that arises, but I think that just setting up a project page dedicated to sorting out this mess or just a subpage Wikipedia talk:Article Incubator would work better as a centralized place where we can work through these articles and discuss what sort of sourcing is necessary or which articles are about nonexistent subjects, etc. But AfD is too narrow a place for this, and after the discussion ends in a week, we'll have to discuss this elsewhere—so why not close the discussion before it gets even longer and have one single centralized discussion instead of ideas proposed at this AfD page and then re-proposed elsewhere or forgotten altogether? /ƒETCHCOMMS/ 04:22, 12 May 2011 (UTC)

This discussion can be copied across to wherever it is continued once it has been closed. To those saying "wrong venue", this seems to be as good a venue as any, as we are talking about removing articles from mainspace, whether temporarily or permanently. The existence of this discussion has been well publicised by Fram. Mjroots (talk) 05:47, 12 May 2011 (UTC)

I personally believe mass AfDs are often more disruptive or more of a problem than individually assessing articles, because we're grouping together all 2,500+ of these pages without verifying the integrity of each one. Just because a bad page is sourced to geographic.org does not mean a good page should be deleted along with it. /ƒETCHCOMMS/ 14:48, 12 May 2011 (UTC)

• Incubate all, with release into article space not permitted without second source and manual attention. ^Bongo_matic 03:27, 12 May 2011 (UTC)
• Comment - Wrong venue, this should go to ANI. Issue is much too big for case-by-case consideration at AfD, there needs to be some broader policy decision. Carrite (talk) 05:34, 12 May 2011 (UTC)
  • ANI is not the place where policy is discussed or decided though. After this AfD has ended, we can see whether new or revised policy is needed to deal with these things in the future, and the probable correct location for this is VPP (with notices at AN, CENT, and whatever other noticeboards are preferred). Please, if you want me to convince that this is the wrong venue, don't suggest one that is clearly less suited for a discussion about the deletion of articles (which this is, after all). And I have put notices at AN and at the Village Pump at the start of this discussion, just because I wanted to have sufficient input on this: this is not an AfD that should be decided by two or three AfD regulars, but by a much wider group of people (which seems to be happening) Fram (talk) 07:05, 12 May 2011 (UTC)

Comment: I have obviously no objection against people deleting the pages (g7) where no obvious improvement can be made, or against people truly improving articles and thereby removing them from the list for deletion, but just removing the source and adding some unsourced info is not really an acceptable way of preventing deletion. Keeping the source but removing the tag for deletion isn't acceptable either. And creating a new unsourced article and then redirecting one of the articles under discussion? Other examples: [4][5]. Perhaps these things are actual improvements, perhaps not, but turning them into unsourced stubs instead of badly sourced stubs is not really the way to go here. — Preceding unsigned comment added by Fram (talk • contribs)

Jeez, will you lay off the stalking/hounding? I didn't have time last night to complete my work. I was working on Al Ayn last night but I found it referred to something else so redirected. I've db authored the ones I couldn't find info for except those which clearly look notable on google maps and likely are notable but we have poor coverage on the web at present for obvious reasons. Of course articles need sources and I'm trying my best to help find them, so please drop your "unacceptable" bullshit and highlighting of only "negative" flaws in my work rather than the good work I did last night.. ♦ Dr. Blofeld 10:37, 12 May 2011 (UTC)

Hang on a minute please

I've recently generated a list of Afghan villages and towns, together with their region and district data, sourced from United Nations OCHA data. See User:The Anome/Villages in Afghanistan index. I've been asked by Dr. Blofeld to check these against the articles that are the subject of this AfD. I can also scour the category tree for other articles created by other sources, and look for duplicates.

Since this is very much a spare time activity for me, and there may be quite a bit of programming and testing involved (for example, working on good string matching for different transliterations of Arabic place names is non-trivial, and I would appreciate any advice other editors could give me), would the other participants here be willing to hold off on the AfD process for these articles for day or two while I do the analysis, and correct any errors I find in these articles, using the UN data and cross-correlation with existing articles as a reference?

