Wikipedia talk:Most wanted stubs

Query

This page was generated with the following SQL query:

SELECT cur_title, COUNT(*) as numlinks 
FROM links, cur 
WHERE l_to=cur_id 
  AND cur_is_redirect = 0 
  AND cur_namespace=0 
  AND LENGTH(cur_text)<100 
GROUP BY cur_title 

ORDER BY numlinks DESC LIMIT 100

The query should be used sparingly as it puts a fairly high load on the server. (see Wikipedia:Database queries)

I will be regenerating the page every now and then according to demand. We can experiment with varying the size of article we look for - the present 100 character limit is very short. Enchanter 18:25 Aug 26, 2002 (PDT)

This page should pop up when Special Pages is selected. Ortolan88 18:31 Aug 26, 2002 (PDT)

Enchanter: Of course, now all of the articles on this page have one more "wanted" link, this page itself. When you regenerate, be sure to delete all the links from this page first, for a more accurate count!  :) DanKeshet

---

Expanded stubs

Turned The Scotsman into an article and removed it from the list. Palnu 03:53, 13 Jul 2004 (UTC)

---

I just did some work on Assiniboine River. Does it still count as a stub? If it doesn't then please remove the stub marker. Canadabear 22:01, 28 Feb 2004 (UTC)

Are we supposed to remove items if we expand them? I did a little more on New Haven, Connecticut and antipyretic. Vicki Rosenzweig 18:56 Aug 26, 2002 (PDT)

Yes, I think it might be a good idea to remove items, and I've changed the subject page to reflect that. It's not absolutely necessary, as the link would in any case automatically be removed when the page is regenerated. However, I think removing items from the list is a good marker of progress. It would also help to give this page lots of mentions in recent changes, which would help to attract more people to fix stubs. Enchanter

I think 'industry' can be removed now. cferrero

I think Conservative Party of Norway can be removed. Maybe not. --Adam Wang 02:40, 16 August 2006 (UTC)

---

[X] This is a minor edit [ ] This is not a stub [X] Watch this article

Wonder if there would be an easier way for people to mark whether articles are stubs, than by writing it in the article...

Suppose can be hard to see what needs doing, since links to stub articles look like links to proper articles, rather than links to unwritten articles... كسيپ Cyp 21:44 Mar 7, 2003 (UTC)

Currently, you can set your set your preferences (Special:Preferences) to change the color of links to smaller articles ("Threshold for stub display", e.g. 100). -- User:Docu

Update list

I found Bhanja also under Music of India. It has a couple sentences down on the subject. What do I do in this situation? Canadabear

Would someone please re-run this to show current results? clarka 24 Aug 2003

Done Enchanter 22:22, 27 Aug 2003 (UTC)

I updated the list once more. Not all stubs have that many pages linking to them as listed and a few are disambiguation pages not formatted as such (currently John Nash, IRS). -- User:Docu

---

I tried to improve "Geographer". I'm rather new at this and not sure if I've done enough to de-stubify the page. Could someone confirm it and remove the stub text from the page? fvincent 17:00, Dec 1, 2003 (UTC)

There's no set limit on what counts as a stub. It depends a lot on whether you think enough is now covered in the article or whether it is still useful to have the stub notice. Personally, I think maybe more could be written but that it would be ok to remove the stub notice now. Up to you really. Good work on it by the way! Angela 19:34, 1 Dec 2003 (UTC)

Ok, I've removed the stub text because I don't think it's a stub anymore but I do agree it could be expanded quite a bit, hopefully by a geographer! fvincent 19:59, Dec 1, 2003 (UTC)

---

Perhaps it's time to recompile this page? Every three months sounds reasonable. fvincent 19:09, Dec 2, 2003 (UTC)

Three months does sound reasonable. Page now updated. Angela 00:27, 3 Dec 2003 (UTC)

This is three months out of date again, but I can't do anything about it now as sysop queries have been limited to 30 seconds, and the query needed to do this takes about four minutes. I've listed it on Wikipedia:SQL query requests so someone with a recent copy of the database can update it. Angela. 17:03, Mar 2, 2004 (UTC)

---

Albuquerque and other disambiguation pages shouldn't be here. 24.6.66.228

Feel free to remove them. The page isn't protected. Angela. 06:16, 31 Dec 2003 (UTC)

---

The updated list just added is mostly bogus--there's a bug that lists links from Orphaned articles a zillion times, so that by far the marjority of these have 0, or maybe 1 or 2, links to them in actuality. I started deleting those, but it looks like most of the list will then go away. Can someone generate something mo betta? Elf | Talk 04:38, 30 Mar 2004 (UTC)

The duplicate links to Orphaned articles is a known bug with the links table. You can request the query be re-run at Wikipedia:SQL query requests. See this version for an example. Angela. 23:19, Apr 1, 2004 (UTC)

Restocked the list

I have restocked the list with 100 items from a user report: User:Topbanana/Reports/This is one of the most linked to stubs. I'm working on getting a list of all most-wanted short articles, but until then this will do nicely. The drawbacks:

• Only four articles are actually less than 100 characters, as the page is supposed to be.
• Only pages with {{msg:stub}} are included, so unmarked stubs are still unseen.

