Re:Using sqlite3 as a notekeeping document graph
by bccdee on Saturday July 03, 2021 @11:28AM
(#27722236)
I suppose it further reduces the chance of a collision. If you're in a totally random space, there's always a chance someone will generate a collision if you wait long enough. But if you follow the timestamp and MAC address rules, then for a collision to happen it basically needs to be the same computer hitting the same number twice on the RNG within one second, which is less likely.
But, since that only holds true if people follow the rules, it's worth standardizing on either v1 or v4, and so it's good to be able to distinguish between the two for validation purposes.
What good rises out of including extra paperwork with each number, essentially saying "attention please, a random number", or "this is a timestamp, a monotonic counter and a MAC address", like if it was a newfangled smartphone? The RFC also says that there is no mechanism for validating a UUID, save for checking if its timestamp part is in the future, for versions that employ such a portion. Why can't any 128 bit value be accepted on the receiving end?