Any articles which could then neither be corrected nor confirmed from the UNOCHA data could then be deleted. -- The Anome (talk) 08:10, 12 May 2011 (UTC)

The AfD will normally run for nearly six more days, so the one or two days you think you may need are no problem. Any article no longer in the category at the end of the AfD will not be deleted (assuming that the AfD ends in delete of course). Fram (talk) 11:39, 12 May 2011 (UTC)

Extension, please I am in favor of an extension in time, allowing editors to rescue particular articles. Administrators, take note. Editors, join me in asking for an extension of time for this mass AfD, please.--DThomsen8 (talk) 15:42, 12 May 2011 (UTC)

I must confess

I'm going to be honest with you. I've mistakenly used geographic.org to cite the source thinking it was GeoNames which you apparently consider acceptable. It twigged earlier when I saw Fram accept an article referenced to Geonames somebody had switched on a Saudi Arabian village and I thought, hang on, aren't they the same? Every Afghan village I started is listed in the directory here which Fritzpoll downloaded in 2008 from the GeoNames server. This fact can be verified by contacting Fritzpoll by email, as he is no longer with us. So you've been ranting on about geographic.org being unreliable and that was not even the site used to compile the list even though it apparently mirrors entries of the geonames... The reason I had to cite a source was because people would moan about unsourced stubs and I genuinely believed that geographic.org was the correct source... The list by province was downloaded directly from GeoNames NOT geographic.org. So this apparent " plague of problems" with the provinces is related directly to GeoNames itself. As for Oman, Kazakhstan and Saudi Arabia I again mistakingly believed geographic.org to be Geonames and anticipated no problems with it being considered unreliable. I simply confused "Geographic names" with "Geonames". I'm sorry. I believed that the source was accurate and would be accepted by the community, given that I know people would kick a fuss up about unreferenced stubs but I got my wires crossed with using it, thinking it valid. It was just intended as an initial seal of verification and then the articles could be built up gradually as more sources become available. So if this mass AFD is to eliminate all entries then basically it is saying we should delete every single article from the geonames server and blacklist Geonames from wikipedia. I know articles really need additional sources but I was operating in good faith thinking geonames would be accepted to verify them,. A bot could be used to simply correct the sources to geonames rather than geographic.org. But any apparent problems with the articles are directly related to the geonames which the nominator has cited as a reliable source (and has cited for verifying other settlements, thus considering them like these to be accurate) which I also believed was reliable and acceptable. The coordinates of settlements equally are derived from the Geonames server and have been shown to be almost 100% correct. I still believe geonames to be a reliable source at least for verification and cooridnates even if they have shown with Afghanistan to have dated information on provincial location. This AFD now comes down to claiming National Geospatial-Intelligence Agency to be unreliable. ♦ Dr. Blofeld 11:54, 12 May 2011 (UTC)

I can't check whether this is true or not, but the articles created certainly don't match with the current geonames database. It certainly contains a lot less errors, looking for the really stupid entries in geographic.org doesn't give any results. It also doesn't seem to contain the errors wrt provinces that are in the articles you created. On the other hand, it does contain "populated places" of such extreme minor importance that we wouldn't bother creating separate articles for them in most countries, e.g. for Belgium they would at most warrant a redirect, if even that. Fram (talk) 12:34, 12 May 2011 (UTC)

I swear I'm telling the truth, it was downloaded in 2008, ask Fritzpoll, he even generated a few articles using it I believe in Badakshan province. Proof is Arun, Badakhshan etc. But given that some of the provinces changed before then it seems odd that the obvious errors existed then. I gather they spotted the error since and have updated it. I only just realised what had happened when I saw you accept that Saudi Arabian village source. Honestly I think you'll find the vast majority of the these 2500 settlements in geonames server and for some of them a scraps at least in google books. Perhaps a bot could be used to change the source. I still believe National Geospatial-Intelligence Agency to be a reliable source for verifying these settlements but I agree the one liners are crappy and need expansion if possible but I would NOT have started them if I believed them not to exist. I think this current situation could be sorted with a bot and some hard work over time to expand these if possible, IN regards to Afghanistan we can certainly get something done with Anome on board and organizing them and updating them. A bot could correct them, I'm certain.♦ Dr. Blofeld 12:46, 12 May 2011 (UTC)

