This page is within the scope of WikiProject Short descriptions, a collaborative effort to improve the coverage of short descriptions in articles on Wikipedia. If you would like to participate, please visit the project page, where you can join the discussion and see a list of open tasks.Short descriptionsWikipedia:WikiProject Short descriptionsTemplate:WikiProject Short descriptionsShort descriptions
This page is within the scope of WikiProject Wikidata, a collaborative effort to improve Wikipedia's integration with Wikidata. If you would like to participate, please visit the project page.WikidataWikipedia:WikiProject WikidataTemplate:WikiProject WikidataWikidata
These are not meaningless, If a country has a president and a prime minister, separating who is the Head of state and Head of government is important, this is the function of such short descriptions. Knowing which is HoS and HoG is important if you are in the search bar or wikilinking.
If a short description serves a useful purpose it is valid. If someone is removing useful short descriptions, ask them to stop. If they do not, it may be a case of disruptive editing, notify an admin as a block may be necessary. Cheers, · · · Peter Southwood(talk): 06:56, 22 November 2024 (UTC)[reply]
SDs for ligaments
Working on pages in "articletopic:medicine-and-health", I've noticed that many ligaments lack SDs and that the existing SDs of ligaments vary in their naming scheme more than other anatomical structures. Most commonly they're either "Ligament of xyz" or "Ligament between x and y" which makes sense considering the ligaments' connective nature (in comparison to e.g. nerves which are often more strictly localized and tend to follow only the first naming scheme).
I don't think there is, so you should follow the general rules at WP:SHORTDESC. Two important aspects that often get overlooked on medical articles are that the SD should be understandable to the inexperienced reader - WP:SDJARGON; and that it should be short - WP:SDSHORT. It's not a definition - WP:SDNOTDEF. If you have some specific articles that you'd like advice on, please say. MichaelMaggs (talk) 15:23, 9 November 2024 (UTC)[reply]
I've just come across someone changing the "and" in a short description to an ampersand (see diff at Andy Kim (politician)), presumably to save characters, and I was wondering if that is in line with the general consensus here. MOS:& prescribes the use of "and" in normal text and headings but also states that [e]lsewhere, ampersands may be used with consistency and discretion where space is extremely limited (e.g., tables and infoboxes). I don't think this specific issue has been discussed before (although there was an adjacent discussion in July 2023, Wikipedia talk:Short description/Archive 15#Using punctuation to shorten description, where Jonesey95 argued to use "and" instead of an ampersand), but whatever the general consensus is, this may be something to make explicit in WP:SDFORMAT following this discussion. (Ping Losipov who made the linked edit.) Felida97 (talk) 19:06, 10 November 2024 (UTC)[reply]
@Felida97 what you mentioned in your message is exactly the reason why I changed the "and" to an ampersand. Sometimes when something is over the limit (like if a short description is 42 characters long), I would change it to an ampersand so it goes down to 40 characters (the limit). I don't know if you took it this way, but I can assure you I wasn't trying to vandalize the article. I just thought it would save space, again like you mentioned. If the consensus is to revert it back to "and", feel free to revert me on that article. I won't object. Losipov (talk) 19:18, 10 November 2024 (UTC)[reply]
@Losipov: Sorry, if that maybe came across in a wrong way, I had not at all suspected that your edit was ill-intentioned (a quick look at your user page and contributions was enough for me conclude that). Felida97 (talk) 19:36, 10 November 2024 (UTC)[reply]
Please don't change "and" to "&" to go from 42 to 40 characters. That is not a useful edit. There is a reason that "&" is discouraged: it is much less reader-friendly than "and". Since the creation of local short descriptions, there has been a persistent misconception that there is a hard limit of 40 characters. It is just not so. Editor time would be MUCH better spent in adding short descriptions to articles that have none than in tweaking existing short descriptions that are fine.
The limit of 40 characters is wise, but not absolute, so using 2 extra characters for "and" is not that evil. Mind you, using just a comma is even shorter! Probably best to use "and" as advised by the MOS — GhostInTheMachinetalk to me19:47, 10 November 2024 (UTC)[reply]
@Jonesey95, @Rublamb, @Patar knight: Do you have an opinion on the use of a comma, both vs. "and" and vs. "&" (for an example, see follow-up diff by GhostTheMachine at the article mentioned above, Andy Kim (politician), where the SD is now 41 characters as a result; just to be safe, I don't mean to call you out or anything and don't think that your edit wasn't well-intentioned either, GhostInTheMachine)? (Sorry, if this question is considered as unnecessarily pedantic; I just thought, we might as well clear this up, too, while we're at it.) Felida97 (talk) 15:48, 11 November 2024 (UTC)[reply]
I don't mind it and prefer it to "&" but I would probably use "and" for clarity. Sliding over 40 by two characters is not a problem. Rublamb (talk) 17:35, 11 November 2024 (UTC)[reply]
I think it's still better to use "and" for clarity since the character save is still not much. However, I think it is better than the ampersand since it is more commonly used and is an additional character saved. -- Patar knight - chat/contributions01:25, 12 November 2024 (UTC)[reply]
40 isn't a hard limit and the character save here is minimal, so we should generally we should be following the MOS here as a standard. If two characters is the difference between critical information being pushed back past 40 characters, the SD probably just needs to be majorly overhauled in a way that an ampersand doesn't fix. -- Patar knight - chat/contributions14:41, 11 November 2024 (UTC)[reply]
People should take the trouble to actually read the whole of WP:SDFORMAT which states that less than 40 characters is/was a common length for short descriptions. It is a statistic, not a rule, and it is probably skewed by the many thousands of automatically generated short descriptions we used to get things going when the project started. Many of the shorter descriptions are acceptable but not very good, and improving them will usually result in a longer short description. In most cases this is just fine. Cheers, · · · Peter Southwood(talk): 05:55, 22 November 2024 (UTC)[reply]
Not sure why you would characterize it this way: it's been made stated repeatedly that 40 is specifically chosen because different display contexts may truncate an SD longer than that in different ways. Remsense ‥ 论05:58, 22 November 2024 (UTC)[reply]
There are a few apps which arbitrarily truncate short descriptions. They are not helpful and were not chosen by the Wikipedia community, so have no authority over what we choose to do. It has indeed been wrongly stated many times that 40 was specifically chosen, but repetition of an error does not make it true. Cheers, · · · Peter Southwood(talk): 06:06, 22 November 2024 (UTC)[reply]
The dichotomy is a bit strong: I'm part of the community, aren't I? "I have to keep telling members of the community that a number isn't agreed to by the community" is a slightly thorny rhetorical position.
