Wikipedia talk:WikiProject Stub sorting/Guidelines
Stub sorting guidelines
editI sincerely believe that we really need to lay down the law in stub sorting, and really provide a guideline. I believe that we should all attempt to reach a general consensus by April 2, 2005 in a set of rules that we can follow. Once we have built a set of guidelines, we can formally create a policy out of those guidelines. We need to define what a specialized topic stub is, how many articles it should cover, when is it appropriate to create it, what defines its need for its creation, what defines its need for deletion, what criteria it should follow, what are the general steps should one take when sorting a stub, whether or not to start using subst: for all templates, whether or not use subst: for all templates created by the meta-templates, and any other matters that may come up in consideration. I thank jag123 for initially creating the subpage for the project at Wikipedia_talk:WikiProject Stub sorting/Guidelines. Even though they have been discussed, I feel that we really need to confirm everything. For that, we should discuss each issue with its own sections, and raise a list of issues that we need to nail down before really continuing on. The English Wikipedia is nearly at 500k articles. Either the MediaWiki software needs to handle stubs such that they can be found with a simple union of categories, or the sorting is done manually by Wikipedians. Personally, I think the latter is less taxing on the server load, especially when we use subst:, which I think would help the Wikipedia out, performance wise. Please make your comments at Wikipedia_talk:WikiProject Stub sorting/Guidelines. I apologize for making this somewhat of a spam notice, but since the project has more members, the project can finally decide on these important issues at hand. -- AllyUnion (talk) 23:22, 15 Feb 2005 (UTC)
Creation of stub categories
editIn an attempt to start the ball rolling on this, these are the questions I ask myself before deciding on (or suggesting) a new stub category:
- will the creating of this category help Wikipedians to find the articles (are the current stubs so buried in a host of other stubs that hunting for them is difficult)?
- does the new category cover ground not covered by other categories, or does it at least create a reasonable subcategory of an area already covered?
- will there be a significant number of stubs in this category (preferably at least 40)?
- does the new category conform with existing stub hierarchies?
- will the creation of this category help to reduce the size of any current stub categories that have a large number of stubs(*)? (* here I use an arbitrary rule of thumb - a reduction by between 10 and 50% for a category with over 200 stubs.)
Ideally an answer if yes to all five of these questions is needed for me to consider a new stub category to be an important addition to stub sorting. Grutness|hello? 23:43, 15 Feb 2005 (UTC)
- These are good; my vote is for them to become the guidelines on creation of new stub categories. I'd like to add that the significant state stubs do fit in all of these categories (5 a little bit of a stretch: the creation of a few will do this, but not one), especially number 1, which is the purpose of the stub notice anyways. --YixilTesiphon 00:00, Feb 16, 2005 (UTC)
Those are good points. When you say Wikipedians, do you mean editors or readers? Stubs are designed for editors who wish to expand articles that are related. The guideline for categories is 100 - 1000 (preferably 500) articles. 100 is a bit steep, but it should be at least 60-70, if not more. In the same respect, creating a new category to reduce an existing one should be done if it meets the above criterias.
We should also establish, for the record, what are stubs. My definition, a way to flag and group articles for editors who wish to expand articles on a related subject. They should be used in cases where the article is obviously lacking important information, and not for articles that are not yet perfect. The length shouldn't be a factor at all; the focus should be the content.
Later this evening, if it's not done, I'll start organising and creating pages to clearly address every point that needs to be addressed. --jag123 00:03, 16 Feb 2005 (UTC)
New topics
editI've added some topics that should be discussed. I'm not really sure if we should set up a new page for every entry or if every point should be discussed on this talk page. I don't want to make this too bureaucratic or complicated that people won't be participate, but I'd like to keep it organised. I'm sure no one will agree on a definition of stub, but I'm sure we can all agree on some points of what should be or what isn't a stub. At the very least, that will narrow the definition. I'm not sure if you all want to vote on every point or just discuss the topic and take it from there. Suggestions would be appreciated. Feel free to add your definition/criterias under each point. --jag123 02:43, 16 Feb 2005 (UTC)
Questioned stub categorizations for discussion
editI don't want to pick on anyone (unduly draw attention to particular individuals, that is), but there are sometimes stub categorizations that I disagree with enough to serve as potential fodder for discussion. Perhaps others will point examples out for discussion as well.
