Wikipedia:WikiProject Accessibility/Top priority issues with templates

From Wikipedia, the free encyclopedia

Here is a small list of long awaited key improvements to make, often important for several aspects of web quality.

As major changes in templates are often long and tedious to make (and also to gather consensus on the said changes), we should aim to maximize the value of these changes, and aim for quality in a broad way. Improving care of accessibility, usability, semantics, and front-end web speed altogether when possible. Back-end template performance should also be taken into account when relevant.[note 1]

Most widely used templates[edit]

We should make a list of the 200 most widely used templates and review them carefully, monitor their changes, etc. See Wikipedia:WikiProject Accessibility/Most widely used templates.

Table templates[edit]

Please help expand this list. To be reviewed

Easy to fix templates[edit]

Infobox[edit]

Without changing the layout nor making big changes:

  •  Done Template:Infobox/row - included in more than 782589 articles. With this most of the job is done, Template:Infobox is awesome and must be used
    • "font-size:88%": I will definitely try my hardest to have the default text size in Template:Infobox set back to default. Small font size are hard to read, and thus reduce usability. This was proved by scientific reading studies, small text size can lengthen the reading rate by 1.5 (for example, a text that you could read in 20 seconds in normal size (12 points) could take you 30 seconds in smaller size (10 points). See Wikipedia:WikiProject Usability/Readability guidelines
  •  Done Template:Infobox artist discography - but very lacking from a design point of view
Remaining Infobox that needs to be converted to Template:Infobox[edit]

Templates-based tables[edit]

Citation templates[edit]

  • Template:Quote box: should use the blockquote tag. DIVs within the textual content should be replaced by SPANs (the enclosing paragraph P can then be generated correctly by MediaWiki).

Harder to fix[edit]

Template-based tables[edit]

Others[edit]

Timelines[edit]

Replacement by tables[edit]

Notes:

Headers templates[edit]

Templates that can easily be misused[edit]

Usage of these templates should be checked on a regular basis.

Cross-wiki ban of HiddenStructure[edit]

See Wikipedia:HiddenStructure. The hiddenstructure class was removed from skins in bugzilla:17009. It may remain in several wikis' local common.css.

Low priority improvements[edit]

Improvements to collapsible menus[edit]

Another script should be made from scratch. Wikipedia:WikiProject Accessibility/Navigation menu will serve as the experimental field for this job. The code is almost ready at User:Dodoïste/navmenu.css and User:Dodoïste/navmenu.js.

Accessibility-wise, this new script should allow for menus made of lists, the sub-menu headers being part of the list (which is currently impossible). And the link or button used to show/hide the menu should be able to be inserted inside a li element, along with its corresponding header. Alternatively, the same thing can be done with the p element. Another alternative is to provide a meaningful title to the link show/hide, in order to made it explicit in context. And neither MediaWiki nor the collapsible script allow to fix this issue, in the current implementation of the script. Related WCAG 2 guidelines:

  1. H30: Providing link text that describes the purpose of a link for anchor elements and G91: Providing link text that describes the purpose of a link
  2. H25: Providing a title using the title element
  3. H33: Supplementing link text with the title attribute
  4. F63: Failure of Success Criterion 2.4.4 due to providing link context only in content that is not related to the link

Usability-wise it should be done with JQuery to make pretty animations, and CSS for button affordance. It implies that the JQuery library should be deployed to all skins, this is planned by the Usability Initiative team.

Notes[edit]

  1. ^ Extremely large templates - especially when they have multiple levels of inclusions (transclusion depth) - lengthen strongly the time necessary for the servers to prepare an uncached page, up to 10 - 20 seconds, or more. Since the current revision of articles are cached, most of the time the templates doesn't have an influence at all. But when users have specific account configurations, browse archived pages or request uncached pages, large templates are an important burden on the servers. We shouldn't focus too much on performance, but we shouldn't play dumb either: many templates have obvious and detrimental performance issues.