Enquête 2015 sur les souhaits de la communauté/Lecture

This page is a translated version of the page Community Wishlist Survey 2015/Reading and the translation is 81% complete.

Rendre la qualité/fiabilité d'un article plus claires pour les lecteurs

Actuellement, hormis les modèles d'ébauche et modèles ajoutés par les utilisateurs, il n'est pas possible de montrer clairement au lecteur de passage la qualité/fiabilité d'un article. C'est un problème, vu la pénétration croissante de contenu promotionnel et autre (voir en:User:Doc_James/Paid_editing). Une solution possible est de 1) activer un gadget pour tous les lecteurs par défaut, 2) développer un script complémentaire, qui afficherait (peut-être à la demande, en étant accessible depuis une barre d'outils ou l'en-tête) des informations sur le nombre de contributeurs suivant la page, les principaux contributeurs et peut-être l'une des nombreuses mesures de confiance présentées dans des articles de recherche et 3) ajouter une liste de vérification pour les problèmes courants (voir ma proposition ci-dessous). --Piotrus (talk) 05:21, 10 November 2015 (UTC)[]


  1.   Support--Shizhao (talk) 09:30, 1 December 2015 (UTC)[]
  2.   Support This bit of literacy is incredibly important for readers - understanding that we have a process for creating content and quality Sadads (talk) 16:07, 1 December 2015 (UTC)[]
  3.   Support Especially the number of watchers would be interesting and a decent simple measure of page quality. Is that number currently publicly available anywhere? Gap9551 (talk) 23:03, 1 December 2015 (UTC)[]
  4.   Support #1 only as it gives an overview of the article's quality pretty succinctly. I don't think readers generally would care about the number of watchers or the other info. Stevie is the man! TalkWork 00:47, 2 December 2015 (UTC)[]
  5.   Support--Manlleus (talk) 15:40, 2 December 2015 (UTC)[]
  6.   Support -- SantiLak (talk) 10:47, 4 December 2015 (UTC)[]
  7.   Support Halibutt (talk) 00:35, 5 December 2015 (UTC)[]
  8.   Support --Yeza (talk) 16:56, 5 December 2015 (UTC)[]
  9.   Support Zamaster4536 (talk) 12:54, 6 December 2015 (UTC)[]
  10.   Comment An interesting approach that could be investigated is providing automated metrics based on the age of the page, number of edits, number of contributors, shape of the distribution of number of edits by contributor and average size of edits by contributor, (c.f. xContribs), average age of words in the page, etc. This would be similar to the "factoids"/"in a nutshell" feature of OpenHub/Ohloh, e.g. https://www.openhub.net/p/mediawiki#factoids --Waldir (talk) 15:47, 8 December 2015 (UTC)[]
  11.   Support Abyssal (talk) 16:49, 10 December 2015 (UTC)[]
  12.   Support --Tgr (talk) 22:20, 13 December 2015 (UTC)[]
  13. Comment: (support) -- The user needs a simple (automated) facility to help them understand a significant number of issues (some listed above per Waldir) related to quality. When people ask me if they should trust Wikipedia, I try to remind them of how to assess trust in other areas of their life. This facility should seek to be an intelligent guide to trust; not a certification of correctness.
    This might include tool-tip like cautions of the more significant issues for the article. Such a system should recognize the age and number of editor templates already on the article. When this system is implemented, the current editor templates should be disabled for IP readers but still appear for IP editing and for logged-in users.
    Both the age and distribution distribution of references in the text affects current topics. However, the scoring must recognize that historical topics more than ~50 years old shouldn't discount quality because of non-current references.
    SBaker43 (talk) 20:02, 14 December 2015 (UTC)[]

Présentation plus adaptée aux appareils mobiles pour les pages de projet et de portail sur plusieurs colonnes

Sur de nombreux portails et projets, les pages principales utilisent des boîtes disposées sur plusieurs colonnes. Cela permet une distribution propre et compacte de contenus liés ou non sur les grands écrans d'ordinateurs de bureau. Mais malheureusement, cette présentation est le plus souvent réalisée à l'aide de tableaux, ce qui est intrinsèquement incompatible avec les appareils plus petits. C'est mon expérience de de-wiki, mais je ne crois pas que ce soit mieux sur d'autres wikis (exemples arbitraires de portails s'affichant mal sur mobile : [1] [2] [3] [4] [5] [6]; pour avoir une idée du résultat sur mobile, ouvrez ces liens dans votre navigateur et réduisez la largeur de la fenêtre aux dimensions d'un smartphone).