More importantly, I happen to think UX is important, and whether I chose the number or not I'd like readers using an app to have a pleasant time as much as anyone, especially as users of those apps are generally the demographic most in need of SDs. Treat it as a cynical fait accompli if you'd like. Remsense ‥ 论06:17, 22 November 2024 (UTC)[reply]
As you say, you are a part of the community. A larger part of the community agreed that there is no hard limit. If you think you represent a larger part of the community, you can start an RfC for a proposed change, but you may find that consensus has not changed as the usefulness of the proposal is low. Cheers, · · · Peter Southwood(talk): 07:52, 22 November 2024 (UTC)[reply]
I would prefer that the developers of apps pay more attention to making the UX user friendly, so we can build the encyclopedia without artificial externally applied restrictions. Cheers · · · Peter Southwood(talk): 07:57, 22 November 2024 (UTC)[reply]
RFC: should WP:SDNONE apply to articles on subnational entities and municipalities in the title? If so, what should be the format?
I have asked this question multiple times and have received no complete consensus to change WP:SDNONE to apply article names such as History of Nottinghamshire and Watersheds of Illinois. When this is resolved, please update WP:SDNONE to comply with this ruling. This RfC is necessary since there are thousands of articles where this applies where there is no consensus for use of WP:SDNONE. Additionally, the guideline is extremely short and vague for something that applies to hundreds of thousands of articles.
Potential options:
Yes, articles on subnational entities and municipalities in the title should have "none as a short description"
Only for cities and subnational entities that an "average reader would know" (which I have seen brought up in the past) should have "none" as a short description (most readers know about Delhi, Texas, and Moscow, not so much Kilgore, Texas, and Ncojane) (This has been brought up in the past by Remsense)
No, none of these apply for WP:SDNONE (a format will have to be provided for what should be recommended in the new guideline)
Please explain why "History of Nottinghamshire" and "Watersheds of Illinois" are not "sufficiently detailed that an additional short description would not be helpful", which is the criteria for WP:SDNONE (not that either article has an SD at all at the moment) - Arjayay (talk) 22:22, 15 November 2024 (UTC)[reply]
From earlier conversations about this (not my thinking, these are OTHER PEOPLE'S quotes:
WP:SDNONE may work, but is not very user-friendly if the term is likely to be unfamiliar to the reader. Prefer "Aspect of Chinese history", "Aspect of Welsh history @MichaelMaggs
I whole-heartedly agree all "Aspects of history" should be gone asap. Like I said below, I was going around and changing those (and similar SDs) to "none" but ran into articles that didnt seem to make sense to be "nones". It was actually while working on SDs in Australia that got me thinking about it, specifically History of the Northern Territory. Unless you know that's a territory of Australia, "none" doesn't make sense. Wouldnt "History of the Austrailian territory" (or something to that effect) be better? @Masterhatch-1ctinus📝🗨23:16, 15 November 2024 (UTC)[reply]
Aspects of history is not usually very useful, but none may be worse. It depends on the specific title. History of the Australian territory is not efficient as it only adds one useful fact, and four of the five words are the same as in the title, but it is an improvement as it is an important fact. Nevertheless it may be possible to improve tit further some day, and better is better. Cheers, · · · Peter Southwood(talk): 06:37, 16 November 2024 (UTC)[reply]
Comment I don't know what the solution is but I do think SDNONE has gone too far. For example, unless you know Saskatchewan is a province in Canada, how does SDNONE help in List of communities in Saskatchewan? Wouldn't something like "Communities in a central Canadian province" (or something like that) be more helpful? Don't get me wrong, I support the concept of SDNONE. SDNONE works well for articles such as History of Canada. That's my two cents. Masterhatch (talk) 02:05, 16 November 2024 (UTC)[reply]
Someone is doing that. It is not SDNONE, it is someone's interpretation and their actions that go too far The advice in SDNONE does not encourage that behaviour. · · · Peter Southwood(talk): 11:11, 16 November 2024 (UTC)[reply]
Another comment Recently there was a tussle over SDNONE at List of Alberta provincial highways. I was on the side of defending SDNONE per the rules set out at WP:SDNONE. User:Evelyn Harthbrooke was on the other side. I think she brought forward another example of where SDNONE might not work. Unless you know Alberta is a province in Canada, how does SDNONE help? After the debate I looked at other highway list articles in Canada and I saw List of Ontario provincial highways wasn't "none". I didn't change it to none because I was starting to second guess if none fits there (Same as the Northern Territory SD caused me pause a while back).Masterhatch (talk) 02:21, 16 November 2024 (UTC)[reply]
Explanation The use of "none" as a short description does not overrule any local consensus to add a non-none short description for any page. If there is a short description available that provides additional useful information of value to the reader it may be boldly added. If others disagree it is a talk page discussion matter. Look at the functions of a short description. Does the proposed short description provide better functionality/reader value? If so, it is a legitimate option. The format depends on the information in the short description. There is no fixed format. A WikiProject can recommend a format, but they cannot overrule consensus for the specific article, particularly if it is of interest to more than one project. As it happens, I am aware of wher Nottinghamhire and Illinois are, but a substantial number of readers probably do not, so a short description could usefully proviede that information. A lot of readers might also not know what a watershed is. The more obscure the topic, the more useful a short description can be, so global cities and federal states probably need less clarification than relatively unknown places. The more relevant question might be whether SDNONE is clear enough that it does not get misused. Cheers, · · · Peter Southwood(talk): 06:08, 16 November 2024 (UTC)[reply]
Done. I spent a lot of time adding short descriptions, and when I dealt with any page dealing with... local elections, subtopics of cities (History of so and so), this came up EXTREMELY often and I had to ignore them all. -1ctinus📝🗨13:11, 16 November 2024 (UTC)[reply]
I think 3 since we cannot and should not assume that most readers are familiar with the subnational entities of the 200~ish various countries that exist, so labeling them as such helps differentiate them from similarly titled articles, which is an WP:SDPURPOSE. For example, typing "History of Mo" into the search bar gets met articles that include countries like Mongolia and Morocco, but also subnational entitles like Montana and Montreal, and board games like Monopoly. At the very least, we should be labelling subnational entities as such for the benefit of the reader.
