Wikipedia talk:WikiProject Geography/Bot/Archive

This is past discussion on a project which is no longer operating.

Gettings things moving

Now that there is an established base of operations for this project, we should tie up some of the loose ends. Links should be redirected here, and discussion centralized. Wikipedia:WikiProject_Missing_encyclopedic_articles/Places should also be integrated into this project space somehow. What country are we working on? How can we tell what places an editor has checked in a list? --NickPenguin_(contribs) 22:30, 15 June 2008 (UTC)

Apparently they old lists need remaking to fit the latest proposal. First will be American Samoa. The reason I didn't set up the project in Missing articles is because the aim of the group is more than just missing articles, it is to clean up existing articles and hopefully use the bot. My concern is that the project is dependent on one single editor to put it into action and if he is extremely busy in real life stuff then it will be difficult. What I'd hate to see is the project fizzle out like happens with so many other projects. Ideally there should be several editors on board who could run the bot not just Fritz. This would take some of the pressure off him. ♦Blofeld of SPECTRE♦ ^$1,000,000? 22:36, 15 June 2008 (UTC)

Unfortunately I don't have any bot experience, nor will I have the time to learn the technical expertise. Any other knowledgeable bot writers out there? --NickPenguin_(contribs) 23:44, 15 June 2008 (UTC)

Delete and move?

Looks like a lot of pages seem to be linking to User_talk:John_Carter/GEOBOT_group in regards to this project. Maybe we should consider deleting this page and moving that one here? That would consolidate most of the discussion here, at least. --NickPenguin_(contribs) 21:40, 16 June 2008 (UTC)

Redirected here:

Trial run of the bot?

Has the trial run already occurred, and, if it has, is there any listing of the articles it created which can help in the cleanup? John Carter (talk) 02:01, 11 June 2008 (UTC)

Cambodia

Don't know what order we're going in, but I've tracked down a decent number of sources for Cambodia. I think the best route might be to make combined district articles (commune articles could work too) with lots of redirects from village names. A model district article is at Ou Ya Dav District (home to around 11,000 people). Mangostar (talk) 07:52, 11 June 2008 (UTC)

Also just found an insanely comprehensive excel spreadsheet about various villages here. Mangostar (talk) 23:49, 28 June 2008 (UTC)

Notability and lists

Inspired by Blofeld and Mangostar, I suggest the following:

Only create "full" stubs on places with demonstrably notable places, per consensus achieved at WP:GEOBOT. For all other places, where the name and details can be verified in two sources of data, we include their names in a list organised by administrative district that includes any statistical info we can glean from sources, such as population, elevation, etc.

Thoughts? (Note that we need input from a few people on this) Fritzpoll (talk) 12:35, 11 June 2008 (UTC)

Would these lists be placed in articles on those administrative regions, or sequestered as subpages here? Ryan Reich (talk) 13:19, 16 June 2008 (UTC)

That would of course be a great idea as it would also substantially improve the district pages which for most countries are one liners. It would also ensure we have a comprehensive coverage of every district "del mundo". MangoStars example is great but I would also incude a geo globe column for location. ♦Blofeld of SPECTRE♦ ^$1,000,000? 15:50, 11 June 2008 (UTC)

The reason there is no geo globe column is that most of the villages have no location in NGIA. I may be able to match up a handful, but I figured I'd just stick it in the notes column (and if we got enough start a new column). Mangostar (talk) 17:57, 11 June 2008 (UTC)

Are you saying there are no coordinates known? Or just not easily harvested? Phlegm Rooster (talk) 19:01, 23 June 2008 (UTC)

I have not found a source that gives coordinates. The NGIA data is not that comprehensive w/r/t villages in Cambodia. Mangostar (talk) 22:00, 29 June 2008 (UTC)

If we decide, which I strongly think we should to create tables of all places anyway by district of each country, I think that the directory should still be generated. We may need to adapt it however if we are to trnasfer them into tables and find other info. Where we can I think we should create as many seperate articles as possible if valid enough but for those which at present it is difficult to obtain data, lists of tables would seem to most comprehensive system. After all people are here to learn, and if they have to visit 100 different pages to get the same data that could be listed in a neat table then our job isn't being done correctly. Some day in the future I can see websites dedicated to districts like there are in places like france and the UK etc all over the world and the days when we found it hard to find a population figure will be laughable, as by this time paragraphs of info on the history/culture of these places will gradually become avilable. It may be a while for some countries to gain access to the web but I think we can be sure that the givernment or elite of even poverty stricken countries will be wealthy eough to publish data online even in countries like Liberia, Guinea etc eventually. ♦Blofeld of SPECTRE♦ ^$1,000,000? 16:09, 11 June 2008 (UTC)

Tables are a good start, and notable places will be more obvious as more countries come online in the coming years. If we develop an adequate article framework, then when the information becomes more readily available, then the depth of coverage on a given country can be modified as needed. --NickPenguin_(contribs) 20:05, 11 June 2008 (UTC)

Possible procedure

Based on what I have seen so far, this seems to at least me to be maybe the best "standard procedure" for proceeding here, up to the point of actively engaging the bot.

• (1) Create one or more lists of all the places to be found in the sources cited on the Wikipedia:WikiProject Missing encyclopedic articles/Places page.
• (2) Separate out the lists into sections of those names which already have articles, those which can be verified as having enough notability (meaning, a nontrivial mention in a reliable source) to potentially have articles, and those which, at this point, don't have that level of notability. We should probably also include in that list some sort of acknowledgement of whether there is an existing disambiguation page or whether the page is currently a redirect, which would probably mean that there should be a disamgibuation page.
• (3) Check the various other existing articles on the subject to see if maybe there already is an extant article on the subject under a slightly different name. If that is the case, then move that entry to the "extant" section, and turn the alternate name to a redirect to the existing article.
• (4) Go through the various sources available, and indicate on the list page which are sourcable, and from which sources, making it a bit easier to expand the articles once created.
• (5) Ensure that all relevant lists, like List of mountains in Afghanistan, List of rivers in Afghanistan and so on, listing all the entries on the basic lists, are created.

