I'm all for adding additional junits, even if they duplicate scenarios covered by the TCK (as long as we're not cut-n-pasting code from the TCK into our svn.) Thought being, that we would be allowing contributors who are generating patches to possibly catch some regressions, without expecting everyone to always setup and run the TCK before checking in changes.
-Donald On 4/21/10 11:57 AM, Carlos Vara wrote: > Hi all! > > Currently I am submitting patches to try to pass more TCK tests, and I'm not > sure if I should enclose unit tests along with the patches. > > Basically, if tests are added, they will probably be a replication of the > tests already available in the TCK, and having duplicated code isn't good. > On the other hand, maybe you value having more "own" test coverage and you > don't want or prefer not to rely on the TCK to ensure proper behavior. > > Please, let me know what you prefer, as by my side it is no problem at all > to add unit tests. > > Regards, > Carlos >