However, there are many aspects of this RFC that are pretty unclear. Is this only about sub-national entities? How are we defining "global cities" and what is a "federal state"? What about subnational entities that incorporate the national name (e.g. Australian Capital Territory)? These things need to be clarified.
If we are opening it up to all instances of "X in Y", I would argue that we are probably overusing SDNONE even for national level entities. In an ideal world, everyone will know what is a country and isn't, the reality is that lots of people are just not very good at geography and an SD would be helpful. For example, it would probably be helpful if "History of Eswatini" had a SD of "National history" or "History of the/an African country" for the same reasons why subnational entities should be labelled. There is no harm in pointing out to readers what things are national-level in scope. -- Patar knight - chat/contributions07:03, 16 November 2024 (UTC)[reply]
SDNONE allows a short description of "none". It does not in any way suggest that when a useful short description is available it should not be used. On the other hand, removing a useless short description under SDNONE is OK. Until someone has a better short description to substitute for "none". · · · Peter Southwood(talk): 11:17, 16 November 2024 (UTC)[reply]
Yes, SDNONE doesn't explicitly ban anything, but practice on this page is to advise against using a SD for things like "X of Y" where Y is a country. (e.g. [1][2][3]) Obviously not every X is the same here (e.g. something like "National economy" or "National flag" are obviously natural SDs), but some consensus/guidance is probably warranted, even at the national level. -- Patar knight - chat/contributions00:06, 18 November 2024 (UTC)[reply]
That practice should be discontinued, as it does not take the purpose and uses of short descriptions into account. Competence is required, not a formulaic rule. Wikipedia avoids unnecessary rules because there are people more concerned with rules than functionality, who rely on rules to win fights which are counter to the purposes of the encyclopedia, and will quote selected parts of rules and ignore those parts which do not serve their agendas. Cheers, · · · Peter Southwood(talk): 03:19, 18 November 2024 (UTC)[reply]
This is one of a series of questions essentially asking for a definite yes/no 'ruling' as to whether "none" is always OK for a particular form of article title, for example for "List of ...", "History of ..." and similar. No such ruling is possible, as the test is based not on the form of the title, but on whether an additional short description would or would not be helpful. That requires editorial judgement on an article by article basis. The answer for any particular article depends not on the form of the title, but on whether something better than "none" can be created which advances the purposes. "none" should never be used when something more informative could be provided.
That said, the fact that SDNONE isn't always well understood does suggest that we should expand the guidance and add some specific examples of good practice. If you can give me a few days, I'll work up a full proposal and post it here for further discussion. MichaelMaggs (talk) 00:50, 18 November 2024 (UTC)[reply]
Sounds great, just try to keep it 1. not Anglocentric and 2. avoid partially redundant short descriptions. A policy that affects like a million articles being two sentences is frustrating and unhelpful for people who add lots of short descriptions. -1ctinus📝🗨01:11, 18 November 2024 (UTC)[reply]
Sometimes a partially redundant short description cannot be avoided without making it considerably longer. This can happen because titles are not required to be chosen with a potential short description in mind. We must deal with reality, and reality is not constrained by our rules. · · · Peter Southwood(talk): 03:24, 18 November 2024 (UTC)[reply]
Previous proposals created an "obscurity test" that was based on what a "typical English reader" would know, which meant that english speaking places were more likely to qualify than non english places. -1ctinus📝🗨11:09, 18 November 2024 (UTC)[reply]
What a strange criterion. Our readers include first language English speakers and people who have only a basic knowledge of English, and while we should not be dumbing down our content, we want to be as accessible as reasonably practicable to all comers. It is safer to assume that the average reader has no local geographical knowledge at all, and those who do, may not know the place by an English language name in any case. · · · Peter Southwood(talk): 11:36, 18 November 2024 (UTC)[reply]
After promising to post a proposal in "a few days" I've had hardly a moment to sit down and it's been impossible to find the time to devote to this. I still plan to post something, but it won't be before next week. Apologies. MichaelMaggs (talk) 10:49, 21 November 2024 (UTC)[reply]
I think 3 and basically leave it as is, flexible as needed or wanted. I would tend to think that when an article such as History of Nottinghamshire starts with a description it has shown that an actual short description would be both useful and easy. Cheers Markbassett (talk) 07:30, 3 December 2024 (UTC)[reply]
Reversion of explanatory note
Hi MichaelMaggs, you reverted an explanatory note stating This is a statistic, not a recommendation to arbitrarily shorten short descriptions to 40 characters or less, and provides no information on the quality of these short descriptions which to the best of my knowledge is simply a description of fact and not a change to the guidance, with the edit summary Strongly disagree, sorry. Perhaps this is just to discourage editors from shortening SDs of 42 characters, but it negates the many months of discussions we have had about SDSHORT, and takes us right back to the early days of interminable arguments with editors insisting that their 50 or 60 word SD is fine because "I need it, and who's to say what 'short' means?". SDSHORT is the single most important guide for keeping SDs within reasonable bounds, and it absolutely must not removed. Is there a part of the explanatory note that you consider incorrect? and why do you strongly disagree with its presence as a footnote?
As far as I am aware, the percentage of short descriptions exceeding 60 characters was not in disputed at the time it was listed, though that may have changed, and I would cheerfully accept a more recent value if it is available.
It is clear from the language that the statement Fewer than 3% of short descriptions are longer than 60 characters, and short descriptions longer than 100 characters will be flagged for attention. is descriptive of the situation and does not prescribe any action, particularly not to reduce the length of a short description simply because of length. We have no statistics on the relation between length and quality of short descriptions, but it is likely that many very short descriptions are sub-optimal.
The header to that item, [Each short description should:] be short – no longer than is needed to fulfill its functions effectively clearly implies that length is secondary to function, and as a general principle, our policies and guidance put function of the encyclopedia as a high priority.