The one major point of potential dispute isn't directly addressed in the above, and that is how to determine whether there is sufficient content on a given subject for an article, if created, to ever be likely to be more than a stub, and I personally have no way of knowing how to answer that. I went through a tour guide on Afghanistan yesterday, and found several locations which are, simply on the basis of that single book, clearly sufficiently well covered for a Start class article. Regretably, several other locations, including individual hotels and restaurants in Afghanistan, often have a single paragraph, and it is hard to know if there is every likely to be more there. I don't myself have an answer to this yet. My initial opinion would be to add all that content to a "Culture of Foo City or Area" section or article initially, and then later expand from there if sufficient interest to do so is displayed. Any other opinions? John Carter (talk) 13:27, 12 June 2008 (UTC)

Exactly what is going on?

If this were an article it should get tagged for omitting the introductory paragraph. I'd be interested in writing articles about the zillion and one missing cities and towns from mostly western Africa (primarily French West Africa), if the bot uploaded the basic templates. What exactly is to be checked by humans, just spelling? There probably are not a lot of spelling issues with large cities and towns in French West Africa without existing Wikipedia articles, smaller towns will be trickier. Also, are rivers and parks and reserves being done, as the question about Afghanistan rivers above? The missing African rivers are a sorrier site than the cities and towns. --Blechnic (talk) 08:42, 16 June 2008 (UTC)

It appears that this is still being worked out, but the procedure in the previous section looks like a plan. Point 2 is what you are looking for: humans will be adding the data from "nontrivial mentions" in whatever sources can be found, to the bot-generated template. The original proposal didn't include any mention of natural features, only inhabited places, but if they were being done, I'm sure the principle would be the same. Ryan Reich (talk) 13:16, 16 June 2008 (UTC) Correction: it seems we have a Scope now, which includes all these things. It is not clear to me either if full articles will be created, but I think the answer to that is, if they can be. Ryan Reich (talk) 13:25, 16 June 2008 (UTC)

We can't bot generate full new articles to begin with. We can do the best we can in getting them set up. It is up to editors to develop them fully later ♦Blofeld of SPECTRE♦ ^$1,000,000? 22:32, 16 June 2008 (UTC)

Which probably won't happen. But something is better than nothing. I'm an Editorofthewiki^{[citation needed]} 00:37, 27 June 2008 (UTC)

Cut'n'paste move

Hold on. There's just been a cut and paste move, breaking up the edit history. Why was the page not simply moved? I can fix this, but not tonight. It requires admin tools and some patience to put Humpty Dumpty back together again. Please don't move pages like this again, thanks. Geometry guy 22:56, 16 June 2008 (UTC)

I'm not too worried about the details, although I'm sure there's enough people around who are ;p I mostly just want to have some work laid out infront of me so we can start on getting things going. --NickPenguin_(contribs) 23:15, 16 June 2008 (UTC)

I know it seems fussy, but all contributions, even to this page, are licensed under GFDL. Also the history contains useful info. Anyway, many thanks to Andrwsc for a perfectly executed fix! Geometry guy 08:25, 17 June 2008 (UTC)

No worries. — Andrwsc (talk · contribs) 16:54, 17 June 2008 (UTC)

WP:GEOBOT

was just redirected by me from the poll/consensus/VP discussion, to this WP page. Revert me if it's too soon, but I think that's what the original idea was once this was up and running as an "official" WikiProject. Also, per the thread above, I belive the cut/paste v. move page has been repaired already. Let me know if admin tools are needed/repair was incomplete. Cheers, Keeper | 76 | Disclaimer 23:27, 16 June 2008 (UTC)

What's the process to "finish" a country?

The Procedure section on the project page is a little vague, so I haven't quite figured out exactly what needs to be done and how I can help. What do we need to do to verify item's on the list, so they can be generated by the bot? Please be specific. --NickPenguin_(contribs) 03:16, 20 June 2008 (UTC)

Creating lists

With the countries where we won't find much info we should generate tabled lists by province and create redirects to the lists. If we can't find much info, rather than skip the country, I think we should still create the lists. Really I think we ought to create a full set up lists whether there is enough info avilabale for a seperate article or not as it makes wikipedia more comprehesnive by area, and any existing articles can be accessed from the lists. This way we know we have a full base of world geography and that the encyclopedia recognizes the existence of all places until they can be started adequately.

