Hi, I’m a new member of the Python Documentation Community Team.
Is there a convenient way to get data on incoming web server requests for missing URLs at PSF documentation sites like docs.python.org? E.g. an incoming request for a URL like https://docs.python.org/3/reference/missingpage.html#missingfragment, where either “missingpage.html” does not exist, or “missingpage.html” exists but there is no anchor “missingfragment” within that page?
At our community team meeting on 7 Nov, we were talking about changing the names of some terms, and those terms happened to be used in fragment names within links. We fear that if we change the terms, we break those links. That is, any person or search engine outside our site who saved deep links to specific locations within our pages might have the old fragment names, and those links would no longer go to the right place in the page. In the case of a page we renamed or deleted, the incoming request would be met with an HTTP error 404.
It would be helpful to inform that discussion with data. Specifically, if we were to change links, could we measure the harm it caused in ensuing weeks? And, can we measure how many requests come in for links which we already broke? How big of a problem are broken links for us?
I’d appreciate insight into whether getting this data is an easy or hard thing to do, and how we would go about getting it. Thank you!