We do accept short descriptions longer than 40 characters routinely, and we do accept longer short descriptions when they are functionally desirable.
As a general principle we do not make restrictive rules without good reason, and we try to avoid making rules overly verbose because of tl;dr, but sometimes people will not read the guidance and stop when they find a part of it that they can wield as a weapon to attempt to browbeat others to comply with a personal agenda which may not align with the intention or letter of the guidance text, or the purposes of the encyclopedia. Limiting the length of a short description to any arbitrary length is a restrictive rule.
This is a big topic and has the potential to start yet another huge discussion on the importance or otherwise of 40 characters. I'm happy to contribute again to that, if needed, but time is short at the moment and if you'd agree I would prefer to concentrate on "none" and the new examples of best practice before going back to issues of length. MichaelMaggs (talk) 17:27, 28 November 2024 (UTC)[reply]
No great rush, but there is always someone mis-applying the guidance to truncate useful short descriptions because they are longer than 40 characters, which is just wrong. · · · Peter Southwood(talk): 04:54, 29 November 2024 (UTC)[reply]
Our guidance on when to use "none" as a short description could usefully be expanded. The current wording – some article titles are sufficiently detailed that an additional short description would not be helpful – is inadvertently encouraging the overuse of "none" on articles whose title is obvious to readers familiar with Western popular culture, but which may not be at all clear to other English speakers worldwide. I suggest re-focusing attention on the primary purposes of short descriptions, and providing some practical examples of good practice. There have also been many queries about best practice in various fields, and we can sensibly extend our general list of examples.
It would help discussion if you were to create a second subpage that contains the current text of the relevant sections. Then we can use Special:ComparePages to generate a custom link showing the differences between the current text and the proposed text. – Jonesey95 (talk) 19:14, 28 November 2024 (UTC)[reply]
Your draft is amazing, but it needs feedback and polish. I’ll start with two simple problems with the example short descriptions:
1. Why aren’t years included in the biography example short descriptions?
2. Why are the gramatical articles (an or a) excluded in the examples?
Ill keep adding nitpicks when I see them, but I like the rescoping. The previous RfC is stable and essentially proved to me that it needs to be completely blown up and rewritten. -1ctinus📝🗨19:55, 28 November 2024 (UTC)[reply]
3. the example given for "List of Star Trek: Voyager episodes" being "American TV show series" seems like it is missing a few words. I assume this is an error? -1ctinus📝🗨20:01, 28 November 2024 (UTC)[reply]
Re question 2 about "a" and "an": Have you read this page's guidance on formatting? Specifically: avoid initial articles (A, An, The) except when required for correct grammar and meaning. Also see the first bit of WP:SDEXAMPLES. – Jonesey95 (talk) 21:01, 28 November 2024 (UTC)[reply]
the emphasis is on the "correct grammar." I'm not sure if "Performances by American actor" is grammatically correct. I'm pretty sure an indefinite article is required (Performances by an American actor) for that to be true, but I'm not a linguist or the grammar police.
For greater visibility, I've moved the following discussion (down to "end of copied discussion") here from User talk:MichaelMaggs/Draft SD guidance. No change to content, but I've tweaked the header levels and tabs.
No more than 40 characters
I object to this as people will use it as an excuse to shorten descriptions regardless of functionality. This is already a problem. There is no broad consensus for a hard limit, and 40 characters is too short for a significant subset of short descriptions. · · · Peter Southwood(talk): 05:07, 29 November 2024 (UTC)[reply]
Propose these should be titled "Suggested short descriptions" as they are not necessarily optimum, for example, does every reader know what the Olympic Gmes are? If so, should "Olympic Games – Major international multi-sport event" not have a short description? Do all readers know what "Ghana – Country in West Africa" is? I would not be surprised if a significant percentage of American readers do not know that Ghana is a west African state. If a suggestion is labelled a recommendation, some people will use it as an argument to remove a short description even if it is useful to some readers. · · · Peter Southwood(talk): 05:27, 29 November 2024 (UTC)[reply]
Done. Changed Ghana to India to avoid questions about how well that country is known. Olympic Games should be OK, though: it's not possible to gloss everything, and attempting to overdo it makes it sound like mansplaining. MichaelMaggs (talk) 09:33, 29 November 2024 (UTC)[reply]
"Not supposed to be a definition"
There are users who interpret this as "Must not be a definition", which is wrong. It does not need to be a definition, but if it is that is OK. A short description that also happens to be a reasonable definition of the topic is acceptable and should not be changed just because it happens to be a definition. This has happened. · · · Peter Southwood(talk): 06:06, 29 November 2024 (UTC)[reply]
I think that's fine, and the wording here is unchanged. The link is to WP:SDNOTDEF which explicitly says "There is no objection to an otherwise-suitable short description that also happens to work as a definition." Can't really be clearer than that. MichaelMaggs (talk) 09:46, 29 November 2024 (UTC)[reply]
Claim that "Concept" is better than a long description that attempts to define the topic, contrary to SDNOTDEF
Not necessarily always true. Please do not make statements that will be interpreted as universally true when they are not. Particularly when they rely on an undefined condition which is open to interpretation (what is a long description in this context?) · · · Peter Southwood(talk): 06:16, 29 November 2024 (UTC)[reply]
Added 'excessively'. Still undefined of course, but may help. Again, a separate discussion (not here) probably needs to be had to tie down the meaning of 'too long'. MichaelMaggs (talk) 09:52, 29 November 2024 (UTC)[reply]
Not yet. As far as I can see, you have not yet created a second subpage that contains the current text of the relevant sections so that it is easy to compare the two pages using Wikipedia's built-in diff function. Please do not ask us to scan line by line to compare each word and clause. – Jonesey95 (talk) 00:33, 5 December 2024 (UTC)[reply]
What happens in the case of redirect articles that don't have their own SD?
I don't have the means to check this myself, so may I beg a favour from someone who does, pleaase?
