Excellent Episode. I actually love the idea of partitioning by the timestamp component. Having a function to extract that. And I believe ONE of the UUID standards allowed for a "character" that could be assigned to the specific server, which was a way cool way to prevent duplicates with the loss of randomness... And the SAVINGS of removing the created_at column... Nice. And not having to index that (right?). Especially if you are mostly sorting, or filtering AFTER the find... That's a lot of small wins! I am curious what ChatGPT thinks about it? (LOL)
Пікірлер: 2