- Probiotic: this is stubbed to Category:Bacteria stubs, which is the distinguishing characteristic of probiotic foods. I'm wondering whether it would be better stubbed to Category:Food and drink stubs though, as the focus is on a type of dietary supplement. So in this case it seems that it is a choice between emphasizing composition or use (bacterial and food, respectively). Courtland 04:42, 2005 Feb 16 (UTC)
- Antifungal drug: this is stubbed to Category:Fungi stubs and I'm thinking it would be better stubbed to Category:Medicine stubs. In this case, peculiar to therapeutics, it's a choice between emphasizing the target or the context (fungus vs. medicine). Courtland 04:58, 2005 Feb 16 (UTC)
In cases like these, where several stubs could equally apply, I pick the one that is more in-line with the focus of the article (until more explicit criterias are established for stubs). In the first case, the emphasis appears to be on bacteria, not the food or medicinal aspect, so I'd leave it as is. For the second, since fungi is barely "relevant", I'd go for med. --jag123 05:24, 16 Feb 2005 (UTC)
- It's tricky, I stubbed probiotic as bacteria, as I thought microbiologists such as myself would be more likely to contribute to the article. 12:25, 16 Feb 2005 (UTC)
Lists are addressed in a policy input request below content of block below suppressed but present in the page
- Bastinado
and Chinese raping chair: both of these have gotten stubbed to Category:Weapon stubs. However, the first is a method of torture and the second began as a method of torture but is thought to have become a "marital aid", so to speak. My thinking is that neither of these constitute weapons in the traditional sense and that there isn't a good category to stub either to at this time. Courtland 06:48, 2005 Feb 16 (UTC)
- the struckout item has been re-categorized Courtland 02:22, 2005 Feb 19 (UTC)
subst considered harmful.
editI'm seriously against the use of subst. (That is, the use of including some static markup instead of dynamically including the template text.) The pages can be cached. This is an unacceptable reduction in functionality and flexibility to pander to what's really a coding problem, a database problem, but not a content problem. The kludge is unnecessary. grendel|khan 04:54, 2005 Feb 16 (UTC)
- Hey maybe it won't matter after we get all that google hardware [1] ;-), If editors have to type in a whole lot of text rather than {{cleanup}} or {{Kiwi-stub}} ect.. then people will be less likely to edit. If it works in the same way as ~~~~ it will still clutter up the page. Onco p53 12:34, 16 Feb 2005 (UTC)
Range of article count
editI propose that the minimum count of article needed to create a new stub / stub category be 75, with the expectation that the category will reach a minimum 100 article in the near future.
Any objections? If no objections are made regarding this proposal within 10 days, it will be added as a guideline for this project. --jag123 15:50, 16 Feb 2005 (UTC)
- hm. Not sure about that. I think that there is no need for a hard-and-fast number. I used 40 as standard when splitting up geo-stubs, but there were definitely cases where logic demanded a new category when there were a handful less than that - as happened with BritOT-geo-stub, which got the last 32 otherwise unsubcategorised geo-stubs. Seventy-five would be a bit high for my tastes in any case. Grutness|hello? 05:38, 18 Feb 2005 (UTC)
- I think the whole point is to pick a hard-and-fast number. I suggested 100 because it's listed here [2]. The purpose is not to subcategorised every article in a "catch-all" category so it's empty, but to create new templates when necessary, such as when a category has hundreds of articles and a new template could split that up significantly. In my opinion, 32 stubs is not enough to create a new category, especially when UK-geo-stub would have been okay. The BritOT-geo stub is specific, and it was a logical grouping, but that reasoning could be applied anywhere really. If there's no specific number, then anyone can create new stub templates to logically group 30 articles and we'll end up with 10,000 categories to maintain. --jag123 23:34, 18 Feb 2005 (UTC)
- Well, with 20,000 stubs, it'd actually be nearer 800 - and that's if all categories had only 30 (most would have far more). I'm just saying we should use common sense to know when a category is needed rather than thinking "hmm...this would be a perfect stub category, but it's only got 95 articles, so I can't create it". I'm sure I've seen guidelines elsewhere which suggest 20 as the number, but that would be far too few IMHO.
