Alternate take to #27683 that builds the modifier class based on min-width, such that list groups are stacked until the class's breakpoint is reached. This is less verbose, matches our primary responsive approach, and prevents an extra class. Unfortunately, I believe supporting flush list groups is too much code here, so I've skipped that and made a note in the docs.
Also added examples generate from our breakpoints data file with a protip for how to do equal width items.
- Move 4.1 docs to 4.2
- Update versions everywhere to 4.1.3 with release script
- Manually bump the shorthand version in package.json
- Add 4.2 to the versions docs page
- Update some redirects
- Fix tests asset URLs
- Bump Nuget and more
* SVG accessibility fixes
- add `focusable="false"` to all SVGs, to stop IE from including the `<svg>` in its default focus cycle
- completely hide decorative SVGs using `aria-hidden="true"`
- add `role="img"` to meaningful/"content" SVGs, ensure they have an appropriate `<title>` and/or `aria-label`
- simplify placeholder default title
* Simplify usage of placeholder.svg in the docs
- in most cases, no need for separate/redundant text and title (now that we make text itself part of the alternative text explicitly)
- no need (expect in rare cirumstances) to describe the placeholder image (unless the look of the image is important/the actual content, as is the case with the image thumbnail class)
* Simplify example.html rewriting of <svg> to <img> code sample
- as the `alt` for the placeholder isn't really important, just add an ellipsis instead (otherwise, to be correct, we'd have to construct the whole "title+text" construct like we do now in the
placeholder.svg itself
* Change svgo configuration to retain role attribute and add/force focusable=false
- regarding `focusable=false`, see https://github.com/svg/svgo/issues/1011 (which would be the "proper" SVGO fix) and the (hacky) solution https://github.com/svg/svgo/pull/817
