Re: UUID v7

2023-09-01 Thread Tom Glod via use-livecode
Oh yeah,that makes sense usually... just a different use case that is all. On Fri, Sep 1, 2023 at 3:02 PM Mike Kerner via use-livecode < use-livecode@lists.runrev.com> wrote: > it's an interesting topic, because in one of the db communities i'm > involved with, they're actually going the

Re: UUID v7

2023-09-01 Thread Mike Kerner via use-livecode
it's an interesting topic, because in one of the db communities i'm involved with, they're actually going the other way, increasing entropy. On Fri, Sep 1, 2023 at 1:56 PM Tom Glod via use-livecode < use-livecode@lists.runrev.com> wrote: > Hi Mike, > > Sometimes you just don't need uniqueness

Re: UUID v7

2023-09-01 Thread Tom Glod via use-livecode
Hi Mike, Sometimes you just don't need uniqueness across all bits, and if you do, use v4. In my example, the ID is used for clipboard clips, but now I don't have to sort them based on their timestamp. Also this improves database performance as the btree is better organized. There are a few videos

Re: UUID v7

2023-09-01 Thread Mike Kerner via use-livecode
why would reducing randomness be desirable in an internet-facing app? isn't the whole point to * ensure uniqueness across the entire space * make it nearly impossible to guess an ID? i would think that once you make the id's sequential, you have eliminated the ability to stop an authorized user

UUID v7

2023-09-01 Thread Tom Glod via use-livecode
Hi Folks, Sharing this because its useful, and also, more eyeballs on the code makes sense. I implemented this with the help of chatGPT. This is a handler that can generate v7 UUIDs. v7 UUIDs work better in databases, because they are not so random, improving performance. And they are