Right now, (this is just an example, it is the principle that I am trying to establish) the article Prebunking is a redirect to Fake news: it does not have its own SD. So my question is this: if someone searches for "prebunking", do they get (a) no SD or (b) the SD for "fake news"? 𝕁𝕄𝔽 (talk) 18:09, 29 November 2024 (UTC)[reply]
Prebunking does not appear to have a short description. You can verify this by going to Page information for that page and looking for "Local description" or "Central description". You can check it a different way by going to its listing on a category page and using the "Show SDs" button generated by the "shortdescs-in-category" script; when you do that, you will see "no shortdesc" next to "Prebunking". – Jonesey95 (talk) 18:42, 29 November 2024 (UTC)[reply]
Depends on the skin. In skins that do show short descriptions (Vector 2022 and Minevra Neue), when a redirect title is typed into the search bar, only the target article shows up in the dropdown (so when you search for Prebunking, only Fake news shows up in the dropdown). Even if a short description is added, the dropdown just won't display the redirect page. In the other skins (Vector legacy, MonoBook, Timeless), the redirect does show up, but those skins don't display the short description. Liu1126 (talk) 19:20, 29 November 2024 (UTC)[reply]
Liu1126 did a much better job of actually answering the question. When I search for "Prebunking" in Vector 2022 (the default skin), I do not see "Prebunking" in the results. I am shown "Fake news" with its description. It appears from Liu1126's response that nobody who searches is shown any short description for "Prebunking". Let us know if that does not answer your question. – Jonesey95 (talk) 20:32, 29 November 2024 (UTC)[reply]
As far as I know you have to click on the backlink to the redirect at the top of the target page to see the redirect page directly, where the short description is visible if you have chosen one of the options that display it, and there is a short description to be seen. · · · Peter Southwood(talk): 17:22, 5 December 2024 (UTC)[reply]
That was the case when I first looked, but it has one now. Someone has added it, which fixes this particular case, but the situation regarding display of target SDs where redirects are displayed in annotated links and on category pages probably remains. · · · Peter Southwood(talk): 06:33, 6 December 2024 (UTC)[reply]
I have just fixed two cases at Sanewashing where this template provided incorrect descriptions for Prebunking and Steelmanning, both of which redirect to a page with a nearly opposite meaning:
Prebunking – The process of debunking lies, tactics or sources before they strike
Steelmanning – The opposite of a straw man argument
While I can see the value of not having to write these annotations manually it seems to me like this is quite a dangerous template in its current form. Would it perhaps be reasonable to display no annotation at all when redirects are linked? Even checking the output manually at the time of an edit adding this template looks to me like it wouldn't be sufficient, since an article might be redirected at any time. I'm not sure how useful the tracking category is given that it is 2900 pages strong (and if the intent is to ignore correct cases it can only grow).
I see there has been discussion of this issue above but it doesn't seem like that was ever resolved. Tollens (talk) 06:56, 29 November 2024 (UTC)
So if someone searches for "Prebunking", the short description that they will see is "a nearly opposite meaning". So in fact this template is doing us all a favour my making the error obvious. Don't shoot the messenger, correct the message. Redirects should have SDs too.
But if someone has time, maybe it is possible to emulate the code that detects and flags 'fallback' SDs taken from Wikidata? --𝕁𝕄𝔽 (talk) 12:36, 29 November 2024 (UTC)
Sorry – I'm not sure where else any reader would be shown the short description of a redirect target without simultaneously being shown the title of that target? As far as I'm aware, the search tools that display short descriptions don't place them next to the title of the redirect being searched, but rather its target. (If this template were to replace the link provided with a link to the target in that same way, I would see no problem there either, but I wouldn't have assumed anyone would want that.) Certainly if this is an issue that's broader in scope than this one template I agree that this template isn't the problem, but I wasn't under the impression that was the case. Tollens (talk) 17:06, 29 November 2024 (UTC)
[ends]
So it looks like my surmise that {{annotated link}} is doing us a favour by highlighting "naked" redirects (especially R to section and R to anchor, which is the only serious problem) is not correct, unless I have failed to find a straw to clutch from your kind responses.
[I will resume the discussion at Template talk:Annotated link#Very poor annotations when annotating redirects rather than here but, fwiw, I will argue that the case for the prosecution is overstated, that ANLIs are extremely useful to readers when presented with a See Also list of meaningless (to them) buzzwords with no clue as to what it is or why they should be interested, thus advancing a major objective of the project, information discovery (library, not legal, sense). IMO, the rare cases sub-optimal behaviour like these two is an acceptable cost, given the reward. Feel free, of course, to participate in that discussion.]
I get big red text after the short description stating (for example:* Prebunking – False or misleading information presented as real Pages displaying short descriptions of redirect targets). I cant remember where I got this CSS, but it sure stands out and lets me know where the problems are. I get a similar message in big magenta text when a Wikidata description is used as fallback. This shows up on pages where Template:Annotated link is used, and where short descriptions are shown as annotations on category pages. The screenshots at Wikipedia:Short description illustrate this quite well. I find it helpful for fixing redirects which need a short description but do not have one yet. As the cases above illustrate, using the short description of the target page can be quite misleading, so I fix them as I find them. I do not know if they are displayed as default for me because of some script, or if this is the general case. A lot of changes have been made to Template:Annotated link over the last few years, and I don't really know how many of them work. · · · Peter Southwood(talk): 14:01, 5 December 2024 (UTC)[reply]
Not broken, it's just that we managed to finish all the articles (don't think that has ever happened since the number was upped to 3000, so well done all) and the bot hasn't added the November articles yet. Probably will be up in a few hours. Liu1126 (talk) 01:34, 1 December 2024 (UTC)[reply]
Sometimes, popular articles without SDs are that way because creating an SD for them is difficult. I have added some links to the "What can I work on?" list that are intended to help newer editors find easy SD work. – Jonesey95 (talk) 00:34, 5 December 2024 (UTC)[reply]
Change coming related to short description of "none"
I haven't tested this upcoming change, but tools affected by the short description of "none" may need to look at T326898. There are some recommendations from WMF developers. – Jonesey95 (talk) 22:47, 5 December 2024 (UTC)[reply]
fd the need for short descriptions in articles that already have self-explanatory explanation
I would like to open this discussion by addressing a critical inconsistency in the application of short descriptions across Wikipedia. While the current guideline suggests that articles with perfectly clear titles may not require descriptions, I propose that adding concise, relevant short descriptions to all articles is a more beneficial and consistent approach.