- To put it another way, the title of this page is "guidelines", not "rules" - may I suggest wording the suggestion as such - say, "Ideally, a stub template should not be created unless a considerable number of items (prefereably over x) will be stubbed with it." Grutness|hello? 00:41, 19 Feb 2005 (UTC)
- My proposal doesn't explicitly state 100 articles, but 75 with the expectation that it will eventually reach 100. Besides, it's not like there is a stub nazi going around deleting templates that have 99 articles since it's not 100. If common sense was common, we wouldn't be having this discussion. The title says guidelines because you can't really make rules here. No one is obligated to respect our criterias when making a new stub template. What if someone rearrange a bunch of stubs into several seperate categories that contain a handful of articles. It's entirely possible that it would make sense but it might complicate things.
- Another reason for the guidelines is somewhat of a foundation for AllyUnion's proposed policy. In it, there's a section on removing unapproved stubs and that bothers me because we can't agree on anything here. Who, then, decides what's unapproved or not? Admins who participate in the stub sorting project (like you)? Are we going to have to examine every stub and decide on a case-by-case basis? If that happens, then it's only a matter of time before someone starts screaming that there is bias, cabal, etc and lacking any agreed-upon "guidelines" that justify our decisions, it would not be a baseless accusation. That's not where I want to see things go.
- I know guidelines are not rules, and no one is bound by them, but at least we can point to this discussion and say "This was discussed already and the Project, as a whole, agrees that we should move in that direction". Previous discussions and previously reached consensus carries a lot of weight, which is why I'd like to see proper guidelines made. I'm not making this stuff up to constrict or govern people of this project; I'm more worried about those who aren't participants and who may not have much "common sense". --jag123 02:04, 19 Feb 2005 (UTC)
- well, okay - we're arguing for exactly the same thing then. Currently there already seem to be too much regulation, hence it already seems to be too much "bias, cabal, etc". Adding more strict guidelines will only increases that feeling. I simply think it's far more useful to have stubs created where there seems to be a valid, logical need, than to wait until it seems that there are 75 articles and an expectation that the number will reach 100. It's good to have guidelines to help us in our decisions, but to try to pin something down to pure numbers doesn't necessairily help to show where stub categories are needed and where they aren't. And I definitely think you're underestimating wikipedians if you don't think that common sense will win out. And as you implied yourself, those that don't have "common sense" will do what they want, whether it's part of the guidelines or not. So what you're in essence doing is adding further guidelines to those who have enough common sense to already be creating stubs logically. Grutness|hello? 03:20, 19 Feb 2005 (UTC)
Why are numbers important at all? The point of stub notices, which I never tire of saying, is to help people find and fix them. Thus we should categorize them based on what makes sense, not what works by some arbitrary number, especially a large one. --YixilTesiphon 03:37, Feb 19, 2005 (UTC)
Why are numbers important? To keep things reasonable. People can still read article names; there's no need to narrow everything in groups of 20-30 for people to find them. It seems like creating stub templates with pretty pictures and organising stubs has now transcended their actual purpose. 100 is not an arbitrary number [3] Regarding regulations, what regulations? What can you do or not do because of regulations? The only thing we ask is that people in the project are kept in the loop regarding creation/deletion of stubs, in case someone is working on something else. It doesn't look like there will be a consensus here, on anything, so I'm done here. --jag123 04:50, 19 Feb 2005 (UTC)
- The point of this page, surely, is to debate and find out what things there can be consensus on. In this case, maybe not - in other cases, yes there will be and is. Your input is important - including (especially?) in this case, and you have kicked off a number of big discussions.. We'll never come to realistic consensus without seeing all sides of an argument. I hope your "I'm done here" doesn't mean you won't take part further! Grutness|hello? 07:21, 19 Feb 2005 (UTC)
What not to stub
editI propose that lists, disambiguation pages (including pages that contain a list of definitions but not marked as disambiguation pages) and definitions of words, expression or slogans NOT be stubbed. Since these types of articles are expecteted to be incomplete (such as lists) or short (such as definitions) or lacking content (such as disambig pages), stubbing them is not necessary and redundant.