Please no, don't try to correct these articles by bot, haven't you learned anything from the problems you have had correcting them by hand? look at all the troubles with Al `ayn, before ending with a redirect to an older article. Or all the trouble to get Dara-I-Pech right. Fram (talk) 12:54, 12 May 2011 (UTC)

No, User:The Anome/Villages in Afghanistan index is downloaded from the UN which clears marks up to date settlements. A bot could easily fix any wrong entries and if there are duplicates it would create a page of what needs to be sorted out manually. Those prior problems occurred from using an old source. There should not be a problem with this. But as Anome says he knows what he is doing and I'm sure knows the best way to deal with this. Al Ayn exists, in fact there are several in Oman.♦ Dr. Blofeld 13:04, 12 May 2011 (UTC)

Comment: extent of problem

Having gone through and re-sourced about 50 of the articles now, I have to say -- the problem, while it exists, is quite over-stated. Now, arguably, using a non-random sample of size n=50 would make even my intro to statistics teacher want to beat me with a herring. But in those 50-ish articles:

• 28 articles had to be moved (mostly for capitalization)
• 2 were not villages as they claimed to be (1 was a neighborhood, 1 was an "other")
• 1 did not have any record in the geonames db
• 2 were too ambiguous to be matched properly to a geonames db entry

90% of the articles checked were fine. (Of course, n=50, sample is not random.) There were only a few cases where I had to change the province name, and all but two of these were for Afghan village articles (which makes sense per Dr Blofeld mentioned above; the others were 1 Omani and 1 Saudi village.)

While it could be argued that the fact that close to 60% of the checked articles had to be moved is a bit excessive, it's certainly not cause for an incubation. And since 90% of the articles were accurate, a blanket delete is quite improper. Now those remaining 10%? Sure, delete them, hands down. But applying a blanket policy with such a disparity of circumstances is not appropriate I think. Just some food for thought for anybody still arguing one way or another. Cheers. --Mûĸĸâĸûĸâĸû _(blah?) 12:41, 12 May 2011 (UTC)

Wait, "had to be moved" means that the name, which is basically fully half the content of those articles, has an error. How is that "fine"? — Coren ^(talk) 12:52, 12 May 2011 (UTC)

Well, I'm sure you could leave it where it was. The page moves are generally just things like Al akhal → Al Akhal. Capitalization really, the sort of thing a bot could do, with supervision. (Or someone on AWB.) The most drastic change I've had to do is Ąaćbaną Pass → Kotal-e Khushk which is just a variant → official/standard name. --Mûĸĸâĸûĸâĸû _(blah?) 13:02, 12 May 2011 (UTC)

So, basically, just like I said from the start, 17 out of 50 one-line articles were correct, all the others had at least one serious error in their one single line or in their name. Coupled with the fact that all of them used an unreliable source and lacked all but the most basic info, I don't see how your analysis differs from mine. Your conclusion does, but that is why we have a discussion. Fram (talk) 12:47, 12 May 2011 (UTC)

The vast majority of the article, I mean at least 98% of these 2500 articles are also verifiable in Geonames. A bot could simply switch sources but hard work will be needed to expand the stubs into something worthwhile, much like the rest of wikipedia. They can virtually all be verified in Geonames and an obvious satellite zoom observation. Its a crappy start I agree but verifiable settlements should be notable, naturally rural villages in Asia are not going to be hotbeds of web information and I think over time they can be expanded.♦ Dr. Blofeld 12:48, 12 May 2011 (UTC)

A bot could not switch sources, most of the articles have no coordinates and no reliable province info, and the name alone is often used for many locations. Couipled with the fact that many articles have been created as duplicates of each other or of older articles, and that a number are not villages, and you don't end up with "at least 98%" but a lot less. Fram (talk) 12:59, 12 May 2011 (UTC)

• Keep Mass deletions, especially where a large proportion are correct, and many of the "incorrect" ones are spelling variants is a "bad idea." As for province changes - they occur, and I would not automatically assume that they qualify as "errors." AGF is key here - and it looks surely like assuming good faith will result in an optimal outcome, whilst deletion is a substantially sub-optimum outcome as existence of a location is not on the order of a contentious claim in a BLP by a few miles. Collect (talk) 12:55, 12 May 2011 (UTC)