Short descriptions serve as a vital tool for enhancing navigation, accessibility, and user experience. They provide immediate context in search results, mobile apps, and external tools that rely on Wikipedia’s API, helping readers understand the scope of an article at a glance. Furthermore, they ensure uniformity across the platform, preventing discrepancies that can confuse or mislead users.
For example, while some articles like History of Painting have retained short descriptions, others like History of Music face contention despite the value they offer. This inconsistent application creates unnecessary debate and diminishes Wikipedia's ability to serve as a cohesive source of knowledge.
By standardizing the inclusion of short descriptions, we can avoid redundancy while still prioritizing clarity. The descriptions do not need to restate the title but instead enhance it, providing critical context. For example:
- History of Music: "The development of music over time" (33 characters).
- History of Painting: "The development of painting over time" (38 characters).
This approach remains succinct, adheres to the guideline’s emphasis on brevity, and ensures that Wikipedia maintains its commitment to accessibility and clarity across all articles.
I propose revisiting the guidelines to allow descriptions for all articles, with an emphasis on crafting short, meaningful summaries that align with the article's scope. By doing so, we uphold Wikipedia's standards of consistency, accessibility, and user-friendliness, and eliminate unnecessary conflicts over what does or does not merit a description.
It's unlikely that you'll get agreement to remove 'none' entirely. There is long-standing consensus for it, specifically to avoid SDs that simply replicate the title in different words, requiring editors to create and users to read pointless text that adds no additional information. It may help that the guidance on 'none' has recently been updated in an effort to improve consistency: Wikipedia:Short_description#"none"_as_a_short_description. MichaelMaggs (talk) MichaelMaggs (talk) 09:44, 6 January 2025 (UTC)[reply]
Thank you for your thoughtful response. I understand the concerns about short descriptions that replicate the title and agree that they should add meaningful context rather than restate what is already obvious. However, I believe this issue highlights the importance of refining the descriptions rather than defaulting to "none" in cases where they might seem redundant.
Short descriptions can and should serve a purpose beyond repeating the title. For instance, they provide immediate value in search results, mobile interfaces, and external tools that display Wikipedia content. The distinction lies in crafting descriptions that enhance understanding rather than merely duplicate the title. Consider the difference between the following examples:
- *History of Music*: "The development of music over time."
- *History of Astronomy*: "The study of the universe's history."
Neither simply repeats the title but instead provides a concise indication of scope, helping users quickly understand what to expect.
Furthermore, while setting descriptions to "none" avoids redundancy, it risks inconsistency across Wikipedia and diminishes the usability of the platform. By focusing on thoughtful and concise descriptions, we can maintain high standards of accessibility while addressing the valid concerns raised about unnecessary repetition.
I suggest that instead of removing descriptions entirely, we explore clearer guidelines for creating effective, non-redundant descriptions. This approach would resolve concerns about redundancy while preserving the utility and consistency of short descriptions across all articles. I look forward to further discussing how we can collaboratively achieve this balance. Ai777 (talk) 18:34, 6 January 2025 (UTC)[reply]
Neither of those short descriptions provides a better indication of scope, and the second one is simply wrong. Do not expect llms to understand the articles. CMD (talk) 04:30, 7 January 2025 (UTC)[reply]
Ai777, you appear to be mis-using AI both to generate the above arguments and also to generate poor quality and often completely wrong short descriptions. This discussion is best continued on your talk page. MichaelMaggs (talk) 17:39, 7 January 2025 (UTC)[reply]
Is none case-sensitive?
I ask because I've just used the Android app suggested edits feature to add a short description of none to an article. The app displayed a warning that the description should be capitalised, but I went ahead and used lowercase anyway. On checking the edit, I discovered that what I'd actually done was change an already existing description from None to none, not add one. So the article had been identified as not having a description, despite already having uppercase None. This suggests that the case might be important, though I don't know whether the same would have happened with lowercase. Musiconeologist (talk) 18:58, 6 January 2025 (UTC)[reply]
"None" is not case sensitive; you can test this in the sandbox, and shortdesc helper will show the standard message for pages with none as a short description. The Android app just asks people to add short descriptions to articles even if the article has "none" as the short description. See T326898 for the bug ticket. Liu1126 (talk) 20:05, 6 January 2025 (UTC)[reply]
Some SDs auto-built by templates do not start with uppercase (capital) letter.