I was thinking something like this with a globe also and any other useful columns we can find. The maps could be shrunk slightly I really think we should use pin maps as well as mini globes. It wouldn't be impossible to generate full lists by district or province of eahc country like this. Aside from this I think it would serve a useful purpose in that it contains the info that other editors create seperate sub stubs for seperate articles. If the exact same info was in the lists and red links were redirected I think editors would think twice about creating a seperate pages for sub stubs unless more info becomes avilabale. This could act as a sub-stub article controller! I for one know if we had something like this I could focus on creating some fuller articles and working on finding sources to make seperate articles possible

Plan

We start on Afghanistan. We consult WP:Afghanistan -see if anybody has any sources. We generate full lists by province anyway with coordinates and the basics into lists from the directory into the mainspace. This process should be repeated for each country. I think we should create a full world list anyway for comprehensive purposes and then make the decision which articles can be started independly in their own right through project consultation and research. ♦Blofeld of SPECTRE♦ ^$1,000,000? 19:30, 26 June 2008 (UTC)

• Stage 1 -we upload the directory into the missing article space as Fritz began doing before
• Stage 2 - we dab them anyway and create redirects later depending on sources available (see below)
• Stage 3 - we copy the directory lists into the mainspace by tabled lists by district/province with coordinates with place names un-wiki-linked.
• Stage 4 -IN consultation with the relevant wikiproject or research team we evaluate which articles can be generated in their own right if additional sources can be found. If not we redirect those in which "info isn't adequate" to the lists until a later date when enough info can be found
• Stage 5-After drawing up the full world lists we are in a strong position to evaluate which articles can be started. This is when the source compling and research begins for those places in which the bot can generate new articles under "village pump criteria" Sound good?

♦Blofeld of SPECTRE♦ ^$1,000,000? 19:37, 26 June 2008 (UTC)

The one clear downside to this proposal, unfortunately, is that I know of at least one very reliable and excellent source, which doesn't break the locations down by the 34 provinces, but rather by closeness to the six largest cities. If there are other sources, though, that would be different. John Carter (talk) 19:52, 26 June 2008 (UTC)

List of towns and villages in Bald Province

Place name	Province	District	Description	Altitude	Nearest settlements	Map/coordinates
Assio						WikiProject Geography/Bot/Archive (Burkina Faso)
Fritz town
Bald City
Giobo
Silouk
Mbogo
Tsawa
Dario
Esewa
Wafene

OR

see Wikipedia:WikiProject Geography/Bot/List proposal

List of cities, towns and villages in Badakhhan Province

Place name	Official name	District	Description	Altitude	Nearest settlements	Map/coordinates
Ab Gach	Āb Gach	Bald District	It is on the left bank of the Ab-i-Wakhan about a mile upstream of the meeting of that river and the Amu Darya.	450m	About ten miles southwest of Kala Panja	36°59′N 72°42′E
Amurn	Āmūrn		Located on the M41 motoway which connects it with Tajikistan near the border
Andowj	Andowj
Anjoman	Anjoman
Arakht	Arakht
Arghandakan	Arghandakān
Artin Jelow	Ārtīn Jelow
Arun	Ārūn
Ashkasham	Ashkāsham
Ashnam	Āshnām
Baharak	Bahārak

Please let me know what you think -personally I'd rather we have a small pin map like the above one but reduced ins ize to not bloat the table ♦Blofeld of SPECTRE♦ ^$1,000,000? 14:17, 21 June 2008 (UTC)

I like stubs better than lists. We could probably contain all the info in Wikipedia into lists and not skip anything. Stubs or red likns inspire article development more than redirects. I'm an Editorofthewiki^{[citation needed]} 14:46, 21 June 2008 (UTC)

I agree; many newbies, especially the ones from third world countries that we are trying to encourge to write about their localities, will not know they are "allowed" to turn a redirect into an article. Better to have a redlink. Phlegm Rooster (talk) 19:45, 21 June 2008 (UTC)

Well yes I would rather we have a full article on every place but if we were to generate two million sub stubs how long do you think "article devlepment" would take?. Truckloads of red links inspire people to create substubs to try to fill in the gap and generally most articles sit around underdeveloped for some time. If we cover exactly the same info that is generated in seperate new articles in lists people will think about a seperate article once more info becomes available. Remember that follwoing the village pump consensus the bot won't be creating seperate articles where there isn't enough content available anyway. Given that we will be actively assessing each country if there is enough info available, where there isn't enough info then we redirect to lists. If we didn't, most of the places in these countries would continue to be ingored because they don't meet the new requirments. Basiclaly you are saying that even if there is no info avialable and there isn't likely to be in the near future it is still worth creating a thousands of sub stub one liners which could be adequately given in one article. Once there is enough info available, then we create seperate pages. At least then the encyclopedia recxognizes those places which the Internet doesn't have much if anything on except location at present and people don't have to browse through 100s of embarrassingly short articles to retrieve the same info. ♦Blofeld of SPECTRE♦ ^$1,000,000? 14:51, 21 June 2008 (UTC)

Yes, and we are not approved to make stubs for every place on Earth, per the discussion at the Village Pump. I look forward to some input from the likes of Ryan Reich on the viability of the list idea form a notability perspective Fritzpoll (talk) 15:12, 21 June 2008 (UTC)

(e/c)I think this is an excellent way to organize instead of substubs that don't link together in the same comprehensive way that a list will (I hope that makes sense to someone not in my brain...). Also, I think I would rather enjoy Bald City. Sounds like a fascinating place. I might do well to open a hat store there. ;-) Keeper | 76 | Disclaimer 15:14, 21 June 2008 (UTC)

Believe me I would rather we had a full seperate article for every place in the world but we have to think realistically in terms of a time frame and level and rate of activity. I've had a good think about it and even if the handful of us tried to create expand every of the possible two million seperate articles it would take tens even hundreds of years to accomplish fully. Givne the scale of this project , yes if everybody on here put in a little contribution to expanding them, yes they might develop quicker, but the reality is that the bulk of the work remains with a limited group of editors and it would not be possible to get these articles up to adequate level seperately. This is why I think each country should be exmained and searched for sources and where we find enough info we generate seperate articles and redirect the rest to the list until more info can be found. I disagree with the notion that many places "aren't notable" because each place has its own history and story to tell however small. However I agree that theres no point in creating zillions of seperate sub stubs which we know can't be expanded realistically as it stands. This is why I think placing the content from sub stubs into a tabled list is the best start to make. It is also a very good way of regulating what we have and in the future if suddenly a lot of info becomes avilabale then we but the bot into action to create the seperate articles. ♦Blofeld of SPECTRE♦ ^$1,000,000? 15:22, 21 June 2008 (UTC)

