> From: "Turner, John" <[EMAIL PROTECTED]> > Sent: Thursday, January 30, 2003 1:33 PM > Subject: RE: Tomcat default directory for servlets
> Actually, the examples use the Invoker servlet, which is not the recommended > configuration. Granted, but it's also a fine line between trying to get marginal success for the user quickly and then moving them forward to a best practice. Having to jump through the hoops to get a simple HelloWorld servlet working is like having to write a Makefile first to compile hello.c. One of the complaints about J2EE is simply that it's pretty darn wordy to get all of this stuff set up. Once you hurdle the burden of all of the boiler plate, then development can be fairly straightforoward. But having to create the web.xml, the servlet.java, the servlet.class, AND put them in the Right Places (i.e. a WAR structure) is quite a bit to do to simply get Hello World. Now, in time, of course, all of this minutae needs to be understood and appreciated. How many questions are answered here that are pretty black and white in the servlet spec, for example. IMHO, the book should have provided their own skeletal webapp that has all of the boiler plate written for the user in order to get their test servlet(s) up and running quickly, and also to confirm the users Tomcat installation. The other benefit, of course, is that the webapp path is more portable than simply the Tomcat extensions path. The challenge is to get good, solid, and quick, success for the user and then migrate them smoothly from the hand holding that the original skeleton gave them to creating, managing and taking responsibility for these attributes on their own. But book authors are as lazy as the rest of us, so far be it for me to cast the first stone... Regards, Will Hartung ([EMAIL PROTECTED]) --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
