Joshua Wehner

:ok: I *think* this is ready for review. I wound up hedging a bit on the ranking query -- the new query maybe isn't quite as "advanced" at ranking users,...

Yes, I was able to replicate. I don't think the problem here is caching, per se, but background jobs & race conditions they bring in: * Prior to https://github.com/forem/forem/pull/11283, we...

Should be fixed by https://github.com/forem/forem/pull/18438

@fdocr - Hey, just wanted to check in and see if there's anything I can do to help move this PR forward?

@mirie Yep! If we find that there's more that needs doing, we can re-open, or open a new issue.

@joshpuetz I'm a little confused on this one? 🤯 I was expecting that an individual reactable would potentially have many, many reactions, and that we'd be more interested in the...

@fdocr - Hey, just wanted to check in and see if there's anything I can do to help move this PR forward?

1. We *are* tracking post-article display ads via the same mechanism as we have had for sidebar display ads (that is: if a user is logged-in, we record an `impression`;...

We will need a migration to remove foreign-key constraints and make `user_id` optional on `display_ad_events`. The history of this is potentially interesting: * There _was_ a `counts_for` column on `display_ad_event`:...

So, I'm just making some thoughts/notes here about the performance effects of these changes. From my perspective, there's maybe two concerns: 1. performance of a query to remove the `user_id`...