Looks good to me, probably make the pushpin maps a little smaller, as some of the lists could be pretty long. Lists appear to be the only option due to the severe lack of editors working on geographical topics, something that is unlikely to change greatly in the short term. EJF (talk) 16:09, 21 June 2008 (UTC)

Yes the maps would be reduced considerably. ♦Blofeld of SPECTRE♦ ^$1,000,000? 16:11, 21 June 2008 (UTC)

I would have reservations about breaking the articles up into provinces, as per the above, because not all sources use that criteria. I know some of the sources I have used don't. However, I can see having the province or region of a country, if known, added to the details about a given area, possibly even as the first item listed. But not all sources directly state the explicit location of every place relative to national subdivisions, and on that basis I think it could be problematic to break up all articles by that detail. John Carter (talk) 16:34, 21 June 2008 (UTC)

Most countries do have information on pronvince location. Those which don't, we would have to work out how to do it, Azerbaijan., Armenia etc. Naturally I'd propose we create lists for the countries which have them first which I think you;d find is the vast majority. ♦Blofeld of SPECTRE♦ ^$1,000,000? 16:37, 21 June 2008 (UTC)

The countries might. However, having gone through several tourist books, I can say in all honesty many of them don't list the subnational entities very often. Breaking the entries up on the basis of those subnational entities might thus make it harder to try to find which page or entry to add a source you've found to. That was why I think a straight alphabetic list, by location name, maybe listing the province/region after the name as the first qualifier, might make it easier to work with the lists. John Carter (talk) 17:01, 21 June 2008 (UTC)

I think this proposal makes sense. The details can always be discussed: what would the "description" field contain? And wouldn't it make sense to have a "population" field? But the idea is basically sound. --Bonadea (talk) 16:52, 21 June 2008 (UTC)

• The other day, while editing Battle of Am Zoer, I considered creating an article on the town of Am Zoer itself. I couldn't find enough information in a timely fashion, so I simply put Am Zoer, 14°13′03″N 21°23′12″E, into the Battle article. If a user clicks on the little globe, they get a zoomable map. (And if they click on the coordinates they get to that GeoHack tool.) Any list would be improved by having the coordinates given for each town in this way, don't you think? Phlegm Rooster (talk) 19:28, 21 June 2008 (UTC)

I think that's a great idea. So we generate lists for every country with a clickable coordinate link to easily speed verification. Then each list page gets broken down into articles currently existing, duplicate entries get removed, problematic spellings being listed, etc etc. When a location has been verified and discovered to have sources demonstrating notability, then they get flagged to be created. Then, when enough (?) articles get given the notability/location verficiation check, the bot does all the magic. Basically? --NickPenguin_(contribs) 00:41, 22 June 2008 (UTC)

Basically the bot will generate seperate articles if there is enough currently available. If not it will redirect to such lists by distirct or province where the info you;d get in a standard sub stub articke will be provided in the tables. So basically I would hope that the encyclopedia recognizez all tow million places eventually evne if many of them redirect to the lists at present. Someday naturally we hope enough info will become avilabale for full articles on everywhere but we need to plan the best course of action in the time frame and resources available. ♦Blofeld of SPECTRE♦ ^$1,000,000? 02:21, 22 June 2008 (UTC)

Come to think of it, can we also generate these lists with a clickable Google search (something in the form of CityName CountryName)? That would speed up determining if a location is notable. Basically, I want to have to make as few a clicks and do as little manual typing as possible. There are several hundred thousand articles that potentially could need to be created, so if we can make this a 2 step verification process rather than a 6 step, this project can be 3 times as efficient. --NickPenguin_(contribs) 15:50, 22 June 2008 (UTC)

Easily done - I can add a separate external link to the initial lists so that we don't lose the ability to dab-check the internal links. How would that be? Fritzpoll (talk) 15:53, 22 June 2008 (UTC)

I like the idea of creating lists subordinate to the provinces (despite John Carter's worries about accessibility for places whose province is not clear in the sources, if we are redirecting to the list, the search tool should address that issue. We can even anchor each list entry and redirect straight to that spot, right?). Notability-wise, the list entries might still be too bare to warrant a full article, but the whole list is a valid and useful addition to a central article on the province itself (in other words, I like this idea if it means the lists get put into, or at least linked from if they are too long, a larger article). What NickPenguin said just above goes for me too: we let the lists do their quiet magic and when editors have fleshed out the "description" enough for an entry, an article will be created. It might be useful to keep the list (of names only) in the province's article even when its entries have their own articles, for completeness and centralization of this (rather disparate) information. I want to repeat Bonadea's comment that there should be a population field, and in fact if we can think of any other standard data that might be included, there should be a field for it too; that way, the "description" will end up having only place-specific data (i.e. notability claims) and once it starts filling up we'll know the article can be launched. This would give an outlet to new editors who don't know what's "allowed" but still want to contribute. Ryan Reich (talk) 15:33, 22 June 2008 (UTC)

I agree that it would be useful to have the list stay around even after the articles have been fleshed out. Specifically, I think that once a region has been sufficiently developed (say 5 start-length articles or some such) then the localities in that region should go into a navigation box so they can be easily interlinked. This would keep the links around on all of the pages involved. Adam McCormick (talk) 17:01, 23 June 2008 (UTC)