Well the idea of course in creation was that they can be expanded like Alishang,Sunaynah, Al Qabil, Jebel Ghawil and Mazim etc which I think are valuable for wikipedia to have. I only created the stubs to try to work towards better coverage of the developing world which over time can be expanded. I most certainly did not intend for them to be problematic or a negative thing. And I think if all 2500 could be expanded like those above people would be very positive. The fact is though they are virtually all verifiable settlements which are plainly more than just visible on google earth and we should have articles on settlements, Hard work is needed to expand them though but given the current web sources are unlikely for many places but that desn't make them non notable. Any errors I don't think are serious enough that a bot couldn't quite easily fix it and we move on and build wikipedia properly....♦ Dr. Blofeld 12:59, 12 May 2011 (UTC)

And what is the benefit of "expanding" them compared to starting from scratch. Take Jebel Ghawil, which you used as an example: there is nothing left from the original article[6], it has been moved three times in a row, and it turns out to be a mountain with perhaps, maybe a village of the same name (note that Geonames has never heard of Jebel Ghawil). I am not arguing that these locations shouldn't have an article, I am arguing that deleting them and starting again from scratch, from a good, more reliable list of locations, would be much more efficient and let less errors hang around. Fram (talk) 13:05, 12 May 2011 (UTC)

What's the point? Abu shaytanah for instance can easily be moved to Abū Shayţānah, and the source changed to this. Is that such a major problem? Zoom in on it on google maps. These are legit settlements but just require some repetitive edits to at least provide a sold base to build upon. A bot really is the best solution to sort these out and those which are duplicates are easily spotted and listed. Your/the real problem is not geographic.org, given that you know these villages are almost entirely all verifiable on geonames and exist. Your real problem is opposition to auto-generated sub stubs generated using a database which is why you posted on ANI and did not get the response there you were looking for so you came here to try and force it to get your own way.♦ Dr. Blofeld 13:17, 12 May 2011 (UTC)

What's the point indeed. Instead of the seemingly random selection of hamlets, villages, and duplicates of major cities that you created, and which you now want to correct by bot, why not start from scratch and let the bot create articles from that list, if it is indeed reliable? You will get a much more complete coverage, with a lot less errors, and a lot less effort, than you get with the current haphazard creation - match - move -correct cycle you are proposing. Why are you so reluctant to just start this all over to get a much better result? Of course you'll need bot approval, but then again you should have gone for that the first time around as well. Fram (talk) 13:25, 12 May 2011 (UTC)

(e/c) Well, I have no problem with automated stub generation (even though I often feel it's unnecessary). The only bit that bugs me right now is the accuracy rate. Sure, we can walk through the 2500-odd stubs that run created manually to check against some other database and fix titles, but how is that a gain from writing those articles by hand from a human-verified source in the first place?

Certainly, you're not suggesting that we leave any of those stubs without a human having checked them for accuracy? — Coren ^(talk) 13:26, 12 May 2011 (UTC)

Absolutely not. I'm proposing that a bot corrects the obvious errors, moves pages to capital letters and then I propose that they are checked and gradually improved with reliable sources as those I've expanded above. My point Corne is that you are claiming articles which are mostly all like Abu shaytanah to not be accurate, which is false. It simply needs a capitalization. The coordinates and location info are correct. At least 90% of these articles are accurate and require simple procedures such as moving page name and a ref switch. The whole point of creating the articles is so that gradually they are edited by humans and nurtured into valuable articles. If nobody chooses to work on Asian villages with me and is happy for wikipedia to blindly ignore half of the world's population then its not my faoult. I agree it is better to create articles with some sources and more info and check each one carefully but the fact is at least 98% of the articles I started are most certainly verifiable as populated human settlements which given time are likely to have sources available for them.♦ Dr. Blofeld 13:30, 12 May 2011 (UTC)

I still think you'd be much better off — and that your time would be much better spent — by starting from scratch and creating those articles one at a time. Surely, it's less effort to walk the (better) database and create a stub than it is to pick a current stub, find its correct match in the good database, and then alter the article to match? (It's obvious you can't just move the article blindly to title caps, you have to check its very existence against the good DB first). — Coren ^(talk) 13:40, 12 May 2011 (UTC)

