conna666

First, let me describe the issue. When using a CDN other than Cloudflare, your static files including images will be served from a CDN subdomain, like cdn.mydomain.com. However, the image URLs in the XML sitemap will remain mydomain.com/image.jpg instead of cdn.mydomain.com/image.jpg. This would be fine IF Google would actually honor canonical tags in the HTTP headers of image URLs that these CDN providers send. However, the sad reality is that they don't honor canonicals of image URLs, only for web search (ie. normal HTML pages including say PDFs). This has been tested by me and others (https://blog.strategiq.co/do-canonical-tags-on-images-work-a-brighton-seo-experiment-512808985296 see last sub heading "Canonical from image to image…") And Google confirm this indirectly. Here: they say that canonical headers are supported for web search results only, which excludes image search. The example they use is a PDF, which shows in web search, not in image search. Here's the article: https://support.google.com/webmasters/answer/139066?hl=en And John Mueller confirms this: https://www.seroundtable.com/google-rel-canonical-for-images-nope-22549.html Here are two articles that describe the correct solution to this issue, which is to simply use the CDN subdomain in the XML sitemap for image URLs instead of the origin domain. See https://crunchify.com/maxcdn-and-wordpress-image-index-seo-issue-in-google-search-console-fix-cdn-image-links-and-avoid-duplicate-content-penalty/ and http://www.gauraw.com/get-google-index-website-images-using-cdn-services/ Now my idea is that it would be amazing if Rank Math would simply add an optional input field to general sitemap settings called "CDN Hostname" (with description "Enter your CDN hostname here without https:// prefix if you host your images on a CDN domain that's indexable and different from your main domain. This will replace the hostname in image URLs in the sitemap with the CDN hostname. Ex.: cdn.example.com") that would replace the hostname in image URLs (ie. <image:loc>) with the input value. So for example https://example.com/wp-content/uploads/image.jpg would be replaced with https://cdn.example.com/wp-content/uploads/image.jpg within the XML sitemap if the input field value is "cdn.example.com". Below is a code snippet that should do exactly that which Alex came up with: function rankmath_cdn_image_seo_fix( $uri ) { return str_replace( 'https://example.com', 'https://cdn.example.com', $uri ); } add_filter( 'xml_img_src', 'rankmath_cdn_image_seo_fix' ); However, it would be much better to have this as an option in Rank Math as described above. It should be easy to implement.

Awesome, thanks a ton! Certainly no reason to apologize - 4 days for a fix that include a weekend is amazing considering this is free. Most dev teams of commercial software probably would have needed longer to close a critical vulnerability than you took to fix this bug that seems to be isolated to specific setups (I assume you would have catched it much sooner otherwise). It all works as expected now. Thanks again!

I'm sorry, but I wouldn't open a ticket if this would be working just fine like you say. Please see redirect with ID 4. I added it with a trailing slash after the destination URL, which got trimmed. And it's not only trimmed in the GUI like you say, but actually redirects to the URL without trailing slash, resulting in a redirect chain.

Hey guys. Awesome plugin, thanks for your hard work! The biggest issue I face with it though is that the redirection part is completely broken. Importing redirects from Redirection plugin messed up all the redirects - none of them is working properly afterwards. Fe. some rules were redirecting home to an inner page because Rank Math somehow changed them during import (no such rule existed in Redirection). Now even without an import from Redirection it doesn't seem to be possible to get redirects working properly. When I add say this rule: https://i.imgur.com/wbi4AF8.png It will end up like this after saving: https://i.imgur.com/4qvAueq.png ... which breaks the redirect rule and also results in a redirect chain because for some reason Rank Math keeps removing the trailing slash from the destination URL.