v2.2

Changed discovery behaviour that non-existing database clusters are not deleted anymore including new parameter --clean-pgclustertab. See chapter Discovery

v2.1

Improved cluster detection (issue #12) to reduce performance overhead
Improved scriptmgr.sh prep call (issue #11)
- Recreate file $PGBASENV_BASE/scripts/.run.${TVD_PGVERSION} only if content of $PGBASENV_BASE/scripts/*.sql changes
- File locking for $PGBASENV_BASE/scripts/.run.${TVD_PGVERSION} to avoid concurrency issues
Fixed issue #10 related to VERSION_NAME query in .psqlrc

pgsetenv.sh now accepts --noscan option. It will source the pgBaseEnv without scanning the filesystem. This argument can be used also in $HOME/.pgbasenv_profile to prevent filesystem scan every time on login.

Additional information about standby status and incoming data stream will be reported. Now status of the wal receiver and master node:port will be reported. Standby status string: Cluster role: STANDBY [ Status: streaming Master: masternode:5435 In recovery: YES ]. As usual all these information will be available from the evn variables TVD_PGMASTER_HOST, TVD_PGMASTER_PORT and TVD_PGSTANDBY_STATUS.
Timeouts was added to find commands to prevent hangs. Parameter PGBASENV_SEARCH_TIMEOUT can be set in pgbasenv.conf to control this timout. By default it is 5 sec.

$PGBASENV_BASE/scripts was introduced to store the user scripts and inject them into psql. The list of the scripts will be available in psql under :scripts variable. Each script can be executed by enetering its own variable name. It is possible to define on which PostgreSQL version the particular script will be visible. Check corresponding section in main README.md file for details.
To prevent pgclustertab and pghometab corruption during simultaneous execution, operations will be executed over flock utility. It will guarantee a sequential access to those files and pgbasenv.sh execution.
New styling was added to the output of the u alias (pgup.sh).
Other minor changes.