Il y a beaucoup de raisons expiquant ces mauvais résultats. Outre le fait que de nombreux portails ont été mis en place avant l'ère du pré-mobile, nous souffrons encore de l'indisponibilité d'un cadres de solutions techniques facilement complets pour construire des portails/pages adaptés au mobile d'emblée, sans la nécessité de manipulations CSS complexes. Je propose donc d'élaborer un cadre, constitué principalement de classes CSS sur un site approprié dans la famille Mediawiki et donc disponible pour l'usage collectif, avec les propriétés suivantes:

  • a flexible set of CSS classes, separated for style and positioning of content boxes; users can add classes to portal elements and define style and positioning by doing so
  • boxes can be encapsulated by the usage of templates with parameters, which I think are better known to wiki users than plain HTML
  • once compiled to a portal/project page, the box arrangement intrinsically adapts to display size, regardless of whether standard frontend, mobile frontend, or a Wikimedia app is used
  • a detailed documentation including use cases will complement this frameset; an average wiki user shall be able to create a mobile- & desktop-friendly portal page
  • bonus task: enable modern CSS techniques in portal namespace, like for instance media queries and whatever is useful; technically skilled users can enhance mobile and desktop experience by using this feature

MisterSynergy (talk) 20:01, 11 November 2015 (UTC)[]


  1.   Support MisterSynergy (talk) 09:08, 30 November 2015 (UTC)[]
  2.   Support --g (talk) 15:13, 1 December 2015 (UTC)[]
  3. Not supportive, mostly because portals should more likely be deprecated as out-of-date and generally of poor quality. We are not hearing from many projects that use them, and they have been useless to readers for many years on (at least) several of the large Wikipedias. A bit of history: on Enwiki, portal creation became popular for a period because it was easy to develop a 'featured portal', and evidence of having contributed "featured content" was a relatively quick way to becoming an administrator. Those portals were then promptly ignored by their creators, and most have not been revised in years. While the motivation may have been different on other projects, it seems they are essentially historical artifacts today. Risker (talk) 23:02, 1 December 2015 (UTC)[]
  4.   Comment I agree with Risker that portals should probably be decommissioned as they are hulks nobody has the time to work on, and I don't think they get much reader attention anyway. The true portals are the main subject articles themselves, because those are what people find when they search. WikiProjects, however, are the "portals" through which some editors go to find out how to help with particular subjects. I wouldn't oppose efforts to make them mobile-friendly, but I think the way to go is to provide a way of having alternative display code for mobile devices, rather than changing the designs of existing WikiProject pages. At any rate, many WikiProjects are short-staffed and will consider mobile access a low priority in my estimation. Stevie is the man! TalkWork 12:32, 2 December 2015 (UTC)[]
    I disagree with deprecating portals. They are a great way to show interesting things to readers. I have developed several sports portals in Spanish Wikipedia, with heavy use of random sections. Hey, nitable competitors appear on their birthdays!
    The problem with portals, other than lack of editors, is that we need to writhe the summaries of articlesvand update them manually. A nice tool would be insetion of page sections, therefore we could just insert the zero section. --NaBUru38 (talk) 04:05, 13 December 2015 (UTC)[]
    I would support some portal-like features built into main subject articles. The current separated portals are hulks that very few visit (I've looked at pageview stats) and almost nobody has the time to maintain. If we have to keep portals around, they need to be "advertised" better from main subject articles. Stevie is the man! TalkWork 17:29, 13 December 2015 (UTC)[]
  5.   Support--Manlleus (talk) 15:41, 2 December 2015 (UTC)[]
  6.   SupportBeleg Tâl (talk) 17:02, 2 December 2015 (UTC)[]
  7.   Oppose - Do many readers/editors using mobile phones want this? I've just looked at some of the above examples using a small laptop (large tablet) size screen and had no problems; the material may not have been laid out quite as neatly as on a larger screen, but you can't have everything. If this went ahead then either (1) existing projects/portals would need to be converted to the new system or (2) we would have a mix of projects/portals using the old and new systems - both would increase complexity for editors. Many projects have some tabbed pages that are pointless (or worse). DexDor (talk) 20:02, 9 December 2015 (UTC)[]
  8.   Support A nice first step would be optional linebreaks. If a row has two columns with an optional linebreak, the two boxes would appers side by side if the screen is wide, and top to bootom if the screen is narrow. --NaBUru38 (talk)