The two examples you gave appear to be properly capitalized though. Shortdesc helper is saying that Russia–United States relations has a sdesc of "Bilateral relations" and Somatic symptom disorder has a sdesc of "Medical condition", both of which start with a capital letter, which checks out with the infoboxes' source codes. Where are you seeing them as uncapitalized? Liu1126 (talk) 02:54, 10 January 2025 (UTC)[reply]
Oh, are you talking about the description that is retrieved by Template:Annotated link? The description provided by that is lowercase for your two examples ("bilateral relationship between Russia and the United States" and "category of mental disorder", respectively). I believe that's because the template gets the Wikidata description if a short description template isn't present in the wikitext (from documentation: If a template is not present in the wikitext of the page, the description will be derived from Wikidata.), which is the case for these pages as the autogenerated short description comes from the transcluded infobox template and hence do not show up in the wikitext, so the annotated link grabs the Wikidata description (which generally isn't capitalized per Wikidata's own policies). Liu1126 (talk) 03:07, 10 January 2025 (UTC)[reply]
Proposed bot to mass-add short descriptions to pages in categories
As in your previous proposal, this suggests applying the same SD to all articles within each selected category. Can you give examples of categories where that would be appropriate? I suspect there will be few. My experience of running ShortDescBot on articles in categories relating to organisms was that the category tree is very difficult to deal with due to the large number of exceptions, errors and oddities. ShortDescBot handled those by generating a unique SD for each article (and incorporating a manual pre-check stage for every edit). Even for very simple categories, your proposal for the same SD for all articles within the category is likely to have a very high error rate. I assume you'll never be using "none", as that would not be consistent with the WP:SDNONE requirement for article-by-article consideration. MichaelMaggs (talk) 22:47, 26 January 2025 (UTC)[reply]
In Category:SES satellites, I count nine different existing short descriptions. In Category:Defunct Texas railroads, I count 37 existing short descriptions, including those on Wikidata. Category:Churches in Telemark has just two articles without SDs (and three different SDs). Category:Satellite buses has 20+ different SDs and just 58 articles. If these are examples of good categories to work on, I don't want to see the difficult ones. I think AWB or a similar editing tool, or just the shortdesc helper gadget, will be needed here, and the bot as described will not be helpful. – Jonesey95 (talk) 23:16, 26 January 2025 (UTC)[reply]
Replying in similar vein, really. Since I use the Android app I see the short descriptions of most articles I visit, and regularly come across generic ones that seem to me to be worse than no description at all, or which desperately need changing but appear on many different articles: Protein-coding gene in the species Homo sapiens, for example, which appeared even for a gene which according to the article we share with mushrooms. Most of those should surely just say Gene or Human gene or Protein, or say what the gene or protein does. I do use the suggested edits feature quite a bit to add descriptions, and my experience is that usually the only safe way to write an accurate one is to read some or all of the lead section.There might be situations where it can work, but for myself I'd rather add a description from scratch than try to work out whether a generic one is true or not. I'd rather see no information than wrong information. Musiconeologist (talk) 23:22, 26 January 2025 (UTC)[reply]
Admittedly, the bot would not be suitable for many categories such as the articles you mentioned. However, the reason I'm bringing up this proposal again is that I went through Category:Spacecraft launched by Soyuz-U rockets, articles which all largely have the same short description, and manually added the same short description to each. The articles in this category cover very similar spacecraft. Lastly, to make sure the bot doesn't add false short descriptions, the operator will need to verify every proposed edit by the bot before it is published. UrbanVersis32KB ⚡ (talk / contribs)23:33, 26 January 2025 (UTC)[reply]
It's the potential for careless use that worries me, I'd say—I can see its value in the situation you've mentioned. I almost feel that what really needs approving is an editor's permission to use it. I'm not dead against its existence, but nervous of it being used incompetently. Musiconeologist (talk) 23:55, 26 January 2025 (UTC)[reply]
To be honest, I haven't really thought of AWB-esque methods for approval. I only know about the regular bot approval list, but it seems like the bot may better be suited for a different method. UrbanVersis32KB ⚡ (talk / contribs)02:23, 27 January 2025 (UTC)[reply]
Frankly, I don't think this is a good task for you. I just looked at Category:Spacecraft launched by Soyuz-U rockets, hoping to verify that the articles there largely have the same short description, and here's a list of the different short descriptions I found there, along with the frequency of each SD:
1 1975 Soviet uncrewed spaceflight to Salyut 4
4 1978 Soviet crewed spaceflight to Salyut 6
2 1979 Soviet crewed spaceflight to Salyut 6
1 1979 Soviet test spaceflight to Salyut 6
1 1979 Soviet uncrewed spaceflight to Salyut 6
1 1980 crewed flight of the Soyuz programme
3 1980 Soviet crewed spaceflight to Salyut 6
1 1980 Soviet human spaceflight mission to the Salyut 6 space station
2 2001 Russian crewed spaceflight to the ISS
1 2002 Russian crewed spaceflight to the ISS
1 2016 Russian resupply spaceflight to the ISS
1 2017 Russian resupply spaceflight to the ISS
1 Aborted 1983 Soviet crewed spaceflight
1 Cargo spacecraft
6 Crewed flight of the Soyuz programme
1 Failed 2016 Russian resupply spaceflight to the ISS
1 First crewed spaceflight to the ISS
1 First international crewed spaceflight mission
1 International Space Station resupply mission
1 International Space Station spacecraft
1 Modified Progress spacecraft used to deliver the Pirs module to the ISS
1 Progress spacecraft used to resupply the International Space Station
1 Resupply mission or crew escape test
1 Russian aircraft
46 Russian cargo spacecraft
1 Russian Earth observation satellite
1 Russian expendable cargo spacecraft
1 Russian Kobalt-M reconnaissance satellite
1 Russian optical reconnaissance satellite
15 Russian spacecraft
1 Russian Spacecraft
1 Russian spacecraft, in service in 2004
1 Russian spy satellite
1 Russian uncrewed cargo spacecraft
1 Russian uncrewed cargo spacecraft of 1997
1 Russian unmanned cargo spacecraft
1 Russian unmanned spacecraft
3 Soviet cargo spacecraft
1 Soviet spacecraft
17 Soviet uncrewed Progress cargo spacecraft
21 Soviet unmanned Progress cargo spacecraft
1 Space mission of the Sojuz program
3 Spacecraft
1 Spacecraft for resupplying the International Space Station
1 Spacecraft that resupplied the International Space Station in 2013
1 Spaceflight
1 Student-built tether satellite
1 Supply vehicle for the International Space Station
Every time I check, the proposed bot operator's statements do not appear to match reality. I have zero confidence in this proposer's ability to carry out this bot task. I'm not trying to be unkind, but we have experience with well-meaning bot operators whose tasks are not analyzed properly and then cause significant damage. – Jonesey95 (talk) 06:26, 27 January 2025 (UTC)[reply]
While these descriptions are not identical, many are very similar. I don't think (if approved) my bot or myself would cause damage to Wikipedia in this manner, either. However, I acknowledge that I am relatively new to the WP bot process, and that this is probably not a good bot task, even with the changes I made to it since my proposal last year. I appreciate the feedback from you and other editors! UrbanVersis32KB ⚡ (talk / contribs)21:18, 27 January 2025 (UTC)[reply]
Others have addressed most of my concerns. It is also worth noting that a high number of the "autogenerated" or suggested short descriptions are not usable as presented, leading me to believe that this is still a project for humans. Rublamb (talk) 00:24, 27 January 2025 (UTC)[reply]
It's not an issue with this single editor but with the tool, e.g. [6][7]. The tool also seems to suggest many short descriptions which are a lot longer than our standard. Fram (talk) 08:49, 29 January 2025 (UTC)[reply]
I note from that ticket that it suddenly got a lot of input from enwiki users since mid-January 2025 after having been dormant for more than 2 years: coupled with the fact that I notice this issue just now, I get the impression that it has very recently been rolled out much wider or aggressively, and that the increased use has lead to increased complaints. I see on the page I listed above that "We will be conducting outreach in September 2024 through November 2024 to inform the Android product team of what further improvements can be made and if communities would like to adopt the feature. " So I guess that this has only been reenabled recently, but without sufficient testing or feedback. @JTanner (WMF): can you indicate if this has recently been activated for enwiki, and where consensus for this was found? Fram (talk) 10:05, 29 January 2025 (UTC)[reply]
Sigh, left side added, right side corrected, both are wrong[8]. This one has been reverted, of course. I thought the developers claimed that it would only be shown to people with 50+ edits, and not be used on BLPs, but both promises seem to have been forgotten. It is used for bad edits like this, and for extremely long "short" descriptions[9][10]. Of the last 500 edits of this kind, 43 have already been reverted (with the revert tag), others have been corrected, and others still need reverting... Fram (talk) 10:34, 29 January 2025 (UTC)[reply]
Could be, the used tag " App description add" links to here which talks about the MAAD Experiment, and this issue seems to have resurfaced only recently, so that's why I thought they were related. In any case, whatever tool is responsible for the suggestions should for starters probably skip any pages which already have the SD=none short description. Fram (talk) 15:44, 29 January 2025 (UTC)[reply]
Just wanted to drop a quick note to confirm that Machine-assisted article descriptions is not enabled anywhere on any wiki, and the edits referenced in the discussion were hand written and not made as a result from from any machine-assistance in the Android app. The team hasn't made any decisions so far and I'm not anticipating that to change until later in the year.
