oooookkk.... If you say so. It is a really simple change. It has been committed and pushed.
On Wed, Dec 16, 2015 at 2:04 PM, Kent Sølvsten <[email protected]> wrote: > Agree they should be unique > > But you could consider replacing the existing instead. > > That way it opens the possibility in assembly to reorder concerns > declared with annotations - or to insert a new concern between existing > (change A,B to A,C,B) > > /Kent > > Den 16-12-2015 kl. 06:56 skrev Niclas Hedhman: > > It seems that in previous versions of Qi4j/Zest, it may have been > possible > > to declare the same Concern (and SideEffect) multiple times and it would > > have been invoked multiple times. > > > > This was detected in TransientAsClassTest, where if I added a Concern, > then > > it was found both as a declaration on the Composite as well as on the > Mixin. > > > > I have add a check that if the Concern/SideEffect is already declared, it > > won't be declared again. Equality is defined by ConcernModel, which is > just > > the class name of the Concern. > > > > Question; Is this the correct semantic behavior? It will be the "first" > > Concern that remains, i.e. the "outer most" one. > > > > Cheer > > -- Niclas Hedhman, Software Developer http://zest.apache.org - New Energy for Java