Liste de lecture

Tracked in Phabricator:
task T120756

Qq uuelques lecteurs (mais normalement ne les éditeurs) trouvent un article intéressant et il leur plairait le lire après, peut-être parce qu'ils n'ont pas suffisant temps dans ce moment. Il serait très bon si ils pussent l'agréger à sa liste de lecture de sorte qu'ils ne l'oublient pas! Se que il y a beaucoup de formes d'agréger marqueurs à des pages en internet (autant dans le navigateur comme en des fournisseurs de services en ligne) mais il est une idée qu'il vaut la peine envisager. 4nn1l2 (talk) 10:28, 10 November 2015 (UTC)[]


  1.   Support 4nn1l2 (talk) 02:59, 30 November 2015 (UTC)[]
  2.   Support בנימין (talk) 07:37, 30 November 2015 (UTC)[]
  3.   Support Ldorfman (talk) 22:23, 30 November 2015 (UTC)[]
  4.   Support Alleycat80 (talk) 09:04, 1 December 2015 (UTC)[]
  5.   Support --Shizhao (talk) 09:31, 1 December 2015 (UTC)[]
  6.   Support -- AvatarFR (talk) 13:30, 1 December 2015 (UTC)[]
  7.   Support simple: non-watching watchlists. --Purodha Blissenbach (talk) 14:32, 1 December 2015 (UTC)[]
  8.   Support --Martinligabue (talk) 15:04, 1 December 2015 (UTC)[]
  9.   Support -- reader focused watchlists? Or something that entices these readers to engage the already available system, but make that list more visible? Public reading lists (maybe the books extension reoriented towards IPS?) could be a good social media tool as well, Sadads (talk) 16:09, 1 December 2015 (UTC)[]
  10.   Support --Urbanecm (talk) 17:46, 1 December 2015 (UTC)[]
  11.   Support --Wesalius (talk) 19:10, 1 December 2015 (UTC)[]
  12.   Support Jules78120 (talk) 19:57, 1 December 2015 (UTC)[]
  13.   Strongly oppose --Usien6 (talk) 20:52, 1 December 2015 (UTC) // Can be designed by the community as a "gadget" or browser extension: no reason to burden WMF with it. By the way, Safari is already shipped with such feature with the bonus of off-line caching and cross-client syncing. Actually, every major browser (and most of the minors...) have the bookmarking feature, which does the job pretty well. Seriously, guys...[]
  14.   Oppose Redundant to the bookmark functionality in your browser and does not improve wiki content or editor productivity. And if you're on a public computer, improvise! A piece of paper, writing on your hand, sending an email to yourself, adding it to a wiki page/watchlist are all easy workarounds. MER-C (talk) 20:58, 1 December 2015 (UTC)[]
  15.   Oppose Not related to improving and maintaining wiki projects. Easily done by browser, as mentioned before. Gap9551 (talk) 23:05, 1 December 2015 (UTC)[]
  16.   Support Helder 23:40, 1 December 2015 (UTC)[]
  17.   Comment If someone wants to develop a gadget or extension for this, that would be all right, but I don't see this going into the core wiki code. Also, Wikipedians can create a reading list on one of their user pages, prettied up using the Todo template if they like. Stevie is the man! TalkWork 12:43, 2 December 2015 (UTC)[]
  18.   Support Why not? Regards, Kertraon (talk) 13:19, 2 December 2015 (UTC)[]
  19.   Support--Manlleus (talk) 15:41, 2 December 2015 (UTC)[]
  20.   Support--MisterSanderson (talk) 01:00, 3 December 2015 (UTC)[]
  21.   Support YBG (talk) 06:36, 3 December 2015 (UTC)[]
  22.   Support Rzuwig 10:57, 3 December 2015 (UTC)[]
  23.   Support Orbwiki107 (talk) 17:24, 3 December 2015 (UTC)[]
  24.   Support SantiLak (talk) 10:47, 4 December 2015 (UTC)[]
  25.   Support Chenspec (talk) 21:16, 4 December 2015 (UTC)[]
  26.   Support Lester לסטר (talk) 18:09, 5 December 2015 (UTC)[]
  27.   Support Zamaster4536 (talk) 12:54, 6 December 2015 (UTC)[]
  28.   Oppose per opposes above. Don't add unnecessary clutter to readers screens. DexDor (talk) 20:07, 9 December 2015 (UTC)[]
  29.   Support Alkamid (talk) 22:33, 13 December 2015 (UTC)[]

