If the storage does not search into document list by itself,
JIO will limit it anyway.
The storage should use the method researchDone() to tell to JIO that
the document list is already limited (avoid double limiting).
If the storage can not support limiting, it can also use
searchDocumentArray() provided by JIO (and no need to use researchDone()).

If the storage does not limit document list by itself,
JIO will limit it anyway.
The storage should use the method limitDone() to tell to JIO that
the document list is already limited (avoid double limiting).
If the storage can not support limiting, it can also use
limitDocumentArray() provided by JIO (and no need to use limitDone()).

If the storage does not sort document list by itself,
JIO will sort it anyway.
The storage should use the method sortDone() to tell to JIO that
the document list is already sorted (avoid double sorting).
If the storage can not support sorting, it can also use
sortDocumentArray() provided by JIO (and no need to use sortDone()).

It contains now
- status : code
- statusText : string associated with error code
- message : an explicit error message
- array : array containing some returned errors from child job(s).

ReplicateStorage will now return the first returned result. (but not for
checkNameAvailability). So OfficeJS changed a little too.

This will set the maximum duration (in ms) a job may wait for.

Node name is now written before its test.
file 'OfficeJS/grunt/gruntall'