WHEN IMPERFECT SYSTEMS ARE GOOD, ACTUALLY: BLUESKY'S LOSSY TIMELINES
Bluesky's Following Feed faced performance issues due to "hot shards" caused by users who follow a massive amount of people. To address this, Bluesky implemented a "lossy" timeline approach that probabilistically drops writes to timelines of users who follow an excessive number of accounts. This sacrifice in perfect consistency reduced write latency and improved overall system performance without negatively impacting user experience.