1
0
Fork 0
mirror of https://git.sr.ht/~seirdy/seirdy.one synced 2024-11-30 15:22:09 +00:00

Compare commits

..

No commits in common. "aa79edeb21dc2fb851b89ca6ceb4f707264b1d2b" and "7c835a14e42247f633ba627696cf3750f9b0f6a8" have entirely different histories.

7 changed files with 43 additions and 44 deletions

View file

@ -2,17 +2,16 @@
* touch-friendly layout.
* Nothing here exists purely for aesthetics; everything addresses a
* specific a11y, compatibility, or critical
* usability need.
* usability need. The only exception is the h1's top margin.
*
* Two exceptions: I re-set the input styles so Safari wouldn't make
* them pill-shaped, and I tweaked some margins to make some things
* evenly aligned.
* One exception: I re-set the input styles so Safari wouldn't make them
* pill-shaped.
*
* I also don't use any classes except to optimize embedded content.
* (e.g. to specify that an image should have pixelated rendering or be
* inverted in dark mode). One exception: a class for narrow width body
* text. My HTML contains microformats2 classnames for IndieWeb parsers,
* but I don't actually use those for styling.
* I also don't use any classes except for image presentation. (e.g. to
* specify that an image should have pixelated rendering or be inverted
* in dark mode). One exception: a class for narrow width body text.
* My HTML contains microformats2 classnames for IndieWeb parsers, but I
* don't actually use those for styling.
*
* To keep myself from caring about minute details, I limited myself to
* only defining spacing in increments of .25em. Pixels are 1px or
@ -78,7 +77,7 @@ html {
* containing lists of interactives, with 8px in between.
* Don't use relative units here; this margin shouldn't scale with
* zoom, or else text will get narrower with zoom. */
padding: 0 16px;
padding: 0 24px;
}
/* 45em is too wide for long body text.
@ -90,18 +89,6 @@ html {
max-width: 35em;
}
/* Enable containment, especially useful for achive pages with
* long lists of content. */
body > :not(main),
main > :not(article),
article > * {
contain: content;
/* Add padding on both sides so that focus outlines don't escape their
* containers. This will let us enable CSS containment without
* clipping overflowing elements. */
padding: 0 .5em;
}
summary {
/* It's not obvious that a <summary> has button semantics.
* "cursor: pointer" is used on MDN, web.dev, GitHub, gov.uk, and
@ -150,13 +137,17 @@ html {
margin: .75em 0;
}
dt > a {
margin: -.75em 0 -.75em -.25em;
display: inline-block;
}
/* <dt> should not be closer to the previous <dd> than the following <dd>.
* That can happen in webmentions. */
dd {
padding-bottom: .25em;
}
dt > a,
aside > a {
margin: -.75em -.25em;
}
@ -164,8 +155,7 @@ html {
a[href="#h1"], /* skip link */
div[itemprop="comment"] dd > a ,
footer > nav,
/* List items with direct hyperlink children should only have one
* hyperlink. */
/* List items with direct hyperlink children should only have one hyperlink. */
li > a,
aside > a,
nav ol a {
@ -173,8 +163,7 @@ html {
display: inline-block;
}
/* The nav has to be distant-enough from the top to make room for a
* skip-link. */
/* The nav has to be distant-enough from the top to make room for a skip-link. */
header > nav {
margin: .75em 0 0 -.25em;
display: block;
@ -338,15 +327,15 @@ samp {
/* Some browsers don't support the hidden attr.
* I use hidden spans in backlinks to provide ARIA labels.
* Some ancient browsers don't support input[type="hidden"] */
[hidden],
span[hidden],
input[type="hidden"] {
display: none;
}
/* Remove list style from data feeds. */
[itemtype="https://schema.org/DataFeed"] > ol {
list-style-type: none;
main[itemtype="https://schema.org/DataFeed"] > ol {
padding: 0;
list-style-type: none;
}
/* Narrow screens: long words can cause content to flow out of the
@ -362,7 +351,6 @@ span[itemtype="https://schema.org/Person"] {
/* Narrow screens: allow horizontal scroll in a pre block. */
pre {
contain: content;
overflow: auto;
padding: .5em;
}
@ -402,7 +390,6 @@ div[itemprop="articleBody"] img {
max-width: 100%;
}
/* Stretch out audio elements so the progress meter is easier to use. */
audio {
width: 100%;
}
@ -428,8 +415,8 @@ form > div {
input {
/* Browsers like Safari make the submit button pill-shaped which
* clashes with the input box. One of the only purely-cosmetic changes
* on this site. */
* clashes with the input box. The only purely-cosmetic change on this
* site. */
appearance: none;
/* Don't shrink the size of the text in forms or make it system-ui. */
@ -479,6 +466,11 @@ form :focus {
}
/* Todo:
* - Some browsers don't scale SVGs properly; the img container
* dimensions crop the image rather than scale it. Investigate
* if this only applies to Internet Explorer or if it applies to
* other uncommon browsers too. If any non-IE browsers do this and/or
* if the spec allows this behavior, try to correct it here.
* - Wait till Webkit fixes its broken-ass default dark stylesheet
* then consider trimming the dark stylesheet I provide.
* */

Binary file not shown.

Before

Width:  |  Height:  |  Size: 976 B

After

Width:  |  Height:  |  Size: 1,003 B

Binary file not shown.

