Hi, I'm proposing to sunset "Extracted page abstracts for Yahoo" dumps. We have to maintain a dedicated extension called ActiveAbstract https://www.mediawiki.org/wiki/Extension:ActiveAbstract just to produce this type of dumps while It's rather easy to extract the same values from the main dump. There are also other reasons outlined in the ticket: T382069 https://phabricator.wikimedia.org/T382069
If you rely on this type of dumps and can't easily switch or if I missed something obvious about page abstract dumps or ActiveAbstract extension, please make a comment in the ticket.
Thank you