Re: EJB vs Servlets

[Kyle Cordes]

Title: RE: EJB vs Servlets

I agree that the standardness is a good argument in favor of CMP beans (as oppossed to another OR mapping tools); I was just making the point that there are plenty of mapping tools and they work fine with servlets, they worked fine before servlets... I think original post below would be more correct to say that development is faster with an OR mapping tool than with JDBC, and EJB CMP beans was their chosen OR mapping tools.   What I want to avoid is the idea that somehow OR mapping is something "new" that is only available via EJB CMP beans.           ----- Original Message ----- From: Magnus Rydin To: Orion-Interest Sent: Wednesday, October 11, 2000 6:07 AM Subject: RE: EJB vs Servlets I would say that the keyword here, and in nearly everything else is *standard*. There are *loads* of books, tutorials, documentation, news-groups, people that uses EJB and likes it. When you are using other, less widely spread, object-wrapper products, you can only rely on a small community for help. With standards comes tools, and we want tools, dont we? WR -----Original Message----- From: Kyle Cordes [mailto:[EMAIL PROTECTED]] Sent: den 10 oktober 2000 20:51 To: Orion-Interest Subject: Re: EJB vs Servlets This strikes me as a straw-man argument.  There is no reason that servlet code must use JDBC directly.  There are many object-wrapper products available that work similarly to CMP beans; such products predate EJB by a long, long time.   -Kyle Cordes     ----- Original Message ----- From: Mike Cannon-Brookes To: Orion-Interest Sent: Monday, October 09, 2000 6:08 PM Subject: RE: EJB vs Servlets I use EJBs in a high volume environment and have had no problems with scalability or speed yet.   I have to say once you know EJBs well enough, dev't is definitely faster than with servlets. The sheer volume of JDBC code and debugging required in a servlet outweighs the quick speed you can do the same thing in EJBs. (See ejb-maker for an example).   Mike