Any objections? If no objections are made regarding this proposal within 10 days, it will be added as a guideline for this project. --jag123 15:50, 16 Feb 2005 (UTC)
- "definitions of words" ... these should be tagged for transwikification to Wiktionary (Template:Move to Wiktionary (placed on the article's discussion page, I think). Courtland 18:30, 2005 Feb 16 (UTC)
- Agree on disambiguation pages - these should not be stubbed. And agree with Courtland on definitions. But lists? Have a look at List_of_cities_by_longitude - you telling me that's not a stub? I think the same rules should apply to them as to articles in general - if there's clearly a shortage of items on the list, or if whole sections of information are clearly missing, then it's a stub. Grutness|hello? 05:44, 18 Feb 2005 (UTC)
- The List of cities by longitude is short, but at what point will you consider it not a stub? 50 cities? 100 cities? 10% complete? 75% complete? How will you even determine that? Do you know the amount of cities on the planet? Assuming the list contains 500 items, but mostly from one area, like the United States, then there is still a clear shortage of items. Either way, a city today could be amalgamated tomorrow, or seperated from a previous fusion (like in Montreal). A list like this will eternally be incomplete and contain some incorrect entries. If stubbed, it would have to be permanent, and that defeats the whole purpose of stubs. --jag123 16:32, 18 Feb 2005 (UTC)
- Why would an incomplete list be always a stub? The decision of when to remove the stub message would have to be arbitrary, that much is obvious, but there is an "incomplete list" template for those circumstances. There is no way that such a list could be completely canonical, but it could at least be comprehensive - common sense should prevail. A "list of cities" that contains only two cities is clearly a stub. One with 100 of the largest cities from around the planet would clearly not be a stub. One with 500 cities, all of them from the US, should be renamed "List of United States cities by longitude". Grutness|hello? 00:36, 19 Feb 2005 (UTC)
- One way to look at it is whether a) items in the list are wiki-linked and b) those wiki-links are live. There have been a very small number of lists I've seen where most of the items are wiki-linked but those links are only placeholders (red) or that aren't wiki-linked at all. I think I'd stub those list types; but a short list with all items linked to articles (or stubs), no ... that's an organizational assist and isn't deserving of "stubbiness". Courtland 00:42, 2005 Feb 19 (UTC)
- Here's a borderline case perhaps: List_of_Alaska_state_highways; here we have a pair of overlapping lists where 1/3-to-1/2 of the wiki-links lead no-where (red). I did some investigating and found support for a redirect from Route 98 to Klondike Highway and that level of work on all those red-links or removing them would drag the list out of its stubbiness ... in my opinion. By the way - I didn't put the stub-notice; that was there when I found it. Courtland 01:43, 2005 Feb 19 (UTC)
- Like I said above, I don't believe in common sense or arbitrary decisions. What if someone decides to stub every single list? What you define as stub-worthy might differ from their definition, and since it's all arbitrary, anyone can do anything really. I suppose we could agree on a set of criteria for lists because I do see some usefulness in stubbing *some* lists, but I really don't care that much about lists to bother arguing over it. Expanding articles is totally different from expanding lists. If there's an incomplete list template, that should be used. --jag123 02:11, 19 Feb 2005 (UTC)
- If... Is anyone likely to stub all lists, simply because they are incomplete? If they do (unlikely) simply remove the stub messages. It's quite clear that some incomplete lists are stubs and some aren't, and that there cannot be a single cut-off point at which you can say "lists below a set size are automatically stubs". As you pointed out further up, it's highly unlikely that anyone would know the number of cities in the world, so it's impossible to set a criterion in actual number or percentage terms as a cutoff. But it's equally clear that some lists are so far below an acceptable standard that they cannot not be considered stubs. The "Alaskan highways" list has loads of information - I don't know whether it's close to listing all of Alaska's highways, but it is a detailed list. The "Cities by longitude" list had two cities on it - that's not a case of "could do with some expansion" - it's a stub. Having said that, I'd agree that 99% of lists shouldn't be stubs. Grutness|hello? 03:07, 19 Feb 2005 (UTC)
- I'll admit it - I was wrong. Every list I've seen cone through category:Stub has simply needed a listdev or similar. Lists probably shouldn't be stubs after all <:) Grutness|hello? 10:35, 15 Mar 2005 (UTC)
One thing that (IMO) should never be a stub (though I've noticed one or two creeping into Category:Stub) - a category. How can a category be a stub? I suppose it's analogous to the list situation, except that at least with additions to a list, they are in the form of edits, and it would be easy to notice that ther's a stub message there. With a category that wouldn't happen. Grutness|hello? 09:44, 20 Feb 2005 (UTC)
Non-articles as stubs
editCategories require {{popcat}}
Lists may use any of Wikipedia:Template_messages/Lists (However, {{expandlist}}, {{listdev}} and {{Stublist}} should probably be merged, I'll submit them to WP:TFD asap.)
