writes. The first worker still uses "pgbench_log.<pid>" for the name, but
additional workers use "pgbench_log.<pid>.<serial-number>" instead.

Reported by Greg Smith.

consistency.

Gabrielle (Roth)

Document this restriction.

Also update tagging instructions, and add id tags to a few documentation
sections.

did with server-side languages.

Docs were unclear on whether or not database=replication was required,
nor did they mention the FATAL error this causes if database parameter is
mentioned explicitly, whatever its value.

with a few strategically placed pg_verifymbstr calls.

Also make a couple other minor editorial improvements.

in recovery_end_command, it always came out as 0 because InRedo was
cleared before recovery_end_command was executed. Also, always take
ControlFileLock when reading checkpoint location for %r.

The recovery_end_command bug and the missing locking was present in 8.4
as well, that part of this patch will be backported separately.

Fujii Masao

Also fix and uncomment an old example of creating a GIST index, and make
a couple of other minor editorial adjustments.

Add ERRCODE_INVALID_PASSWORD sqlstate error code.

In particular, the assertion that shmall is sufficiently sized by default
is slowly becoming untrue.

7.4.28.

Tim Landscheidt

file instead of ~/.psqlrc on startup.

about the behavior.

Document that quotes in to_date, to_timestamp, to_number skip input
characters.

into proper sections, per suggestion from Tom.

two adjacent long-running queries have much less than max_standby_delay
before query cancel is possible.

from master to slave, and discourage its use during slave/master
keep-xid-alive connections.