The reason why I wouldn't want to nuke them all and start again? Aside from being a waste of time, Fram and co are now so strongly against stubs on villages created enmasse that they block people indefinately for it and he would use every excuse as to why I should not create them. Of course if I recreated everyone with 30 minutes work and sources (not that that's possible) then he couldn't complain but the likelihood of that is slim; I do not have the time to spend on that amount of articles and wikipedia should be a shared project in which we all work together to better articles. Of course I prefer researching articles and writing them if possible, but I do not have the time to write every article on wikipedia.. Might I point out that if Fritzpoll had been permitted to run his bot back in 2008 creating articles on world settlements (directly verified) this would never have happened and by now the stubs could at least start to expanded. ♦ Dr. Blofeld 13:30, 12 May 2011 (UTC)

Dr. Blofeld, please don't accuse me of things I haven't done, and am not planning to do either. I have blocked (and unblocked) one editor whio continued to create articles against a clear consensus at WP:ANI, a discussion with which I had no involvement whatsoever. You did follow me to that user's talk page, even though you had had nothing to do with that discussion either, and now you misuse this here. If there is a consensus for a mass creation of articles (eg. by bot), and if they are tolerably corrct in general, then I have no opposition against that. Claiming that you don't want to delete these articles because I would stop their recreation is beyond the pale. (and can you please try to use less edits to make your point? You are constantly causing edit conflicts) Fram (talk) 13:54, 12 May 2011 (UTC)

OK then so if after this is fixed I go on and create say 38,000 articles on Burmese villages using geonames as a source and where possible a google book source you would support me? ♦ Dr. Blofeld 17:31, 12 May 2011 (UTC)

(ec/4) Well, we'll have to agree to disagree on this then. I remain unconvinced that you are, in fact, saving any time by doing it this way: you still have to manually check and correct all of those stubs, and in that interval they remain visible to be disseminated around with inaccuracies. — Coren ^(talk) 13:55, 12 May 2011 (UTC)

Ye have little faith in me Coren. You do not need to delete an article entirely from wikipedia to recreate it have you know...♦ Dr. Blofeld 17:15, 12 May 2011 (UTC)

I argued strongly back in 2008 that Fritzpollbot could be used to generate articles on settlements by country verified by geonames directly and where possible the plan was for it to find population data where possible and try to produce a decent starter stubs which could be built upon. The reason being was that I know humans tend to cock things up in large scale operations and are very inconsistent from editor to editor. The reason was precisely because I did not want errors occurring as with what happened here and people creating poor quality stubs on Indian villages without coordinates or anything. There was considerable support for this, however because of the fact that 1/3 of the editors protested against auto-generated "useless sub stubs" and because Fritzpoll was such a good guy who wanted to please everybody he dropped the proposal. I couldn't agree more that articles do need human nurture and to be written properly but given that I consider all populated settlements certainly above 50 people to be notable and i think this has been proved in many places then I think the best way they can be tackled, and this goes for any geographical landmark is to be started using geonames and a few other reliable sources if possible. We had a plan almost 3 years ago to the day in which Fritz would do trial runs and it to be checked for errors and then to run it. By now we could have articles on most geographical features in the world and I could be spending more time expanding them rather than creating them. If another bot was proposed and authorised I would happily nuke all of these and get it to do it properly with more sources and data if possible. Like Tanzanian settlements all have data in the official national statistics website. A bot could be used to extract coordinates form geonames and compile info. I know that it is best for humans to work on each article individually and create them like that but the massive amount missing on what are notable places makes the task too big manually. If there was renewed support for a bot to create articles on settlements and geo features such as mountians, lakes and rivers from geonames and if possible using other sources to create new articles then I'd happily start again. But given the view of many it seems they are completely against anything auto generated and would oppose it. Unless something is done with world settlements people will continue to make desperate poor quality stub runs again and again. It is most certainly unfeasible for any normal editor to want to create so many articles manually with lots of research. ♦ Dr. Blofeld 14:13, 12 May 2011 (UTC)