Regarding the lenght of the descriptions, I became aware of the note on on the phab task and the team were last week exploring options to reduce the length of article descriptions on the English Wikipedia. Either through a hard limit or through some kind of explicit guidance.
Thank you. Then I guess that the suggested edits tool needs some changes, as I can't imagine someone inventing "Wikimedia list" as a description on their own. But at least the MAAD isn't enabled, thank you. Fram (talk) 16:09, 29 January 2025 (UTC)[reply]
Seddon (WMF) and HNordeen (WMF), thanks for engaging. Please be aware that T326898, editors being guided to change valid "none" short descriptions, is still happening and should be fixed ASAP. Also, previously, WMF staff agreed that new editors would not be guided toward adding short descriptions, but that limitation has apparently been removed and needs to be reinstated; see T297341 for diffs. – Jonesey95 (talk) 17:12, 29 January 2025 (UTC)[reply]
This template must be followed by a complete and specific description of the request, so that an editor unfamiliar with the subject matter could complete the requested edit immediately.
Edit requests to fully protected pages should only be used for edits that are either uncontroversial or supported by consensus. If the proposed edit might be controversial, discuss it on the protected page's talk page before using this template. Consider making changes first to the template's sandbox and test them thoroughly here before submitting an edit request. To request that a page be protected or unprotected, make a protection request. When the request has been completed or denied, please add the |answered=yes parameter to deactivate the template.
On the first line, in the "then" parameter to the #ifeq, the argument:
<nowiki/><!--Prevents whitespace issues when used with adjacent newlines-->
should be changed to:
{{SHORTDESC:}}<nowiki/><!--Prevents whitespace issues when used with adjacent newlines-->
That is, {{SHORTDESC:}} should be added to the 'none' clause. The wikidata and app developers have agreed (phab:T326898) to respect "empty string" as a signal meaning "no short description is wanted for this article", ie the same thing {{Short description|none}} indicates, in a language-neutral way. Further discussion can be had on phab:T326898#8654683, but the apps will be updated (if they have not been already) to interpret an empty string as the short description to suppress prompts to add "missing" short descriptions to pages. C. Scott Ananian (WMF) (talk) 16:39, 29 January 2025 (UTC)[reply]
I tried that change in January 2023, and I'm pretty sure it didn't work in my testing. Here's the discussion about that. I have reinstated the above code in the sandbox. It should be tested in all relevant cases, including normal usage, usage in conjunction with infoboxes that use 2=noreplace, showing short descriptions on category pages, and display of short descriptions using the shortdesc helper gadget. There may be other use cases that need to be tested. – Jonesey95 (talk) 17:25, 29 January 2025 (UTC)[reply]
Yeah, it was a good try but unfortunately it wouldn't have worked before because it needed a change at the parsing level first to accept it as a valid input. Seddon (WMF) (talk) 13:08, 30 January 2025 (UTC)[reply]
Test results
I installed the sandbox version for testing at Alliance (New Zealand political party), which uses "none" as an SD and has an infobox that generates a short description if a local SD is not detected. Here are my results:
Test results
Test condition
Live template
Sandbox template
SD shown in Page information
"Local description" table row not displayed
"Local description" displayed with empty cell for the description
SD shown in article by shortdesc gadget
"This page intentionally has no description." (correct behavior)
"Missing article description" with "Add" link; (incorrect behavior)
So far, it looks like the sandbox is not working correctly. The shortdesc helper should show "This page intentionally has no description." I do not know whether the Page information should show no line or an empty line in the table, but the behavior is different. The category script is broken in both situations, which is a known issue. I haven't tested the mobile v. desktop situation yet. – Jonesey95 (talk) 17:58, 29 January 2025 (UTC)[reply]
And it looks like, from the response above, that the mobile v. desktop issue may be fixed now that the parsing has been modified. I suppose we should go ahead with this change. If it breaks anything significant, we can either fix that thing or revert. Do the WMF developers have any opinion or comment on the first test condition, "SD shown in Page information"? IMO showing a blank cell seems better, but I don't know if anything is dependent on the nonexistence of the table row rather than an empty cell. – Jonesey95 (talk) 16:47, 30 January 2025 (UTC)[reply]
Thanks for testing these changes; I would advocate for making the change, and then seeing if it breaks anything significant (and fixing/reverting if it does). This would also help us confirm whether we need to fix anything else on the client side of Apps, and/or the parsing side of our services. (I do also agree that re. Page Information it makes more sense to show a blank cell rather than nothing at all; This way, any tool that relies on Page Information will now be able to differentiate between a nonexistent description vs a deliberately-empty description.) DBrant (WMF) (talk) 19:06, 30 January 2025 (UTC)[reply]