Back to the discussion of table layout... I like the idea of a pinmap, but maybe we could include both a map of the region within the city country and a numbered map of the locations in the region? Adam McCormick (talk) 21:30, 26 June 2008 (UTC)

I;d rather a small pin map too but remember that we need to keep the tables as condensed as possible, particularly if there are many places in a given district ♦Blofeld of SPECTRE♦ ^$1,000,000? 21:39, 26 June 2008 (UTC)

Which is why I'd suggest one map for all the locations, it would save space over a map per place but still allow for visual identification. Additionally, it might be possible to template this along with the table itself so that it would generate automatically. Adam McCormick (talk) 21:55, 26 June 2008 (UTC)

One map for all locations? Nono it doesn't work like that. You can't have hundreds of places on one map anyway ♦Blofeld of SPECTRE♦ ^$1,000,000? 22:59, 26 June 2008 (UTC)

That particular template doesn't allow more than one location but the template syntax could easily be copied and modified to allow for as many "pins" as you might want. Adam McCormick (talk) 23:04, 26 June 2008 (UTC)

Interesting. It would have to be about 800px though at the top of each list. If that could be done, the globe coordinates could go in the final column rather than map ♦Blofeld of SPECTRE♦ ^$1,000,000? 00:18, 27 June 2008 (UTC)

Two aspects of progress

We have to proceed in two directions. Getting a list of places to have articles on would not be difficult, which is not to mean we don't have to prepare one. But, there's another aspect—how effectively the bot functions. I'd made a suggestion when voting was on, which had largely got ignored that if and when the BOT is created, we should have a prototype country, say "Wikiland", which will have a few cities in it, say "Wikicity1", "Wikicity2", etc., with a certain language currency, religion, etc. This is important as it will be equally important to decide what all headings a place will have to be described under (like, location, capital, language, major religions, and so on). Also, this will make the idea of a BOT-generated article clearer and attractive to all the people.

—KetanPanchal^taLK 15:24, 21 June 2008 (UTC)

Not sure what you mean by "how effectively" the bot functions. The bot has already been demonstrated to operate competently from a technical perspective, after tests and demonstrations requested by the Bot Approvals Group. Determining a template that it will use can either be standardised for all countries, or determined on a country-by-country basis. Fritzpoll (talk) 15:29, 21 June 2008 (UTC)

Well, I really didn't know that such a BOT has been tested. May I know where one can get to try it? If the basic issue is if we should prepare a list of places, then I'll say "most definitely yes". —KetanPanchal^taLK 16:09, 21 June 2008 (UTC)

Well, I'm the approved operator, and it is code under my control, however the approval is at Wikipedia:Bots/Requests_for_approval/FritzpollBot. Hope that helps. Best wishes, Fritzpoll (talk) 16:19, 21 June 2008 (UTC)

Thanks Fritzpoll, I already tried the template. So, now I understood how it works. Filling in the template would generate an infobox that would be available to you and when you're satisfied with the data, you'll order the bot to created the article. That's perfectly fine.

One of the issues raised during voting was also regarding what notability-criteria are to be considered for creating an article? —KetanPanchal^taLK 11:00, 22 June 2008 (UTC)

Yes, per the instructions on this project page - only articles that satisfy notability guidelines will be created. What's being proposed above is placing all places into lists but not creating full articles without sources Fritzpoll (talk) 11:20, 22 June 2008 (UTC)

Two million articles/places

There's this talk of creating two million articles through this bot. So, I'm assuming one article would stand for one place, and each place will have exactly one article covering it. Right? If this indeed is what's planned may be somewhere our calculation is wrong.

The current population of the world can be considered to be 6 billion, and if there'd be 2 million articles, then that comes to an article for every 3000 people, which is a very small population comparing with the criteria that were being considered earlier.

Please do let me know if my analysis is flawed in any way.

—KetanPanchal^taLK 11:09, 22 June 2008 (UTC)

No, there is no talk of creating 2 million articles with this bot, at least, not within this project Fritzpoll (talk) 11:16, 22 June 2008 (UTC)

My doubt was in reference to the question How many articles? on the FAQs page, and the figure of 2 million in one of the preceding threads on this page itself (sorry that was an approximation, but that still doesn't really change the calculation much [3333.3 people instead of 3000.0 people!]). To be honest, I'm not much bothered by the number of articles; just that I thought that this was some kind of miscalculation. —KetanPanchal^taLK 14:30, 22 June 2008 (UTC)

Ah, the FAQ is out of date and requires alteration. Thanks for drawing attention to it Fritzpoll (talk) 14:49, 22 June 2008 (UTC)

That's alright then. All the best! —KetanPanchal^taLK 15:24, 22 June 2008 (UTC)

Tables vs. stubs

Having been asked by Mr. (Dr.?) Blofeld to comment on the table idea (above), my view is that it might be a good idea, but the implementation would be up to the WikiProject, instead of being applied universally. If a WikiProject wants all of the places in a table, give 'em a table, otherwise do the stubs. That's my opinion, anyway. Best,—Ëzhiki (Igels Hérissonovich Ïzhakoff-Amursky) • (yo?); 16:15, 23 June 2008 (UTC)

I would myself favor a basic "list" format. One think that could very likely arise would be that, for example, certain very notable buildings might have architect and construction data available on them, while others might not. It would probably be a good idea to include it where it is available, but it might not happen often enough to warrant a separate column. There are additional pieces of data which might be available on only some locations, as well. I think a list would be the most easily adaptable format to use for all such entries, as it would allow inclusion of whatever data is available. John Carter (talk) 16:39, 23 June 2008 (UTC)

