Tweet · 14 Jun 2019, 19:32 UTC

Frankly, by today’s standards a local cache of article titles probably wouldn’t be _that_ much. Less perfect than matching the content, but should catch lots, could be refreshed regularly.