The following documentation comment has been logged on the website:
Page: https://www.postgresql.org/docs/16/git.html
Description:
>ISO 8601 specifies the use of uppercase letter T to separate the date and
time. PostgreSQL accepts that format on input, but on output it uses a space
rather than T, as shown above. This is for readability and for consistency
with RFC 3339 as well as some other database systems.
https://www.postgresql.org/docs/current/datatype-datetime.html
Short answer: T (or t as discouraged alternative).
After reading on this as much as I could, it turns out the time separator
must be a T or t. What has made think this way is first of all this thread
in the GNU lists where F. Alexander Njemz contacted the authors of RFC3339
Graham Klyne and Chris Newman asking if T is mandatory and got this response
from Mr. Klyne:
> In short: "yes"
>
> Per section 5.5, the intent in this draft was to specify a timestamp
format using elements from and compatible with 8601, but eliminating as far
as reasonable any variations that could make timestamp data harder to
process. This includes making the 'T' mandatory in date+time values.
Just for clarity's sake, this is stated in the section 5.5:
> Simplicity is achieved by making most fields and punctuation mandatory.
This clearly clashes with a non-mandatory T and strongly makes me think that
the this syntax in that problematic passage refers to ISO8601 and not
RFC3339.
https://stackoverflow.com/questions/63783868/what-are-valid-date-time-separators-in-rfc3339-strings/63882162#63882162