=== on 15 Oct 2014, 16:43:20 Ann Campbell wrote:
Assumptions

* issue will be logged on import statement of individual class
* cycles within a package will not be included in first implementation
* false negatives in initial implementation on cycles of more than 2 classes

Note, no overlap is intended with RSPEC-1196

=== on 15 Oct 2014, 22:12:40 Freddy Mallet wrote:
According to me this rule is valueless @Nicolas but I let you decide.

=== on 21 Oct 2014, 15:29:18 Nicolas Peru wrote:
I don't see the difference with RSPEC-1196 : if you have cycle between classes you will have a cycle between package. 

If both classes are at the same package level, then I don't see why you should not do something like that. This is valid code and I don't see any special issue that won't be covered by 1196.