We have a piece of code in legacyurls.py in the website project that handles URLs that are from the old website. And by old website I don't mean the one from half a year ago, I mean the one from 10 years ago.
This one handles:
/about/press/presskit/<ver>.html.<language>
/about/news.<nnn> for news articles (it has been /about/news/<nn>/ for 10 years now)
/about/event.<nnn> for events (it has been /about/events/<nn>/ for 10 years now)
/community/signup (it's now /account/signup)
I think all those should just be dropped. We could investigate if they are actually hit by anything, but meh, they're 10 years old and those are links that are just going to have to 404 in my opinion.
Then it handles:
/what/ever.html -> /what/ever/ rewrites.
Slightly less sure about that one, but again, maybe we should just kill that one as well once and for all?
Finally it has:
/mailpref/<listname> - which used to redirect to mj2 and now redirects to pglister.
That one we should probalby keep around a while?
Finally, we also have some rules that are commented as "legacy even on the old site", meaning they were legacy 10 years ago:
/developer/sourcecode/ -> /developer/coding/
/developer/bios/ -> /community/contributors/
/docs/tecdocs/ -> wiki
/docs/faqs.FAQ.html -> wiki/FAQ
/docs/faqs.FAQ_DEV.html -> wiki/Development_information
/docs/faqs.TODO.html -> wiki/Todo
/about/license -> /about/licence
Of those I suggest we keep the license redirect (some people consider American English not to be legacy after all), but nuke the rest.
Thoughts? Any of them worthy of saving? Or does somebody actually think it's worth investing the work to track exactly how much that old stuff is used?