Hello
@Alpha1
Thank you for your notes.
In general, we are not planning large enhancements for the product, and we are adding only features we consider essential or very important for its functionality. Any such add-on can be enhanced to have 10 times more features than they have, but each such enhancement is very time consuming. The product has been intially developed for admins to "find dead links", and it has grown into a product with tens of times more features than just "finding the dead links", so I hope the feature set in general is satisfactory.
Here are some additional notes regarding some of your comments:
I have noticed that the link checker does not distinguish between deleted posts and normal posts. In the case of normal posts, we want to fix or delete our dead links in order to improve user experience and SEO. In the case of deleted content we want to keep dead links, because this is often evidence of abuse. user experience and SEO are no factors with deleted posts, so there is no need to check hundreds of thousands of links in deleted posts.
Link checker is not meant to validate your content, it validates links, based on HTTP response they give - no matter if the link is on your board or on someone else's board. None of the validation features it has depend on internals of your database, but only on HTTP status received from the link.
It can be enhanced and its link index can be used to find content on your board based on its status in the database, but those are not features are planning to add into this add-on.
Please also consider an option to check internal links that are restricted to guests. We have hundreds of thousands of links which are only accessible to registered members.
All links are validated as "guest", simply because that's how HTTP works, that's how search engines crawl your website. Links accessible to registered users will have appropriate HTTP status for search engines, and that's the status you see in the product.
One thing that would be mighty handy to have is a list top dead link urls. i.e. display the 40 most common url patterns that have dead links. For example if you have a lot of internal dead links because a directory no longer exists, then display the url to the directory. This would make it easy to identify the main issues that we need to fix.
I am afraid the feature is a bit hard to implement, especially based on a "pattern". You can use SQL queries to find such queries, as all links are indexed in the database along with their status.
Since google has forced pretty much all sites to change from http to https, this change accounts for a very large percentage of dead links found. Probably around 40%. The links work fine, but no longer on http. As is we would automatically delete hundreds of thousands of valid links because the addon lists these as invalid. Please add function to select http links with a specific status and check for the https version.
I am not quite sure what you mean by Google forcing https anyhow making http links invalid. Sites should still return correct response code for http links, e.g. both links in the screenshot are valid:
Anyway, if you think you have links, which are shown as invalid just because they use http://, you can use the batch replacement feature to turn them to https://
You can of course filter the list by the status, to make sure you work only with http:// URLs which are currently invalid.
5. Currently when removing dead links it only seems possible to either keep or remove the link anchor text. But this botches up a lot of posts as many posts have part of the written text hot-linked and removing that means removing words or phrases out of a text.If there is no anchor text or if the link anchor is just the url again, then it makes sense to completely remove it. However, when the link anchor is normal text then it makes sense to keep the anchor text.Please add an option to delete if there is no anchor or if the anchor has URL syntax, while keeping the text if the anchor is other text.
The product works almost as described, except that if the anchor text is a URL, it would keep the URL:
We have updated the product to handle that case as well now, in case of the last row in the screenshot above the link will also be completely deleted, as it contains a URL as an anchor text.
6. When there are a large number of dead links for a status, it would be helpful to be able to review and process those in batches.
You can set the number of items you see in the admin panel preview using the option "Preview Limit". In the preview, use the checkbox on the top-left corner to select all links or just some links to process:
7. It would also be really helpful if there was a function to go to the URL of the dead link.
We have added the link in the preview:
8. For those of us who run cookieless domains it would be nice to be able to define those domain as internal domains.
Internal domain check is done with basic text match in MySQL, so adding an ability to specify multiple such domains is error prune. I also don't see why exactly cookie-less domains should be considered as internal links, as they are simply not and the fact they are cookieless does not anyhow affect the HTTP status of the links pointing to them. As I understand this is not a crucial feature to implement for now so we would prefer not to modify this aspect of the product and risk causing further issues related to it.
We will release the new version within some hours.
Thank you!
