Well, I'm not experienced with J2EE but here's my take anyway. Actually, I should say I'm parroting what I heard (though I try to be an informed parrot :-) earlier this week. I personally *do* like Java but, based on a EJB user group get-together earlier this week, I would shy away from using EJB at this time for a production site. This meeting was a first meeting, and it's quite possible that the group had no experts with all the right answers. However, the take was that EJB is not quite ready for prime time. It's the right direction, but there's a whole lot of undefined (spec) stuff and a lot more needed on implementation to get *all* the promised advantages. IOW, EJB doesn't buy everything development wise that it seems to on the surface. Basically, manage your expectations. My take away understanding is (at this current time), EJB offers transaction management transparency. That's the big win. Other stuff ends up being too brittle to be useful on a dynamic web site. Web site development using EJB seemed to be an afterthought. The picture isn't great yet. In particular, the consensus by the end seemed to be to avoid using Entity Beans for now (those are the persistent ones) and just use Session Beans. Instead of Entity Beans, just put a layer of custom database access objects to do the database stuff between the session beans and the DB. If Entity Beans become desirable later on, just upgrade the DAO layer to use entity beans instead of accessing the DB directly. As an aside, if you want pure power from a commercial platform, you might want to take a look at Apple's Web Objects. I've only heard good things about it (scalability, tools, etc.) though I know it's proprietary. FYI, Apple is rolling out Java2 (finally) with their new (mach/unix) OS X. (Also, FYI, Java2 won't make it to OS9 since all Java development by Apple is now for OS X.) Zope figures into this equation for sheer flexibility. I would think Web Objects would work for pure performance (scalability), documentation, and support (though $). Before I get pummeled for my (perhaps incorrect) opinion on performance, let me just say that WO has the advantage (performance-wise) of Objective-C underneath. O-C is compiled. OS X, btw, will be Java on top with O-C working underneath as necessary (O-C is somewhat more flexible than Java type-wise, so it's necessary some places for WO even if Java is generally preferred). That's my capsule perspective. Good luck, = Joe = Wei Tao wrote:

Nemeth Miklos wrote:

...

In our department I have decided (as a CTO) to use Zope for an ecommerce application server. And I also would like to use it for another application for a customer. I will have a presentation about Zope and our plans in a 3-day technical conference of our corporation.

Can I have some (primarily technical) arguments why Zope is better to some (especially Java-based) specific web application server tool (WebSphere, Oracle Application Server, Oracle8i, NetDynamics, etc)?

I'd be very glad if some of you share his thoughts about the topic.

I found a report titled by "The Technical Benefits of EJB and J2EE technologies over COM+ and Windows DNA" several days ago: http://web2.java.sun.com/products/ejb/pdf/j2ee_dnatwp.pdf

( And there is a sister report titled by "The Business Benefits of EJB and J2EE technologies over COM+ and Windows DNA": http://web2.java.sun.com/products/ejb/pdf/j2ee_dnabwp.pdf)

It is written well. Though personally I dislike java, I have to say that J2EE is well cooked.

_______________________________________________ Zope maillist - Zope@zope.org http://lists.zope.org/mailman/listinfo/zope ** No cross posts or HTML encoding! ** (Related lists - http://lists.zope.org/mailman/listinfo/zope-announce http://lists.zope.org/mailman/listinfo/zope-dev )