Robert Haas <> writes:
> On Wed, Nov 9, 2016 at 9:02 AM, Peter Eisentraut
> <> wrote:
>> On 11/7/16 1:13 AM, Haribabu Kommi wrote:
>>> Yes, I agree that default zone is the main use case of the original thread.
>>> From the RFC 4007, the default zone is used for the global addresses,
>>> This may be the main use case with zone id. How about currently just
>>> ignoring it and store the actual IP address with the attached patch and
>>> handle the rest of the actual zone id support later once the it gets
>>> properly standardized?

>> Well, according to the RFC, the default zone is 0 "typically", which is
>> a very weak requirement.  So just ignoring it is probably also not right.
>> So far we have only heard one use case for any of this, which is someone
>> wanting to store ::1%0, which is not even a valid address according to
>> that same RFC.  So this is all on very weak ground.
>> I think we should just forget about this.  It's all a bit too dubious.

> +1.

Agreed, let's wait until more standardization emerges.  Anything we do now
risks painting ourselves into a corner, and there is not so much demand
for a feature in this area that we need to do something about it Right Now.

                        regards, tom lane

Sent via pgsql-hackers mailing list (
To make changes to your subscription:

Reply via email to