Jacob Champion <jchampion(at)timescale(dot)com> writes:
> On 7/19/22 09:14, Andres Freund wrote:
>> IMO this should replace the existing ascii escape function instead.
> That will affect the existing behavior of application_name and
> cluster_name; is that acceptable?
I think Andres' point is exactly that these should all act alike.
Having said that, I struggle to see why we are panicking about badly
encoded log data from this source while blithely ignoring the problems
posed by non-ASCII role names, database names, and tablespace names.
regards, tom lane