TODO

Most issues and todos are tracked on the issue tracker at
http://github.com/abh/ntppool/issues

multi server:

  - move "score_ts" into a server-per-monitor table
  - calculate score per monitor + aggregate into the total "score_raw", too.
  - http api to get IPs to check
  - api (anyevent daemon?) to send back monitor results
  - server calculates scores

multi server:

  - monitor script
    - use Getopt::Long
    - return score results every N servers instead of just at the end

  - monitor api
    - handle multiple monitor clients with the same API key

Not necessarily in order of priority:

  * Improve the scoring mechanism to better track long term stability
    (right now the score is basically based on your last couple of days)

  * Let people mark their server as "available to other pool members".
    Use the long term scores to include / recommend / ... servers from
    that list to new admins who want to join the pool.

  * Deploy more Soekris/ALIX boxes for monitoring stations and "super nodes"

allow an organization name and http link for each user for attribution
  - schema change
  - UI to update it on the manage page
  - Logic on the zone pages to show the organization url relative to
    the current total score for the user or some such

check if the servers do HTTP redirects 
  - check for request to pool.ntp.org and foo.pool.ntp.org
  - ok if redirect to www.pool.ntp.org
  - ok if none
  - warning if redirect somewhere else

update use.html page (be more explicit about number of zones)

make the country zones use data from the continents if they are otherwise too small

Make the monitoring servers use ntpd (noselect) to test the servers
  - figure out if this is even possible/practical for 5-10k servers
  - figure out the magic ntpdc commands to add and remove noselect servers
    (is that even possible?)