Sous-pages : supprimer les liens vers les pages de niveau supérieur

Wikibooks uses Subpages in order to structure a book in chapters. Wikiversity does the same for a course and lessons. Each subpage shows links to ancestor pages automatically. Nevertheless the authors offer to navigate between chapters and the content page using navigational templates. In these cases, you see the link to the ancestor page as well as the navigation bar -- e.g. b:en:Geometry for Elementary School/Introduction. I'ld prefer to suppress the automatically created link if a navigation bar is shown. Double information may disturb users.

Users: Wikibooks, Wikiversity and other projects that use subpages. Current situation: Links to ancestor pages can't be suppressed. Solution: Add a behavior switch __NOANCESTORLINK__ or something like that.

-- Juetho (talk) 10:45, 18 November 2015 (UTC)[]


  1.   Support - This would be useful occasionally on English Wikipedia, too - for example, there is no reason for Talk:P/poly (the talk page of the P/poly complexity class) to link to Talk:P (the article being about the letter). עוד מישהו Od Mishehu 21:21, 30 November 2015 (UTC)[]
  2.   Support, and only with the magic word or a comparable feature that works page-by-page. Having it added automatically by algorithm (e.g. "if the page displays X, don't display the ancestor page") would risk its omission by a false positive, but merely making it suppressible page-by-page wouldn't be a problem. If a project needs to remove ancestor links from a whole batch of pages, it can run a bot to add the magic word to the pages in question. Nyttend (talk) 21:49, 30 November 2015 (UTC)[]
  3.   Support using magic words, with a wiki-by-wiki default setting variable, for wikis like Wikiversity where these links tend to be unhelpful. --YodinT 02:43, 1 December 2015 (UTC)[]
  4.   Support --Shizhao (talk) 09:31, 1 December 2015 (UTC)[]
  5.   Oppose or make it a user preference option with appropriate wiki-wide default. I !@:#\ things being too different between wikis I am using. --Purodha Blissenbach (talk) 14:30, 1 December 2015 (UTC)[]
  6.   Oppose --Usien6 (talk) 20:55, 1 December 2015 (UTC)[]
  7.   Oppose Helder 23:40, 1 December 2015 (UTC)[]
  8.   Support per Nyttend. Stevie is the man! TalkWork 12:49, 2 December 2015 (UTC)[]
  9.   Oppose -- Overall wiki navigation should remain consistent throughout. Navigation templates can be redesigned or moved to resolve this issue. -- Dave Braunschweig (talk) 22:20, 2 December 2015 (UTC)[]
  10.   Support--MisterSanderson (talk) 01:00, 3 December 2015 (UTC)[]
  11.   Support Juetho (talk) 07:32, 8 December 2015 (UTC) -- I dont't want a wiki-wide way to suppress these small links. I only want a page-by-page option set by an author manually.[]