Hopefully I (or someone wanting $10) will have a chance to run the original query. Between now and then, this list gives plenty of opportunities. Just as before, if you expand an article to the point that it is no longer a stub, remove the stub message and remove it from the list. Happy writing, --Ben Brockert 02:33, May 25, 2004 (UTC)

Do we remove them, or strike-them through? How long until we can get this updated again? Bo-Lingua 17:47, 8 Mar 2005 (UTC)

History Calls to Arms

I was intrigued by the fact that so many topics along the lines of: XXX BC had the same number of referenced links... Upon checking it out, this whole group needs to be considered a superstub of sorts. This is an worthy attempt to build a timeline of history along the lines of an almanac. If there is a 'wikiwide' communciation means, especially one that can be used to target history interested parties such as my humble self, I have a suggestion to clean this up expeditiously. Ask any and all members, for a two to three month genesis, to make an edit toward this worthy goal (entries are essentially 'bulleted single line items') inside the appropriate decade anytime they run across a dated historical event. After the genesis period, 'stub' can probably be safely removed from said articles -- some decades were (apparently <g>) boring so that little of historical note happened. This should shorten next years list significantly, assuming success in mobilizing the masses! A project like this needs publicity. High kudos to whomever concieved the idea! Fabartus 02:06, 30 May 2005 (UTC)

New data available

Hello,

Dmcdevit told me about this cleanup project. I have written my own software to generate a report for you which you can find at Wikipedia talk:Most wanted stubs/20050623-triddle. I would of integrated the new list into the page myself but I thought it would be best if a member of this project did it. Also sorry about the broken links; they should point at real articles but something is subtly wrong with the name. I'm not quite sure why that happens. Triddle 06:26, Jun 26, 2005 (UTC)

The trouble with the red links is already fixed in this list and later on Triddle's list. Is due to a misspelling of non English characters by the database dump. Perhaps the software could be fixed so that the list replaces the wrong non English characters for the right ones? It would be nice. Vale, Lcgarcia 30 June 2005 18:52 (UTC)

Merging new dataset myself

After checking the history of these project pages I've decided its likely that there is no one here to read this message or my previous one about the new data. I'm going to take it upon myself to replace the old list with my new one and list this page on the active cleanup projects template. Triddle 19:58, Jun 26, 2005 (UTC)

Many links, no difference

I've checked two of the items on the main list (Kilis province, a Turkish province and an American athlete), and each had the same problem: they were linked out of a list that was repeated on every member page. Every Turkish province page has a list of all the other provinces, so Kilis gets numerous hits, and every US Olympic medal winner (athletics) has a table of all the other ones as well. This means basically that those two (and probably many others on this list) aren't really important or popular (based on the number of links), but just the victim of a bad use of lists. I guess that theoretically, all the pages linking to thse two (except for the parent / category) should be cleaned, but this seems a bit radical... Fram 12:56, 22 August 2005 (UTC) After checking a few more, I come across places in Singapore, districts in Uganda, plays by G.B. Shaw, ... all of those are included here for the same reason, making this list virtually useless. Is it standard practice that in a low-level article, a list of all other articles of the same category are included? It seems like a strange thing to me, but I'm new here. Fram 13:00, 22 August 2005 (UTC)

Well, I think including a list of other similar things is frequently quite useful, and I have seen it done often, but by no means all the time.

For the purposes of this list, it does complicate things however... Maybe what would be more useful would be a list of stubs most linked from non-stub articles. So the number of incoming links to say, Nebbi would be greatly reduced... Districts of Uganda would be counted, but Nakapiripirit and Mukono, for example, would be discounted.

I've no idea whether creating a list like this is feasible? --David Edgar 10:19, 24 August 2005 (UTC)

Most of the time, all those links are due to templates. A smarter way of counting would be to count a template only once. But then you'd need a smart way of counting links while ignoring links from inside templates - any ideas ? It looks like templates don't include any metadata in the html source, maybe ignoring links in tables (since they are templates 99% of the time) would be a good enough heuristic ?

