Issue solution not tested: 8

- updates in comments
- Location of topology file fixed for network examples

- Fixed energy calculation: if fromUtilization == 0 and toUtilization == 0, it's assumed that the host is switched off and therefore energy is 0.

- Fixed host.isSuitableForVM - only accepts a VM, if the maximum requested MIPS among all virtual PEs < PE capacity

- Removed PowerPe and Power.PeList

- Fixed MIPS allocation to PEs for time-shared VM scheduling: MIPS of single VM can be split among several PEs with the only constraint that MIPS of the VM < MIPS of a single PE, as VMs cannot be parallelized

- New VmScheduler: time-shared scheduling supporting over-subscription

processing was never updated, because it was always rounded to 0.

when it should return estimated finish time.

Added a Maven pom that will be included in the distribution package along with the Ant's build file.

Reverted revision 41 - it caused another problem. At this moment there is a small bug: sometimes not all cloudlets return to the broker after the execution.

Fixed a bug in the PowerDatacenter::updateCloudletProcessing() method that caused removal of completed VMs before the Cloudlets have been sent to the broker. This didn't affect the simulation results themselves, just the output.

Datacenter: send ack for cloudlet submission when requested
FederatedDatacenter: send ack for cloudlet submission when requested; 
DatacenterBroker.java: avoid utilization of uninitialized VMs; removal of schedulingInterval field; schedule of internal events.

- TODO replace returning the field from Vm.getCurrentAllocatedMips() by a call to getCloudletScheduler().getCurrentMipsShare()