News for hackers. Stuff that matters.
Re:Exploring PostgreSQL 18's new UUIDv7 support by Illniyar on Friday October 17, 2025 @06:50PM (#45623046)
If leaking creation time is a concern, can we not just fake the timestamp? We can do so in a way that most performance benefits remain - so like starting with a base time of 1970 and then adding base time to it intermittently, having random months and days to new records (or maybe based on the user's id - so the user's record are temporally consistent but they aren't with other user records).

I'm sure there might be a middle ground where most of the performance gains remain but the deanonymizing risk is greatly reduced.

Edit: encrypting the value in transit seems a simpler solution really

Re:Exploring PostgreSQL 18's new UUIDv7 support by hu3 on Friday October 17, 2025 @07:01PM (#45623144)
In that case, auto increments can also be bumped from time to time. And start from a billion.

They're more performant than uuidv7. Why would I still use UIID? Perhaps I would still want uuids because they can be generated in client and because they make incorrect JOINs return no rows.