Reviews:VojtechSvatek about PeriodicInterval

VojtechSvatek about PeriodicInterval (Revision ID: 11172)

Overall suggestion (score): 0 - needs major revision

Review Summary: The problem is important, but needs much more careful and systematic treatment, see below.

Reviewer Confidence: High in ODPs in general; not particularly deep in temporal modelling, yet I believe the pattern is unsophisticated enough for me to confidently evaluate it.

Problems: It is necessary to distinguish between 'gap' (the distance between an end and a start, for subintervals) and 'period' (normally, the distance between *two* starts or *two* ends).

In fact, the given pattern could be generalized to that of a periodic interval with regular starts and variable length - e.g. an executive meeting that always starts on Wednesday 1pm but takes as much time as needed. One could even imagine the opposite case, e.g. preparation of coffee for the meeting can start (variably) in advance but always ends when the meeting starts - for it to be drunk suitably warm :-)

In the pattern description it is not even clear if the 'start' and 'end' refer to start/end of the whole non-convex interval (start: Tuesday, x-th January, 2010, 0:00) or of the subinterval (start: Tuesday, 0:00; with year 2010 as externally defined context).

Community Relevance: Highly relevant. Temporal modelling is one of the hottest issues.

Relation to Best Practices: The alternatives should be explicitly described.

Reusability: Highly reusable (after thorough revision...).

Relations to Other Patterns: Missing, although there are other temporal patterns on the portal.

Overall Understandability: Partial, see the Problems section.

Clear Problem Description: Yes

Clear Relevance and Consequences: Not quite clear consequences at the moment; alternative solutions should be explored.

Clear Figures and Illustrations: OK

Missing Information: See above.