Before

Width:  |  Height:  |  Size: 1,017 B

After

Width:  |  Height:  |  Size: 1,004 B

View file

@ -71,7 +71,7 @@ The website is built on well structured, semantic, [polygot XHTML5](https://www.
This site sticks to Web standards. I regularly run a local build of [the Nu HTML Checker](https://github.com/validator/validator), `xmllint`, and [html proofer](https://github.com/gjtorikian/html-proofer) on every page in my sitemap, and see no errors. I do [filter out false Nu positives](https://git.sr.ht/~seirdy/seirdy.one/tree/master/item/linter-configs/vnu_filter.jq) and report them upstream when I can.
I also perform cross-browser testing for both HTML and XHTML versions of my pages. I test with, but do not necessarily endorse, a large variety of browsers:
I also perform cross-browser testing for both HTML and XHTML versions of my pages:
- I maintain excellent compatibility with **mainstream engines:** Blink (Chromium and Edge), WebKit (Safari, Epiphany), and Gecko (Firefox). The hidden service also works well with the Tor Browser.
@ -79,7 +79,7 @@ I also perform cross-browser testing for both HTML and XHTML versions of my page
- I also regularly test compatibility with **current alternative engines:** the SerenityOS browser, Servo, NetSurf, Dillo, Kristall, and Goanna (Pale Moon's Gecko fork). I have excellent compatibility with Goanna and Servo. The site is usable in NetSurf, Dillo, and the SerenityOS browser; the always-expanded `<details>` elements might look odd. [The SerenityOS browser doesn't support ECDSA certificates](https://github.com/SerenityOS/serenity/issues/14160), but the Tildeverse mirror works fine. It also has some issues displaying my SVG avatar; it does not attempt to use the PNG fallback.
- I occasionally test **abandoned engines,** sometimes with a TLS-terminating proxy if necessary. These engines include Tkhtml, KHTML, Internet Explorer (with and without compatibility mode), Netscape Navigator, Opera Presto,[^1] and outdated versions of current browsers. The aforementioned issue with `<details>` applies to all of these choices. I use Linux, but testing in browsers like Internet Explorer depends on my access to a Windows machine.
- I occasionally test **abandoned engines,** sometimes with a TLS-terminating proxy if necessary. These engines include Tkhtml, KHTML, Internet Explorer (with and without compatibility mode), Netscape Navigator, and outdated versions of current browsers. The aforementioned issue with `<details>` applies to all of these choices. I use Linux, but testing in browsers like Internet Explorer depends on my access to a Windows machine.
Some engines I have not yet tested, but hope to try in the future:
@ -87,7 +87,6 @@ Some engines I have not yet tested, but hope to try in the future:
- [Kozmonaut](https://github.com/twilco/kosmonaut)
- [Moon](https://github.com/ZeroX-DG/moon)
- [hastur](https://github.com/robinlinden/hastur)
- [Flow Browser](https://www.ekioh.com/flow-browser/)
Machine-friendliness
--------------------
@ -164,7 +163,3 @@ Privacy
This site is **privacy-respecting.** Its CSP blocks all scripts, third-party content, and other problematic features. I describe how I go out of my way to reduce the information you can transmit on this site in [my privacy policy](../privacy/).
[^1]: Opera Presto isn't really abandoned. Opera Mini's "Extreme" mode still uses a server-side Presto rendering engine; see {{<mention-work itemprop="citation" role="doc-credit" itemtype="Article">}}{{<cited-work name="Opera Browsers, Modes & Engines" url="https://dev.opera.com/articles/browsers-modes-engines/" extraName="headline">}}{{</mention-work>}}. That being said, I do test with the outdated desktop Presto engine in a sandboxed environment.

View file

@ -69,7 +69,7 @@
<a rel="alternate" href="http://wgq3bd2kqoybhstp77i3wrzbfnsyd27wt34psaja4grqiezqircorkyd.onion/">Tor mirror</a>
</li>
<li>
<a href="{{ site.BaseURL }}meta/privacy/">Privacy</a>
<a href="{{ site.BaseURL }}privacy/">Privacy</a>
</li>
<li>
<a href="{{ site.BaseURL }}about/#location-seirdy-online">Contact</a>

View file

@ -19,7 +19,7 @@
{{- if and (eq .Level 2) (.Page.Section) (ne .Page.Section "notes") -}}
<!--crimes-->
<aside role="none"><a
href="#{{ .Anchor | safeURL }}"
id="{{ .Anchor | safeURL }}-anchor" href="#{{ .Anchor | safeURL }}"
aria-labelledby="{{ .Anchor | safeURL }}-prefix {{ .Anchor | safeURL }}">
<span id="{{ .Anchor | safeURL }}-prefix">Permalink to section</span>
</a></aside>

View file

@ -15,7 +15,19 @@
)
or
( # the search page has raw templates, let those slide. I validate the final dynamic search page manually.
.url | test ("/search/index.")
(.url | test ("/search/index."))
and (
(.message == "Text not allowed in element “ol” in this context.")
or
(.message == "Bad value “{{ .URL }}” for attribute “href” on element “a”: Illegal character in path segment: “{” is not allowed.")
or
(.message == "Bad value “{{ .URL }}” for attribute “cite” on element “q”: Illegal character in path segment: “{” is not allowed.")
)
and (
(.extract | test("{{"))
or
(.extract | test("}}"))
)
)
) | not
)