This is somewhat tangential from the current AfD, but personally, I would be very happy with automated creation of location articles from a single database, if that database was better - more reliable and more detailed. However, if we want more than merely coordinates and a (not very reliable) demonym, the most likely sources are databases owned by national governments, and in each country, different details will be recorded. We already seem to have a lot of automatically-generated stubs for some developed countries, which have pulled more detailed information from national statistics bodies - they have population stats, land area, membership of various administrative divisions &c. Although many developing countries have weaker government institutions, at least some expose similar databases which we could use to create an adequate stub. Even something like the generic USA CDP articles - dominated by 200 words of barely-readable demographic statistics - would be an improvement on what geographic.org gives us. (Personally, I disagree that every settlement is automatically notable, but that's an argument for another day and I realise a lot of people disagree with me). bobrayner (talk) 14:55, 12 May 2011 (UTC)

The best reason to include articles about some obscure places is when there is a link from another article, XXX was born there, the bridge from a big place to YYY ends there, and that kind of thing. I created a CDP place in Pennsylvania because a bridge went there. My experience with generated articles done by another Pennsylvania editor was that much could be done with such articles in a sandbox, before making them mainspace articles. In this particular matter, I have mixed feelings. Is the category getting smaller as name change and other editing is done on articles? (Right now, 2435 articles in the category.) What about addition of talk page templates? (I have been doing that for Oman.)--DThomsen8 (talk) 15:34, 12 May 2011 (UTC)

Currently 2224 articles in the category, down 211. Some editors must be doing some good work. --DThomsen8 (talk) 01:22, 13 May 2011 (UTC)

I agree Bob, that is the idea. A geobot generates articles consistently using goenames for coordinates/maps and a government source if possible for population data. There are many countries which have available data and they could easily be created with some useful info which is useful as a start. If automated creation is strictly tested and checked prior to creation it is an excellent way of creating content consistentally without the errors and horrible inconsistencies we seen in articles and provide a start to build on. The problem of course is editors lacking to develop them further. Rambot whilst performing a much needed task and much need info it showed some disappointing results in subsequent article development and the articles are mostly unloved and frustratingly generic at times. 8 years down the line and most of them barely have anything of substance added to them apart from the detailed demographic data, which is now 10 years out of date. So that's my concern about automated creation. But the fact is articles will continue to be stubbed poorly in the future by all sorts of people as shown with the recent Indian villages. If they had been created with a bot and some data then the editor could be expanding them.. Personally I would rather articles existed and were generic rather than missing or created haphazardly with errors and inconstencies. It is inevitable that eventually wikipedia will be covering most settlements so I think we should be planning the best way to go about it and stop people form having to create lousy stubs with errors. ♦ Dr. Blofeld 15:44, 12 May 2011 (UTC)

• keep I am not sure where to state my opinion here! The stubs are not against policy, even if they are not helpful when they are wrong. Therefore I would ask that they be fixed instead of creating more and more at this point. The statements on how to improve are already covered above: correct, redirect, proper source. Instead there could be a topic ban limiting the rate of creation, say to ten per day so as to make sure that they can be checked and improved. Any mass create of multiple articles per minute should get a bot approval, even if it is done by a human. Graeme Bartlett (talk) 21:24, 12 May 2011 (UTC)

Well naturally articles like Taldysay would best be created using google book sources, but we have the stubs now so they must be nurtured into something worthy. But if there is to be concern about manual editing and errors and an insistence on bot approvals I strongly urge that a GeoBot is created which generates articles on geo features to stop this sort of thing happening again. It would alos mean that I wouldn't have to resort to "sub stubs" and could concentrate on building up content. ♦ Dr. Blofeld 21:43, 12 May 2011 (UTC)

Keep all - Notability is more important than anything else. There's nothing wrong with stubs, and the idea is to start from somewhere in order to make a more comprehensive encyclopedia. There are many editors on WP, and each day many of them work on different existing articles. That's what I think this project is all about. If there's a problem with something, it's better to just be bold and expand them, source them, and improve them, because after all they deserve to have their own entries on Wikipedia. Fram himself said "a lot of articles are correct", so how can all of them be proposed for deletion? If anything, this proposal to blindly delete them all despite the fact that many of them are worthy, is puzzling, and IMO, simply wrong. Shahid • ^Talk2me 22:11, 12 May 2011 (UTC)

