Potential problem with modelling of epo:Period

[csnyulas]

In ePO 4.0.0 the epo:Period class is modelled like this:

Questions:

1. What is the meaning of the epo:hasTimePeriod property, and why is it mandatory (cardinality 1..*) ? Looking at the value set at-voc:timeperiod this does not seem to make too much sense.
2. How can one express a period with a known starting date(time) and a duration of 2 months? This representation is often used in the eForms notices (see fields BT-36-Lot which often is used in combination with field BT-536-Lot).

Wouldn't it make more sense to make epo:hasBeginning attribute mandatory (unless we want to allow the flexibility of expressing periods for which we know the end instead of the beginning) and either point the epo:hasTimePeriod or a new property (say epo:hasDuration) to the epo:SpecificDuration class? The property pointing to the epo:SpecificDuration class would have of course cardinality 0..1.

[andreea-pasare]

The at-voc:timeperiod should be removed since it is unnecessary to have it linked to epo:Period. This will be analyzed and implemented in ePO v5.0.0.

Proposal to align with https://www.w3.org/2006/time#ProperInterval