User:Isaacl/Visual diff test

VisualDiff is a tool that displays diffs in a more WYSIWYG manner, rather than just a Unix diff-like comparison of the source text. It is particularly good for browsing through a large number of diffs, providing context around each.

Because of the greater use of horizontal screen space, VisualDiff can use vertical space more efficiently.

There are situations where it hangs or produces difficult-to-read output:


 * Mis-nested markup: when an HTML inline element incorrectly spans across multiple block elements, VisualDiff hangs and doesn't produce any output.
 * Numbered lists: VisualDiff seems to include the entire list, presumably to get the numbering correct. Unfortunately for long lists, this is cumbersome.
 * Unnecessary changes in list styles: unfortunately, it is common for editors to unnecessary change a leading prefix of asterisks and colons to some other combination, which results in lists being closed and opened. VisualDiff treats this as changed text, and so a lot of content which actually hasn't been changed will get flagged as altered.

Miscellaneous issues (using description list for testing) :
 * Although the text and background colours used by VisualDiff meet the WCAG guidelines for colour contrast ratio, I find them difficult to read and would refer a lighter background colour.
 * Colour contrast ratio isn't a one-stop solution; some colour combinations can still have poor legibility next to each other. For example, some colours have a vibrating appearance when adjacent.


 * Some project discussion pages or editor talk pages have an opening  tag with styling to produce a border around all comments. This relies on the browser implicitly closing the   element. VisualDiff isn't able to isolate changes on such pages; it will mark the entire   as being changed.

test adding text within tq template


 * test adding second instance of text within tq template within an unbulleted list

More testsvis
Here is an introductory paragraph
 * followed by a first-level bullet item
 * Inserting a third-level non-bulleted item, and now editing it


 * then a second-level bullet item