• Keep all. If they can be fixed, then that should be done. If most articles come from GeoNames, which is reliable, then just look at the remaining ones. Give him time to work. Decide which sources are reliable, and get the information from there. If you have a bot that does that automatically, it shouldn't be a problem. Dream Focus 00:16, 13 May 2011 (UTC)
  • Dr. Blofeld claims that these articles come from Geonames, but no evidence for this has been provided, and it seems dubious. The articles match the geographic.org source more closely than the geonames source, and the errors pointed out in the nomination are not in the geonames database. At best, he used a very outdated version of Geonames, but I don't buy it. Fram (talk) 06:54, 13 May 2011 (UTC)

So you're calling me a liar? Email Fritzpoll he will verify this and make your lousy invalid AFD massacre look like the pathetic desperate act of revenge it really is because you didn't get your own way at ANI. You're the one with the attitude problem, not me. If you just assumed good faith and worked with people you'd get on a lot better on wikipedia. I've given you the chance to quit being like this and actually work constructively but you persist with remarks like this.♦ Dr. Blofeld 08:24, 13 May 2011 (UTC)

I did not call you a liar, it seems to me that you are mistaken. Please keep your personal attacks to yourself. Fram (talk) 08:49, 13 May 2011 (UTC)

• Keep While I don't think mass creations like this are really the best idea, a mass deletion will probably cause more harm than good. Probably better to take a bit more time and sort through them individually. Qrsdogg (talk) 02:09, 13 May 2011 (UTC)

• Keep, per Mukkakukaku's research & The Anome's approach of using material from the OCHA database. Yes, there are probably a lot of entries here that might need to be deleted, but I doubt it is as many as Fram believes it to be. A blanket delete of all of these articles will cause more problems than it might fix. This kind of response to Dr. Blofeld serves as a disincentive to other people like myself who also create articles on geographic locales -- even if I take a more conservative approach to the problem. -- llywrch (talk) 06:17, 13 May 2011 (UTC)

It's the kind of response he gets when he vehemently refuses to believe that an article is incorrect, even when it is pointed out to him repeatedly, forcing an unnecesary AfD to get him to realise that whatever I stated earlier and the solution I proposed, a redirect, are correct after all (the whole Dara-I-Pech debacle). Note how in this very AfD, his story has changed three times or so, and his opinion on what should happen, what was the cause of the problems, and so on changes constantly. Despite an ANi discussion, a RSN discussion, a previous AFD, discussion at DGGs talk page, and my tagging of all these articles, he still didn't realise that he indicated the wrong source (geographic instead of geonames), and that his list was seriously outdated (not to mention absolutely random in its choices of what articles to create, with many tiny hamlets created, and much larger villages ignored). The problem is not the response to Dr. Blofeld, but the response from Dr. Blofeld. And I haven't yet seen how Anome is hoping to automatically correct these pages. Fram (talk) 06:54, 13 May 2011 (UTC)

Above you've basically called me a liar. You are the one with the attitude problem. Fritzpoll Bot generated the list from Geonames . He began creating them like this using the same source. Don't start calling me a liar because I've revealed your pathetic mass AFD to be a joke given that geographic.org was not even the source used.♦ Dr. Blofeld 08:25, 13 May 2011 (UTC)

No, I didn't call you a liar, I said that a claim you made seemed dubious to me. You can make mistakes without being a liar. Apart from that, I have no reason to assume that Friitzpoll did anything else than what you claim, but this not about whatever Fritzpoll did, this is about what you did. Fritzpoll created pages like Wikipedia:WikiProject Missing encyclopedic articles/Places/afghanistan/page1. You created User:Blofeld of SPECTRE/Missing Settlements/Afghanistan. A page like Bona see Buna doesn't appear in Fritzpoll's pages, but it does appear in your page. So please stop with the arguments about whatever Fritzpoll did, and accept your responsability for all this. Fram (talk) 08:49, 13 May 2011 (UTC)