- How about Disambiguation pages - unless they form part of an article on one of the items disambiguated? That is, a simple disambiguation list is not a stub (although it might conceivably get a listdev template or similar), but an article which is about one subject but adds at the bottom "this term also could refer to:" followed by a list is fair game. Grutness|hello? 00:00, 18 Mar 2005 (UTC)
- If I find a stub on a disambiguation page, I remove it. If I find a stub that is a non-templated disambiguation page, I remove the stub and add the disambiguation message. If I find a stub on a list, I remove it; I generally do not add an expand-list type of message, but perhaps I should; that's a toughy because most lists are expandable, some (almost) infinitely so. Courtland 01:17, 2005 Mar 18 (UTC)
Use of subst
editSubst:stub
editI propose that "subst:" not be used. There does not appear to be much support for it's use and other alternatives should and can be explored.
Any objections? If no objections are made regarding this proposal within 10 days, it will be added as a guideline for this project. --jag123 15:50, 16 Feb 2005 (UTC)
good move. All that substub does is break Category:Stub in two according to length of article - hardly the most useful criterion. Grutness|hello?
Comment: I wonder if there is a misunderstanding here? What are we talking about: The "Substub" template {{substub}} or the method of substituting templates to static wikitext with {{subst:Template name}}? -- grm_wnr Esc 14:54, 16 Mar 2005 (UTC)
- Oops - my bad. Please ignore my comments above (although the thread that developed from it does indicate that there is some antipathy to Template:substub as well!). I hope, for that reason, that Courtland and YixilTesiphon do not mind that I have split this section into two separate parts and moved their comments into the next section to reflect the two different topics! Grutness|hello? 07:31, 17 Mar 2005 (UTC)
Use of substub
editI propose that "subst:" not be used...jag123
(User:Grutness misread this as "I propose that "substub" not be used - the following thread resulted)
- good move. All that substub does is break Category:Stub in two according to length of article - hardly the most useful criterion. Grutness|hello?
- Agree I've wondered why someone would use "subst:" when "stub:" seems to cover things all the way down to a handful of words. Will a bot be run to convert them, or will this be a manual re-categorization process? Courtland 04:00, 2005 Feb 18 (UTC)
- I don't know; though I do Agree to a certain extent in that subst just seems to be a more pathetic plea for expansion. --YixilTesiphon 23:00, Feb 20, 2005 (UTC)
Update: It seems, from inspection of the Category:Substubs for the last month, that few new substubs are being labelled as such - the category's population seems to be a stable 3000 or so. I am trying to work through at least ten a day, merging, vfd'ing or upgrading to stub subcategory wherever appropriate. See also note below at Wikipedia_talk:WikiProject_Stub_sorting/Guidelines#Substubs. Grutness|hello? 07:31, 17 Mar 2005 (UTC)
Category criterion
editA suggested addition to the "what is a stub" - irrespective of length, any article is a stub if it hasn't got at least one category (Ideally, a category should be found rather than a stub message added, but this isn't always possible). This ensures that all the articles are listed somewhere other than the complete list of articles, rather than "free-floating in cyberspace". This, technically, could come under "article missing important information", since its place in the category hierarchy is important information. Grutness|hello? 02:39, 18 Feb 2005 (UTC)
- Categories are cheap and easy to make and I'd seriously be surprised if there is a topic that doesn't already have related categoriess. I believe there is an uncategorised pages project (if not, there are definitely people who do this) which handles the task of categorising pages. If anyone finds an uncategorised page, I certainly encourage them to categorise it but I think stubs should be left for articles that are lacking content. --jag123 02:57, 18 Feb 2005 (UTC)
Number of stub categories
editIf there are, in fact, 20,000 stub articles then interestingly enough if they are evenly divided into stub categories each with the minimum of 100 articles that would be 200 categories... Which happens to be the number of categories displayed at once at category:stub categories. What a coincicdence. RJFJR 03:50, Feb 18, 2005 (UTC)
Division of biography stubs?
