re: sendRedirect(), also jikes

[Bart Locanthi <ba...@sabl.com>]

nfi on sendRedirect(url) vs setHeader("location", url). i'm trying to
decide whether to try 3.2b2 or not. i'm using a build as of a couple
weeks ago.

btw i finally discovered why i couldn't get it to use jikes.

1) conf/web.xml is never read. ever.
2) webapps/<context>/WEB-INF/web.xml has the goods
	if you make a copy from conf/web.xml, you also need to fix the
	specification of jsp servlet's full name, eg
		org.apache.jasper.servlet.JspServlet
3) specifying a compilerPath, as below, also breaks the use of jikes
        <init-param>
            <param-name>compilerPath</param-name>
            <param-value>/usr/local/bin/jikes</param-value>
        </init-param>
4) in light of (1), from *where* does tomcat get default values?
5) the first time a .jsp is compiled, jikes or no, it takes a very long
time
	independent of whether the jit is enabled or not. why?

> Date: Fri, 14 Jul 2000 18:14:58 +0200
> To: tomcat-user@jakarta.apache.org
> From: GOMEZ Henri <hg...@slib.fr>
> Subject: RE: sendRedirect()
> Message-ID: <36...@css4.cs>
> 
> We got something like this when using request dispatcher.
> include (lDispatcher.include( pReq, pResp )) didn't send the JSESSION
> COOKIE.=20
> We have to use instead forward (lDispatcher.forward( pReq, pResp )).
> 
> Any idea why ? Or did it is a bug corrected in 3.2 ?
> 
-----
> 
> > i had an issue with sendRedirect() that i traced to it not sending the
> > session cookie with the redirection.
> > 
> > when i replaced sendRedirect() by setHeader("location", url), it worked.