In order to minimise eventual changes of supervisord configuration, and so
avoid needless restarts, sort the list of services. They are read from
filesystem, and os.listdir does not guarantee same list on each invocation.

Sometimes one incident (network is down on some location, a computer pbs is down ) and multiple Support Requests are generated due this problem.

The Incident response is used to Group all support requests related to an unresolvable incident, so we can track (or ignore) the support requests waiting for an incident be revolved.

This is helpfull for development

/reviewed-on !117

render function should take 0 time otherwise, it may be called several times by
parent function.  In this javascript, the render function is actually
requesting a software release, so we don't want it to be called several
times...

/reviewed-on nexedi/slapos.core!118

Sometimes history link (on header) contains a link to its parent when the link comes from some external source.

  Avoid to excessively post to master if the state is the same as the once expected,
  this decreases significantly the number of posts to master.

Mostly make it more reliable against slowliness and particular corner cases (slow indexation, data formating)

/reviewed-on nexedi/slapos.core!115

Stop date is used to determinate until when the Acknowledgement will be displayed.

The order of execution of activities are not linear anymore, so this require ensure objects are indexed to be found by the alarm.

Monitoring and Upgrade scope of a Slave handled by the owner of the Software Instance where the Slave is allocated into.

This will require to the user input title, else the form will indicate that information is missing.

Reorder the field to fit better the screen.

/reviewed-on !114