Martijn Faasen makes the point that it is important for the Zope 3 developers to have good use cases. So here we go. Here is my distributed Object use case for Zope 3 developers. Can I do this in Zope 3? I run a job board for each technology. I have one job board for zope/plone developers, and another one for Linux System Administrators. Right now I have quite low traffic, and I run all my job boards off of one zope instance. But if this should ever be a hit, I can imagine running one zope instance per job board, or even multiple zope instances for one job board using zeo on its own server. The available parallelism is quite strong in Zope 2. Now here is the problem. Say one of you is both a Linux Admin and a Zope/Plone developer. Right now you would have to post your resume twice on two different job boards. If you changed your resume, you would have to change it in two different servers. Sometime soon, I hope that you will be able to just apply once. You would fill out the Zope skills question box, and the Linux Administrator skills question box, and the server will automagically create copies of your record on both job boards. If those are on separate servers, then that is a distributed migration of objects. When you return to edit your data and update your resume, that information needs to be propogated to the other job board on the other server as well. I hope that is a helpful use case for the Zope 3 developers. Of course maybe things are done totally differently in Zope 3 than in Zope 2. Here is another distributed use case, which I do admit is more speculative. Performance reasons might encourage me to have one server in the US, and one in India. Reportedly Google localization takes into account where the server is located. So if I want to compete globally in the US and India, I would want my Indian candidate objects to migrate to my Indian server, and my US candidate objects to migrate to my US server. I think I could do this with ZClasses. Can I do it with Zope 3? I hope that helps. Chris