4 ms·
This brings up an interesting ergonomics problem. By naming them UUIDv4 and UUIDv7, is it going to be this never ending confusion for people to have to remembe
by atonse 3y ago
This brings up an interesting ergonomics problem.
By naming them UUIDv4 and UUIDv7, is it going to be this never ending confusion for people to have to remember which one is good for databases and which one good for one time tokens?
Not sure what the backwards compatible solution here is either.
In elixir the function is UUID.uuid4() to generate a v4 UUID.
So we could theoretically scan code for its use I suppose. But all this increases chances of errors.
- hot_gril 3y ago> is it going to be this never ending confusion for people to have to remember which one is good for databases and which one good for one time tokens Yes, because this is what's been happening already with the past versions. It's not just sequential and random, there are also hash-based UUIDs. They shouldn't have sequential (heh) version numbers.