editInspite of the fact that I'm a big proponent of merging stub categories I'd like to hear if there is interest in splitting bio-stubs into two categories: modern-bio-stub and historical-bio-stubs. It sounds like they'd be worked on by editors with different interest. I'd suggest that if we did then we could arbitrarily make 1900 the dividing line. People who's notable contributions were mainly before then would be in historical and if after that would be modern. Comments? RJFJR 03:50, Feb 18, 2005 (UTC)
- Judging by comments on stub sorting in general there seems to be more of a move to divide bio-stubs by profession and/or nationality. I agree that they need to be split up, and thie historical criterion is a viable one, but I think I'd prefer the other ways of splitting the stubs up. Grutness|hello? 05:34, 18 Feb 2005 (UTC)
- Me too, I'd prefer other ways of sorting. There's some problem with picking an arbitrary dividing line as there are historical eras that are better categoricals for binning biographies. For instance, rather than 1900 for the dividing line between "old" and "new", a fuzzier "Industrial Era" category for modern biographies might be better. Further, one could go farther back along generally accepted era boundaries in the West and East, etc. This provides and orthogonal and alternative axis for sorting to the profession/place-in-history categorization that is currently in place; I wouldn't say it replaces that pre-existing axis. Courtland 00:26, 2005 Feb 19 (UTC)
- I prefer other ways of sorting. Dividing people by era seems unnatural. I think it's more likely that someone will have knowledge of scientists or army generals or authors who've lived in any century than someone with knowledge of all kinds of people who've lived in so-and-so time period. --jag123 00:36, 19 Feb 2005 (UTC)
- Yes, it should be by profession or nationality. --YixilTesiphon 23:01, Feb 20, 2005 (UTC)
Beginner Mind
editAs a brand new user of Wikipedia,I would like to say that it is an amazingly wonderful and valuable thing you are all creating. I came here for information on English pronunciation, particularly phonology and I found a lot of useful stuff. I also found some valuable related info on a page noted as a "stub," which I followed to get a definition and later encountered this discussion. I have no technical expertise to offer, but in the debate on whether to kill stubs or not, it does not seem like a good idea since much valuable information might be lost. Would it be wishful thinking to hope for gathering this related info into a master document? Remember, I said I brought no technical expertise, but perhaps there is something in this beginner's perspective that can help you in this complex discussion. Keep up the good work! --Lamusette 19:41, 18 Feb 2005 (UTC)
- Don't worry - at least on this page the people are involved with sorting stubs rather than removing them - to make it easier for editors to find the articles and expand them. Wikipedians tend to have what are known as 'inclusionists' (who want to keep as many articles as possible) and 'deletionists' (who think that tiny articles are worthless and should be deleted). Thankfully there are enough inclusionists that most of the articles deleted have something majorly wrong with them (like vanity pages) or are so small as to have no information at all. Oh, and thanks for the compliments! Grutness|hello? 01:53, 19 Feb 2005 (UTC)
Substubs
editNot sure whether this belongs here or elsewhere(I'm going to cross-post it to Criteria as well for this reason), but this came up in discussion recently on tfd ater Template:PBS-substub was discovered. The suggestion was made (which I agree with) that substubs should not be subdivided in the same way as stubs.
Substubs seem to fall into four types: dicdefs that should be moved to Wikipedia, scraps of information which should be merged into larger articles, potential vfds that will never amount to anything, and potential articles which should be at the very least expanded into stubs. At some point, some effort from WP:WSS should probably go into sifting through Category:Substubs to work out which ones go where and doing something with them. Those that can be made into proper stubs, all well and good, they can them get changed into whatever subcategory of stub is appropriate. The rest should be dealt with according to their needs. Grutness|hello? 22:56, 15 Mar 2005 (UTC)
Porn-stub and Bio-stub: sorting between these
editI've started sorting Bio-stubs for porn actors/actresses into the Porn-stub rather than double-stubbing them to Porn- and Bio-stubs. What do you think about the merits of this decision? Would you do the same? Thanks for your input. Courtland 21:43, 2005 Mar 24 (UTC)
Sounds fine. RJFJR 22:51, Mar 24, 2005 (UTC)
- I'm beginning to think that anything that reduces the size of Category People stubs is a good move. Go for it. Grutness|hello? 00:41, 25 Mar 2005 (UTC)