On Sat, Feb 7, 2009 at 7:31 AM, tao wen <[email protected]> wrote: > Hi, > > In the book, I see different bounded context requires a translation map or > anti-corruption layer. So I guess that means two bounded context will not > share same structure of domain model, otherwise why they need a translator?
Well, if you read up on Conformist and Shared Kernel, it could be that translation is not needed, or needed to a much smaller extent. Since the book, he has also introduced "Partnership" where two domains collaborate, again with or without translation. > In qi4j, we also use the word context. But in the implementation, it is a > composite extending several interfaces. Interface is a abstraction of role > played in a context or collaboration. So, in qi4j, there are no translation > between context, they all share same state with same structure. Therefore, I > think the "context" are different, in Eric's book and in qi4j > implementation. Yes, perhaps. Perhaps it is so insane that Qi4j is actually trying to make the GUM a possibility, and that that might be a pipe-dream. Cheers Niclas -- http://www.qi4j.org - New Energy for Java _______________________________________________ qi4j-dev mailing list [email protected] http://lists.ops4j.org/mailman/listinfo/qi4j-dev
