Damdit 1.2: the whole site, not only the pages it opens
Broken links on every page, orphan pages, links to other websites, a missing 404, canonicals, duplicate pages, oversized images and handover checks. Included in your subscription.
Damdit 1.2 is out. If you already have Damdit, it will offer the update the next time it checks; nothing changes until you say yes.
The whole site, not only the pages it opens
Damdit opens up to 500 pages in a real browser and checks each one in depth: accessibility, performance, content. Until now, a site with 3,000 pages meant 2,500 pages Damdit never looked at.
Now it reads the rest of the site too, without a browser, up to 5,000 addresses. That’s enough to follow every link and read every title and description. A broken link, a redirect or a missing description on page 4,000 shows up like one on the home page. Coverage tells you how many pages were checked this way, so you know what was measured in depth and what wasn’t.
What that makes possible
- Orphan pages. A page your sitemap or Webflow lists, but no link on the site leads to. Visitors can’t click through to it. Damdit says so only when it read the whole site; on a bigger site it tells you it couldn’t tell, instead of guessing.
- Pages too deep. A page four or more clicks from the home page, with the click path that gets there.
- Links that can’t work. A link to a page that redirects, an empty link, a link to
localhost, a mistypedhppts://. Each one lists every page it’s on and where on the page: header, navigation, footer or content. A broken link in the footer is one fix, not 500. - Links to other websites. Each external address is asked once, up to 1,000 per audit. Old articles are where they pile up: a startup that closed, a blog that moved, a page that’s gone. A site that refuses automated requests reads “Couldn’t verify”, never broken.
More checks
- A site with no real 404. Damdit asks for an address that can’t exist. If the site answers with a page or sends you to the home page, every mistyped link looks fine and search engines can index anything.
- Canonicals. Two different canonicals on one page, a canonical to a page that redirects or asks not to be indexed, and a canonical a script changes after the page loads.
- Duplicate pages. Pages with the same main content, the menu and footer left out, with the cause: a trailing slash, www, a URL parameter, or the same text at different addresses. Pages that are 90% alike go to Needs review with the lines that differ.
- Image sizes. An image larger than it’s shown, with how many kilobytes a fitting file would save, and an image without a height, which makes the page jump while it loads.
- The share image. A relative og:image address, a file over 5 MB or under 600 × 315 px, and a shape that networks crop.
- Link texts. Links that say only “Read more”, “Click here” or “Více”, in English, Czech, Slovak and German.
Your score will probably drop
Most of the new checks count in the score, so after the update a site can score lower without anything on it changing. Damdit 1.2 uses a new scoring version for that reason: your next audit isn’t compared with audits from earlier versions. History says the score isn’t comparable instead of showing a drop, and issues whose check changed read “Inconclusive” once instead of a false “Resolved”.
Calmer to read
The issue detail now starts with the fix and where to make it. The evidence is one example at a time. Needs review groups an element that appears on 500 pages into one decision, not 500. And when the next audit doesn’t reach the page where an issue was, the issue stays open as “Page not checked again” instead of looking fixed.
Handover checks are new too: each project keeps the things only a person can check before handing a site over, like “the contact form reaches HubSpot”. They’re in Needs review and in your client and developer reports.
Something looks wrong, or a check misfires on your site? Reply on Mastodon or write to support@damdit.com.
— Tony