The single largest group of broken links reported via the Google webmaster tools, and also noted by @mmokrejs (see http://mailman.open-bio.org/pipermail/biopython-dev/2016-April/021345.html) are URLs like this:
This page is still available via the primary OBF mailing list domain name:
They are also available via this secondary alias:
Ideally we'd automatically redirect http://lists.open-bio.org/pipermail/* to http://lists.open-bio.org/pipermail/* which was easy under Apache, but becomes harder under GitHub Pages.
See also #19 and #32 for more general issues about broken links.
We could easily add stub pages for:
Even adding the monthly overview pages would be quite easy. However, all the useful URLs pointing to these pages are to individual emails in the archives.
I think that the only way to fix all the links within GitHub Pages is literally to copy the mailman archives into a GitHub pages repository - perhaps another separate project one, i.e. branch gh-pages of https://github.com/biopython/pipermail in order to be visible as http://www.biopython.org/pipermail/aka http://biopython.org/pipermail/
This would have some advantages: It would be a nice backup, and we can put styling on top of the minimal mailman layout, and a one-off conversion/dump would fix all the existing URLs.
Potentially we could keep the GitHub mailman archive updated automatically - this would require a script running (ideally on the OBF mailman server which is possible as long as the load is minimal). Unfortunately that would create an unwelcome maintenance burden.
Tell search engines to ignore /pipermail/ mailing list archives
See issue #49, the mailman archives are still via e.g.
However, with the move of biopython.org to GitHub Pages
we can no longer do this on our main domain.
I've redirected the related /mailman/ URLs with 07d9e88