New Templates

During the discussion on this bot I volunteered to be the Template goto guy, so here goes. I'd like to start making templates but I need to know what everyone thinks are necessary (maybe project banners and such for now). Anyhow, the following is a list of templates I think we'll need:

• WikiProject Banner (for all the new talkpages created by the bot)
  • I've buit an initial version at {{WPGEOBOT}}. Let me know if this will suffice
  • I should think we'd use WP:GEO's assessment banner so ours is just a "This was created as part of this effort" sort of thing
• Need to decide what stub tags we'll use
• User box (which I've taken a crack at in {{User GEOBOT}})
  • Also, let me know what you think of the logo I put together at {{WPGEOBOT Logo}} and used on the user box as i think it might work for this project's logo.
  • Thanks to User:Stepshep we now have a stand-alone logo (which is now placed by the template)
• Nav boxes (some standard box to link a province or region)
• Maybe a templated version of the listbox proposed above for the bot to place
  • As part of this I've been working on a multiple-point version of {{Location map}}. An early version is at User:Alanbly/GEOBOT map with an example at User talk:Alanbly/GEOBOT map
  • I think though, if we are going to use this kind of template we'll need regional maps, not just countries. Otherwise maybe it only makes sense for labeling subregions. Adam McCormick (talk) 01:46, 28 June 2008 (UTC)

Template Discussion

These are just my thoughts, I'm sure there are others, feel free to add more. Adam McCormick (talk) 20:04, 23 June 2008 (UTC)

I think the logo looks really good and you've done great work on the other template too. The wording on the WikiProject banner could possibly be altered, maybe change balance to expand? --Kaly99 (talk) 18:47, 26 June 2008 (UTC)

I just threw some text up, It can be changed easily Adam McCormick (talk) 20:34, 26 June 2008 (UTC)

GEOBOT logo looks good but I;d suggest cropping it so only the two biggest cogs on the end are included ♦Blofeld of SPECTRE♦ ^$1,000,000? 20:09, 26 June 2008 (UTC)

I'm confused, there should only be 2 cogs in the image. Do you mean the scaling factor example? Adam McCormick (talk) 20:17, 26 June 2008 (UTC)

Ahh soz I was looking at the scale. Nice job! ♦Blofeld of SPECTRE♦ ^$1,000,000? 21:30, 26 June 2008 (UTC)

Those look amazing! Fritzpoll (talk) 20:26, 26 June 2008 (UTC)

Thanks, I need to make a PNG alternative for IE (it's pretty ugly for IE) but it will just be an option, unless there are some SVG folks out there who could combine the images more permanently... Adam McCormick (talk) 20:34, 26 June 2008 (UTC)

Excellent! The combination of cog wheels and globe is spot on for the project. Your implementation of it looks really good, too. --Bonadea (talk) 07:26, 27 June 2008 (UTC)

So can we fill up this list?

I see no reason why we cannot generate lists for these countries. Certainly not everyone is interested in improving articles in the same country, so if we have decided on the format the lists should be generated, then we can let people pick and choose what they want to work with, and also evaluate the total possible number of creatable articles. Perhaps we can also add clickable links to the American Samoa list; a lot of editors have put work into organizing the list, but the verification and notability check is perhaps the most time consuming aspect of this project. --NickPenguin_(contribs) 05:53, 1 July 2008 (UTC)

I agree we should get all the two million places onto here in a list. I sincerely hope that Fritz can use Mr Bot to generate them once he finishes moving and is settled ♦Blofeld of SPECTRE♦ ^$1,000,000? 20:55, 3 July 2008 (UTC)

Contacting the national projects in question?

I know that several of the extant national/regional projects, even if not all of them, have several editors who are actively working on their content. It would certainly be faster if we could get them involved in creating the lists. What do the rest of you think of dropping notes on the talk pages of the relevant geographical projects and asking them to help in the construction of the lists? John Carter (talk) 20:51, 1 July 2008 (UTC)

I love leaving notes on Wikipedia. Why? Because they linger on talk pages, and more people than just then intended audience tend to catch wind of goings-on. --NickPenguin_(contribs) 21:05, 1 July 2008 (UTC)

Why is India on hold ? We , WP:INDIA with over 55 active workgroups , will be glad to help -- TinuCherian ^{(Wanna Talk?)} - 05:05, 2 July 2008 (UTC)

India was on hold originally because we learned of a possible 638,000 settlements, I think it was because there was so much Fritz decided to pass it till later ♦Blofeld of SPECTRE♦ ^$1,000,000? 20:53, 3 July 2008 (UTC)

Visualizing the task

Worldwide density of geotagged Wikipedia entries

Worldwide density of GeoNames entries (2006)

One should not dwell too much on the fine print (probably not every GeoNames entry deserves a Wikipedia article, many Wikipedia articles still await geotagging, and I'm not sure if both use the same color scale), nevertheless I find it quite enlightening to compare these two images. Regards, HaeB (talk) 02:21, 8 July 2008 (UTC)

Amazing, thanks. emijrp (talk) 14:11, 1 September 2010 (UTC)

Sources

Can there please be some kind of information on what sources are being considered for use by FritzpollBot and this project? The FritzpollBot page just says that it uses "a public domain source". This page isn't much better, stating only that "a variety of sources" have been identified. What are the sources? Kaldari (talk) 19:11, 14 July 2008 (UTC)

Well, so far as I can tell, the primary source from which a listing would be added to the list of potential articles will be created would be the Geographic Names Information System. The information like coordinates and the like would be added to the list based on the information from that site. Then, we would go through the various other sources which some of us might have available, whatever they might be, and either mark the list entries as being long enough to have separate articles and/or maybe adding what minimal information on the locations is available. In such cases, the content regarding the location would be added to some article with a slightly broader scope. However, I hope you can understand that, given the wide variety of sources available for all the various locations out there, it's rather difficult to identify a specific set of most commonly used other sources. John Carter (talk) 19:19, 14 July 2008 (UTC)

