On 2024-Mar-01, Euler Taveira wrote:
> I don't like to break backward compatibility but in this case I suspect that it
> is ok. I don't recall the last time I saw a script that makes use of -d option.
> How often do you need a pgbench debug information?
I wondered what the difference actually is, so I checked. In -i mode,
the only difference is that if the tables don't exist before hand, we
receive the NOTICE that it doesn't. In normal mode, the -d switch emits
so much junk that I would believe if somebody told me that passing -d
distorted the benchmark results; and it's hard to believe that such
output is valuable for anything other than debugging pgbench itself.
All in all, I support the original patch.
--
Álvaro Herrera 48°01'N 7°57'E — https://www.EnterpriseDB.com/
"I love the Postgres community. It's all about doing things _properly_. :-)"
(David Garamond)