Link targets need to be focusable to work with VoiceOver.
Making <main> focusable causes some side-effects, like making the TAB
key go to the beginning of <main> instead of the element after the
currently-clicked region.
Also removes the annoying outline around "main" in some non-mainstream
browsers, without having to add extra CSS.
IBM Equal Access A11y Checker caught an <aside> without a label. Figured
this was a good opportunity to instead use the site description.
While I was at it, I expanded said site description and used it
properly.
The WCAG "label in name" SC requires visible labels to contain
accessible names, preferably by having accessible names start with
visible labels. This commit makes footnote backlinks display as a
hyperlink reading "Back" to meet this SC.
- We don't want autocomplete on the webmention form, since users should
not be entering a link multiple times.
- Not all browsers support the "required" HTML5 attribute, so add the
"aria-required" HTML attribute too.
- Adjust transcript shortcodes to also support <audio> elements.
- Add audio element shortcode based on <picture> shortcode
- Make <audio> elements match container width.
- Display reply content in webmentions, when it's available
- Truncate titles and redundant content from webmentions
- Add note on a11y issues regarding badly-formatted webmentions from
brid.gy's mastodon integration.
- Add a11y metadata to transcribed images to communicate the presence of
a transcript
- Fix relative urls in navigation: make them absolute urls, so that the
parsed navigation elements have the correct destinations.
This also switches image transcripts from a section with a heading to a
div with an ARIA label. That should reduce duplication between the
<summary> content and the heading while still being semantically sound.
- Switch an abs link to a relative one
- Account for a site move
- Manually correct a couple dead links to point to the Wayback Machine
- Automatically switch some webmention links to the Wayback Machine
The RSS feeds use escaped HTML instead of XHTML, which improves
compatibility with certain feed readers (e.g. Microsoft Outlook).
Mention that Outlook uses its own weird engine for feed contents in my
web best practices article.
I increased a bunch of tap targets to fit the recommended 48x48 sizes
and 56x56 non-overlapping regions, but unfortunately this caused
outlines to overlap with each other. This commit turns these elements
into inline-block elements and makes the necessary fixes to accomodate
this change.
- Drop copyright symbol: I put it there because certain programs
explicitly look for it, but between rel=license, schema.org microdata,
and creative commons RDFa, I think scrapers should be covered.
- Update the theme-color and friends to work with my site's updated dark
theme.
Explicitly mark RSS feeds as "legacy" because I have joined the A-Town
Atom gang and they have a strict full-time attendance policy.
Thanks Forever for pointing out the spelling/grammar error here.
- Re-order rel for my homepage's representative h-crad
- Fix the aria-current for the homepage
- Microdata: remove unnecessary "sameAs"
- Reduce "update without adding date updated" timeout to 15min
- Add u-uid and rel=me to my h-card on every page.
It's useless and invalid in those contexts, and removing it
significantly trims the file sizes.
Edit a post featuring a microdata code snippet to avoid conflicts.
- Remove XHTML content-type meta header from HTML documents, reverting
back to the meta charset
- Give XHTML documents their own XHTML declaration
- Remove now-redundant meta charset from XHTML
- Since XHTML and HTML documents differ now, compress after running
xhtmlize and make xhtmlize only act on uncompressed files.
- Validate XHTML using vnu
The site now has polygot markup and can handle both XHTML5 and HTML5
parsing rules. My staging site will be XHTML but my main site will be
HTML5, just in case of parse errors.
If other tools (e.g. LightHouse) end up supporting XHTML5, I'll consider
switching the content-type to XHTML.
I go out of my way to make sure the site works on viewports well under
240 CSS pixels wide. Some pages are very long, so a single mistake in
one part of the page shouldn't cause small text throughout the page.
Some article extractors get confused when the article body references
another author with microdata. Declare the author itemprop a second time
to make them happy.