I have not used those lists in over three years. Fritz downloaded the full list into the workspace directly from Geonames. I'll ask him to email you and confirm this. I am accepting responsibility for this which is why I've been working diligently to fix the problems while you sit around moaning.♦ Dr. Blofeld 09:19, 13 May 2011 (UTC)

Another personal attack, after I corrected errors you introduced at Al Hawiyah and Al Khaydar and removed your incorrect and duplicate Al `Utaybiyah from sight? Please, you are not helping yourself with these posts... Fram (talk) 09:30, 13 May 2011 (UTC)

You bother because you are a WP:TROLL who stalks the edits of the others with the intent to identify only their errors. If you genuinely were operating under good faith you'd be working with me to fix the others. I am refraining from saying what I really think of you as I'd be blocked.♦ Dr. Blofeld 09:45, 13 May 2011 (UTC)

Why would I go around fixing pages I propose should be deleted and recreated from a good, more complete, and more error-free list instead? That's not logical at all. I only fix those pages where you (or others) have avoided the deletion, but have left small (coordinates) or large (duplicate article on a neighborhood, claiming it is a village) errors. And considering the number of attacks I have received from you (here and earlier), and the lack of cooperation since the beginning of this story (where I had to demonstrate every error you made repeatedly and exhaustively before you acknowledged that there might be something to it, instead of, you know, assuming good faith and checking immediately and with an open mind whether my remarks were perhaps correct). Anyway, if you really don't want to get blocked, it may be better if you not only stop making new personal attacks, but also strike all the previous ones in this AfD. Fram (talk) 09:53, 13 May 2011 (UTC)

If you were genuinely a good person and keen to improve content you'd stop making snide remarks implying I was a liar and "dubious" and start helping me out with the others and refrain from making such remarks here. Somebody acting in good faith does NOT rant on at ANI about the edits of others and does not open a mass AFD of 2500 articles without assessing each one first. Somebody acting in good faith would have contacted me on my talk page, pointed out that there were errors in my work and have kindly asked me to fix them and expand them. ♦ Dr. Blofeld 10:01, 13 May 2011 (UTC)

Dr Blofeld, please try to concentrate on the articles and not the editors. Many people have a sudden rise in blood pressure when their articles are nominated for deletion - that's understandable - but the best way forward is to work on a solution to the problem, rather than sniping at others. You might not agree with Fram (or me) but Fram surely has the encyclopædia's best interests at heart and they have put in hours of work looking for problems in article-space and trying to fix them on your behalf - which is, oddly, the outcome that you appear to want from this AfD (although presenting a fait accompli is not a valid justification for a "keep" vote). Fram has also patiently listened to your varying (and occasionally contradictory) explanations. If you still want to argue with people, take it to my talkpage (I'm a thoroughly disagreeable person so I'm used to it) and let other people here discuss the real problem in peace. Also, it would be very helpful if you could put a little more effort into indenting, as it's getting quite hard for other editors to follow this long thread. bobrayner (talk) 11:32, 13 May 2011 (UTC)

And I suggest you also put more effort into building an encyclopedia rather than wasting your own time making biased overviews on AFDs. I will agree not to comment here anymore if Fram agrees not to comment anymore and we let other people discuss it. If I had been approached on my talk page and Fram had suggested I go through these and sort them then I would have done so with the other countries and would probably have db authored the Afghan lot and started again using the update to date. Now the scenario is we are faced with an open AFD and people passing judgements with "keep" or "delete" when they are not the ones who've created the articles or know exactly what can be done to ensure they are all fixed and improved. It should have been sorted between ourselves and in a spirit of collaboration. If you want sweet comments from Blofeld don't mass nominate 2500 articles without even notifying me first. ♦ Dr. Blofeld 12:03, 13 May 2011 (UTC)

• Delete all, or incubate with a time limit of 6 months. I don't like the idea of incorrect one-liners sticking around longer than that, even in incubator space. As bobrayner said, "This project is not a race to create as many zero-quality placeholder articles as possible from an unreliable list, in the hope that somebody else will come along and clean up the mess later." --Fang Aili ^talk 08:01, 13 May 2011 (UTC)