Only counting links from non-stub articles also sounds like a good idea :)

Heck, there's no reason not to make *several* most wanted stub lists on different pages, one that ignores links from stubs, one that ignores links from inside templates, etc. Then it's up to the editor to choose what he wants to work on :) Flammifer 07:39, 27 August 2005 (UTC)

Defining a stub

• A better definition of what makes a stub a stub may be needed - perhaps anything marked as a stub + anything shorter than 250 chars? - TB 10:07, Jun 9, 2004 (UTC)

I would just go with anything that's actually tagged with a template that has "stub" at the end of the title. Anything that's short and not tagged as a stub should show up on the "Should this be a stub?" page. Anything that's long and tagged as a stub should show up on the "Is this really a stub?" page. -- Beland 21:44, 30 Apr 2005 (UTC)

Removing a listing

1. Frederick_Law_Olmsted (151 links) was on the list at 350. It's a nice long article that isn't tagged as a stub, so I removed it from the list. Metarhyme 06:53, 31 December 2005 (UTC)

1. I struck out Coat_of_arms and ore; neither of them were tagged as stubs at this time. tweekus 02:56, 6 January 2006 (UTC)

"Most wanted stubs" is a waste of time, keep user-pages and talk-pages out of the statistics!

The link numbers have no significance whatsoever. Most of the hundreds of links come from user-pages and talk pages. Why are they even included in this statistic? How do links from a user-page that is probably accessed once every three months warrant the attention of anybody but the user? I think the talk and user namespaces should NOT be searched when these numbers are gathered. Mstroeck 13:10, 29 January 2006 (UTC)

• The user and talk namespaces are not searched when counting inbound links to a page. What leads you to believe that they are? -- Beland 07:40, 20 February 2006 (UTC)

It's worse than that. Often large numbers of links come from one specific project where every page links to every other page. The list of asteroids that always pops up as being hugely desirable is, uh, undesirable. Stevage 20:33, 29 January 2006 (UTC)

• Well, the theory is that if many pages link to an article, it's more likely that readers will follow a link that brings them there. If you have an idea for an alternative way of quantifying "most wanted", I could change the algorithm. We currently don't keep track of page views, so either some software development would need to be done to fix that, or it would need to be something where we only look at the wikitext of articles and other pages. -- Beland 07:40, 20 February 2006 (UTC)

Stale!!

Hasn't there been a database dump since January? Or do we only get a refresh when everything is crossed out? Or, does someone manually do this? Or a bot? Should there be a bot if there isn't one? Would a bot sophisticated enough to do this mind-numbing task eventually learn and grow until it tried to delete everyone's userboxes whenever Jimbo sneezed? --James S. 10:55, 29 March 2006 (UTC)

Very stale!! Personally, I'd like to see a fresh list, in order to find good candidates for collaboration of the week. -Scottwiki 10:57, 3 April 2006 (UTC)

Woefully Out of Date

I followed a number of links in the rather lengthy list of most wanted stubs, only to find that they weren't stubs. They had been expanded to full articles (or were articles to begin with?).

As is the list seems rather pointless.... angrykeyboarder 03:59, 6 May 2006 (UTC)

Someone should just delete all the articles(from the list) that are struck through. That would reduce the size of the list a bit, and make it a bit easier to manage.dbalsdon 02:06, 30 June 2006 (UTC)

I've just manually removed all the struck through articles from the list. I don't see how this would be a problem, but if it is, just revert, and please accept my apologies. dbalsdon 18:23, 5 July 2006 (UTC)

I've just done another purge. -- Beland 19:21, 9 September 2006 (UTC)

Wikipedia Integration: solution to some stubs

I have worked with a few individuals to form a project to eliminate poor stubs and duplicate content in favor of more comprehensive articles. I agree some articles will always be "stub" in size, but those that are suspect to development in more than 1 article for the same subject matter should be merged, and maybe split off again when deemed necessary.

WP:ʃ

Cwolfsheep 23:17, 12 July 2006 (UTC)

Usability?

I noticed that the rank of the majority of stubs is skewed by includion of them in certain templates. It is one thing that a term is used in the text of an article; it is totally another if it is in a brainless list of small towns in an even smaller canton of Szitzerland.

I noticed the same problem when backtracking links to certain articles, which were very difficult to sort in terms of relevance because of includion in templates. Is it difficult to have a tool that traces "what links here" but ignores transclusion? Mukadderat 04:47, 6 October 2006 (UTC)