Last I heard GEOnet was the primary source that FritzpollBot was going to use, not the Geographic Names Information System. Is it going to use the GNIS for locations in the US and GEOnet for locations outside the US? It's certainly OK if this page doesn't list every source we're going to use (since they aren't going to be consistant). But it really needs to have some sort of minimal information on the main sources we are using to generate the place names at the very least. Kaldari (talk) 19:24, 14 July 2008 (UTC)

Your right, GEOnet is the primary source for foreign sources. Same government, wrong entity, sorry about that. I do think it is worth considering to maybe consider generating articles on US area locations other than populated places that would otherwise qualify as notable. Basically, we're going to use the US government sources to generate the lists, and then let the relevant national/regional projects and their members use whichever sources they want to use for additional verification. Hopefully, they'll have a better idea of what sources are best, etc., than members of this project will. John Carter (talk) 20:17, 14 July 2008 (UTC)

Bot run tomorrow or thursday

Notification that I will be running the bot in the next 48 hours to begin uploading the countries as tabled lists. I have been asked to begin with Afghanistan - is there consensus for this as the choice of country? Fritzpoll (talk) 01:34, 23 July 2008 (UTC)

Could you explain in more detail what it means to "begin uploading the countries as tabled lists"? Kaldari (talk) 02:34, 23 July 2008 (UTC)

Also, I'm interested to see what the output of each new article would look like. Is there a sample article that we can see? — Andrwsc (talk · contribs) 04:44, 23 July 2008 (UTC)

Are you going to create one list for the entire country or several lists for the different provinces? Have you verified that no such lists already exist? Are you intending on creating redirects from town names to these lists (as suggested in a discussion above)? If so, are you aware that a similar plan (to turn redlinks into list redirects) was adamantly rejected at WikiProject Tree of Life just this past week (and would be considered extremely controversial by some)? What sources are going to be used to create these lists? What are the list articles going to be titled? "List of populated places in Jimbo Province"? "List of towns and villages in Jimbo Province"? Are you going to create lists of rivers? Mountains? Other geographic features? Any other details would be appreciated. Kaldari (talk) 06:35, 23 July 2008 (UTC)

Thanks for giving us a chance to monitor and comment on your progress, BTW. Kaldari (talk) 06:54, 23 July 2008 (UTC)

Not a problem - it's what I promised I'd do! :) This stage doesn't involve creating anything in the mainspace - they would just be the lists for checking by editors, and include fields, which may be blank if the information is unavailable. This allows places to be verified, disambiguation to be resolved at a later stage, and also for new sources to be found and material added within the table. Without sufficient information, a place would never leave the list, and never be made into its own article. I'll also insist on verification of each place by some means other than the centralised databases, such as a reference to census data.

There is talk above about creating the lists into articles with the titles Kaldari suggests, but I won't be happy with this without a second reliable source confirming the existence of a place.

I have no plans at this time to include rivers, mountains or other geographic features, as there is no consensus for the bot to do so. Hope this answers all your questions - let me know if you have any more. Fritzpoll (talk) 07:55, 23 July 2008 (UTC)

Thank you, thank you! My paranoia has been allayed :) I just really don't want to see another repeat of what happened to the species articles due to Polbot. See my discussion on Blofeld's talk page if you'd like to understand my crazy paranoia further :) Kaldari (talk) 16:42, 23 July 2008 (UTC)

If the Afganistan list works out, it would be great to have lists for all of the countries. --NickPenguin_(contribs) 20:20, 28 July 2008 (UTC)

Has Afganistan been run? Where should we expect to see this list? Kaldari (talk) 20:31, 28 July 2008 (UTC)

Not done - the bot needs a bit of recoding to take account of the new guidelines, and it is taking longer than I expected. As I said to Blofeld on my talkpage, very sorry for the delay Fritzpoll (talk) 20:32, 28 July 2008 (UTC)

Any luck on the Afghanistan list? Is the main source going to be GeoNames+maplandia or have other sources been found? Kaldari (talk) 19:25, 5 August 2008 (UTC)

No other sources - I was rather hopeful at the start of this process that other people would help look for them, as I have quite a bit to do on the bot front... :) Fritzpoll (talk) 20:21, 5 August 2008 (UTC)

The Google geography directory has some good resources, but I'm not sure which one's you're using and what you're looking for in a source. World-Gazetter look like it would be easy to run through a bot and generate country lists. This page or maybe this page might have some good links for general and country specific info. --NickPenguin_(contribs) 04:21, 7 August 2008 (UTC)

I tried finding a better source for Afghanistan myself, but wasn't able to locate anything on the web :( Kaldari (talk) 17:21, 11 August 2008 (UTC)

Notability - Places

I've started a discussion at Wikipedia_talk:Notability_(Geographic_locations)#Using_an_Atlas_as_a_source_for_notability which menbers here may be interested in contributing to. I would appreciate your views! Thanks AndrewRT(Talk) 13:35, 27 July 2008 (UTC)

Refactoring

I have a question. How is the geography bot going to maintain the articles it creates? One maintenance task would be to update population figures, and another, more complicated task would be to convert articles to newer styles.

