On 11/6/19 10:49, Peter Eisentraut wrote:
> On 2019-04-09 08:04, Peter Eisentraut wrote:
>> On 2019-04-08 23:06, Andres Freund wrote:
>>> The randomness based UUID generators don't really have dependencies, now
>>> that we have a dependency on strong randomness. I kinda thing the
>>> dependency argument actually works *against* uuid-ossp - precisely
>>> because of its dependencies (which also vary by OS) it's not a proper
>>> replacement for a type of facility a very sizable fraction of our users
>>> need.
>> Yeah, I think implementing a v4 generator in core would be trivial and
>> address almost everyone's requirements.
> Here is a proposed patch for this. I did a fair bit of looking around
> in other systems for a naming pattern but didn't find anything
> consistent. So I ended up just taking the function name and code from
> pgcrypto.
>
> As you can see, the code is trivial and has no external dependencies. I
> think this would significantly upgrade the usability of the uuid type.
Yes, indeed. Thanks!
This is definitively a good step towards removing external dependencies
for general usage of UUIDs. As recently commented, enabling extensions
at some MSPs/Cloud providers can be a bit challenging.
I wonder whether re-implementing some more of the extension's (ie. UUID
v5) in terms of PgCrypto and in-core makes sense / would actually be
accepted into core?
I assume that Peter would like to commit that potential patch series?
Thanks,
/ J.L.