Thread: Re: Variable substitution in psql backtick expansion

I assume we want to keep that pre-existing behavior of \set in
psql, that is, making a copy of that string and associating a
name to it, whereas I guess pgbench computes a value from it and
stores that value.