Quick list of outstanding items / bugs / feature requests for CP:

NOTE: All bugzilla items are now on github

* Add --setup and --reset flags to replicate_row to automate things.

* Add documentation on privs required for each action, and recommended grants for roles, etc.

* The same_schema action does not check indexes. See bugzilla #54

* Perform automatic creation of views and function to allow all actions to be run
  as non-superusers.

* Have the configuration file checker only report once on a changed checksum,
  then store the changed one somewhere for the next run

* Put on CPAN - worth doing as this in now in packages?

* Fix up the constraints checking for same_schema. It sometimes reports odd
  constraints with the name repeated many times.

* Ensure that all stats output is 100% Nagios certified,
  e.g. 'dbname'=val1;warn;crit;extra

* Make sure all the regexes (e.g. slurp =~ //) work on WIN32.

* Add a --output=plain, for simple human-readable output.

* Add a --output=csv, for importing into various targets.

* Language testing: force to EN?

* Allow a flag to determine if skipped into goes into perf or not

* Make a (semi) automated way to allow non-superusers to run all actions

* Put complete Nagios templates on a page somewhere

* Get more language translations

* Investigate report of an empty fsm_pages causing 255 out of range error

* Check the current logfile is not bigger than X KB/MB/GB

* Have the checksum action only warn once and save the result somewhere

* Reconsider PGBINDIR and see if there is something more standard

* Support for pg_bouncer checks

* Full and correct support for Nagios, e.g.
  http://nagiosplug.sourceforge.net/developer-guidelines.html#AEN203

* Test for all the Nagios output of each action.

* For last_analyze and last_vacuum, way to skip null (never been hit)

* Check at least one table has been vacuumed/analyzed since X time

* Check every table has been vacuum|analyze at least once. (no NULL)

* Check at least one table has been vacuumed|analyzed since X hours/days/...

* Check the current logfile is not bigger than X KB/MB/GB

* Make the bloat estimation more accurate (lots of small bad assumptions in there)

* Consider a more accurate bloat check, using the expensive pg_stattuple function

* For the Nagios 'perf' output on last_analyze and last_vacuum (and other things with a time), 
consider making a single global timing call.

* Create a Perl::Critic test suite