« D. Trebbien (talk) 03:03, 6 August 2008 (UTC)

We now have 900 incorrect articles about Armenian settlements

It looks like Carlossuarez46 used the list of Armenian settlements created by FritzpollBot to generate 1000 or so articles on these settlements. The problem is he did not use the sources available to discern whether the settlements were "Քաղաքներ" (towns) or "Գյուղեր" (villages). Instead he just decided that they were all "towns". Unfortunately, 90% of them are actually villages (Գյուղեր), so currently we have about 900 articles on Armenian settlements that contain misleading information. I can provide a list of which settlements are officially towns and which are officially villages if you could program FritzpollBot to correct them. Kaldari (talk) 23:33, 11 August 2008 (UTC)

BTW, we also have a similar issue with all the Bangladesh articles created by Blofeld. There is no distinction made as to whether the settlement is a gram or a mouza (or both), they are all simply "villages" (which is typically translated to gram). Unfortunately, there doesn't seem to be a comprehensive source available to correct these articles with. Without verifable sources to clarify the status of these places, we should default to referring to them as "settlements" rather than "towns", "cities", or "villages", as those terms often have very specific meanings (which vary by country). Kaldari (talk) 00:16, 12 August 2008 (UTC)

Closing down this project

I think its time we closed down this project as unfortunate as it is. It has been active for months and I see little evidence that in all honesty it will ever go into operation. It is now October and I doubt anything will happen before Christmas. It could have worked very well but I guess we'll have to continue manually and get there by 2030. The Bald One ^{White cat} 11:31, 2 October 2008 (UTC)

I have to agree - but the problems were not urely technical. One of the problems was that noone ever opened up one of the pages per the instructions given, and gave me source material to work with. That's been a large part of the problem - with no apparent interest in giving me the sources, and satisfying the requirements that the community laid down, such as participation and interest shown by other users, it has been difficult to inspire myself, when I do have the free time, to sit down and run it. After all, run what? I know the format desired, but I have no reliable dataset that I can cite as having been checked by people. Without this, it was doomed to fail. Fritzpoll (talk) 11:45, 2 October 2008 (UTC)

I presented the data for the Nepalese settlements to you here. American Samoa and Ethiopia (Llywrch offered his data here) had available data and also Senegal (which was greatly needed). I even asked if you could see about writing it into the bot as it had data on all the aspects of demographics in a choice of file. I felt quite certain to information given here was suitable for the bot and it would have saved me and ed a lot of time creating them manually as we have been doing with Valam etc and also a bot would have started them with as much detail as possible with the result all now started would look far better. I didn't get any repsonse though, but I think it coincided with sponge bob square pants messing you about over admin conductivity. The Bald One ^{White cat} 11:57, 2 October 2008 (UTC)

Never saw it, I fear - ping me with the data again, and I'll take another look. Fritzpoll (talk) 12:00, 2 October 2008 (UTC)

CLick on one of the district and highlight a settlement. You can even highlight them all in one district and choose one of the options from Economic Activity,Literacy,Marital Status,Religion, Population (total pop and number of houses), Population (in 5-year divisions) School Status. You even then have the choice of which format to save it as. I was certain that ALL of the data for each settlement could have been saved and the processed so the bot could have started the articles with a fixed text format and just adding in the data in the gaps for each one The Bald One ^{White cat} 12:04, 2 October 2008 (UTC)

Well, provided others agree that the sources are reliable, this can easily be done. Just start the sub-page, fill it in, and ping me once you're done. I could be running this as early as tomorrow evening (which I have to myself) Fritzpoll (talk) 12:06, 2 October 2008 (UTC)

Shall I create a mock up page? The data is undoubtedly reliable given that it is from the official census and compiled with the governors of each VDC. The Bald One ^{White cat} 12:11, 2 October 2008 (UTC)

See User:Fritzpoll/Nepal The Bald One ^{White cat} 12:31, 2 October 2008 (UTC)

Make it public, then everything will be in place: Wikipedia:WikiProject_Geography/Bot#Procedure Fritzpoll (talk) 12:36, 2 October 2008 (UTC)

Do you mean creating the list of missing villages manually? Also see the progress here. If I list the district that haven't been covered could you copy the settlements or shall we do it? The Bald One ^{White cat} 12:42, 2 October 2008 (UTC)

OK I;ve started Wikipedia:WikiProject Geography/Bot/Country List/nepal. The Bald One ^{White cat} 12:53, 2 October 2008 (UTC)

Even more beneficial would be getting all the municipalities of Mexico onto here. See Guachochi (municipality). Every municipality has a truck load of data which could be processed e.g from here or/and using spanish wikipedia which could be used to create start class articles. Then I could concentrate more on translating and reowrding text to expand all the articles to B class. The Bald One ^{White cat} 12:39, 2 October 2008 (UTC)

So we seem to have got back to the original 'approval' issue, is a minority of fervent supporters combined with a bot looking for a reason to exist, a sufficient justification to approve the rapid creation a legion of permastubs on wikipedia? MickMacNee (talk) 13:23, 2 October 2008 (UTC)

We're not creating permastubs, as well you know if you've been reading the above, and if you read the closing consensus and proposal. No stubs will be created on locations that have no notability of their own - this bot has a reason to exist, and whilst I appreciate your concerns about bots, the community voiced its opinion and disagreed with you. Fritzpoll (talk) 13:26, 2 October

2008 (UTC)

Perma stubs Mick? Of course nobody could expand articles using this as a source could they? What you;ve just said illustrates to me you know little about the manual work that has been going on all these months at expanding some of those "perma stub" articles. La Palma, Chalatenango and Apastepeque were once part of those one line Central American stubs we have in abundance on here which nobody thought would ever be expanded. OK the latter it needs referencing but at least it has translated content and images. The Bald One ^{White cat} 17:18, 2 October 2008 (UTC)