How it works / 4

What a reader sees and can use

Each feature below was checked by fetching live pages from this site. Where a check only proved that the feature exists in the page, and not that it works, the card says so.

On every page

Built into the page frame

Main navigation

A labelled navigation bar and a menu toggle for small screens. Present in the HTML of every page fetched.

Search overlay

Open and close search controls and a search box are in the page. Typing a query was not tested here.

Dark mode

A control labelled for switching between dark and light mode is in the header, and the styles carry dark variants.

Back to top

A back-to-top control is present on every page fetched.

Sticky call to action

Sticky call-to-action markup is present in the page code. How it behaves on screen was not tested here.

Consent banner

A consent banner and a privacy-choices control are present. What each choice switches off was not tested here.

Images. Images are served from the site's own /img/ path as WebP files with a one-year immutable cache header. On the article pages checked, the first image loads eagerly, carries a srcset and fixed dimensions, and later images load lazily. I did not find a high fetch-priority hint on the pages I fetched, so I claim eager loading only, not priority loading.

What was not checked. On-screen behavior of the sticky call to action, search results, consent choices and form submission. Those are listed as present, not as proven.

Content a reader can take

Feeds, forms and files

RSS feed

The address /rss.xml returns an RSS feed of the site's articles.

Search index

The address /search.json returns a list of 21 articles with title, excerpt, address, tags and date.

Newsletter sign-up

A newsletter sign-up form is present on the lab landing page. Submitting it was not tested here.

Downloadable PDFs

Articles link to PDF files served under /pdf/. The PDF article pages returned 200 and carry those links.

Questions a client asks

Is there a separate search page?

No. The address /search returns a not-found page. Search is an overlay opened from the header.

Do these features need extra scripts per page?

I did not measure script weight, so I make no claim about it.

Keep reading