Can't fix Tags moved to private still show in sitemap

[Mr Lucky]

Affected version

2.0.9

I have been getting errors in Google console due to tags on threads that were public but subsequently made private but are still indexed in the sitemap.

The thread itself is no longer in the sitemap, but tags on that thread are still in it.

Surely if a tag becomes not viewable publicly, it should not be in the sitemap. In this case the tag in question is only used once on that now private thread.

From a current sitemap:

View attachment 183006

I also notice that if you are not logged in (so theoretically without permissions to view that tag) and then search in tags for repairer, then the autocomplete shows the full tag, inviting you to click on the tag and you then get No results found. (as opposed to

View attachment 183007

As expected, any tags created on that thread after it became private don't show in the sitemap or tag search autocomplete.

I think this will explain why Google Console is complaining and I believe that threads moved to private are no longer in the sitemap, so neither should be tags that are unique to a private thread.

 

[Chris D]

The attachments didn’t work for some reason.

What are the actual errors in the search console?

 

[Mr Lucky]

The attachments didn’t work for some reason.

It was a saved draft, could that be why?

I've uploaded again and edited into above post.

Google Console was soft 404

 

[Chris D]

Soft 404 pages aren't necessarily "errors" or huge problems. They're more informational than anything.

The difficulty with tags is that we can't actually display a hard 404 here because that wouldn't be semantically correct. The tag actually does exist, but there just so happens to be no visible tags to the visitor. At any point in the future, that page could start re-listing content on it as new threads use that tag or the thread may be moved to a more visible location. Not displaying a hard 404 means those tag pages won't drop off the index (which would be problematic in case they ever do contain real visible content in the future).

We also don't (and won't) track the viewability of tags in this way on a per tag basis. If we're generating a sitemap that contains thousands of tags, it's not practical to analyse every single tag usage to determine if they are visible or not.

So, all things considered, the way we currently handle this is the best way to, and not something we're able to change, but also not technically wrong, either.

 

[Mr Lucky]

Fair enough.