If the path is different then they are completely separate cookies. So you could have a cookie named "fruit" and a path of "/root" with a value of "apple". You could have another cookie named "fruit" but with a path of "/root/anotherpath" and it has a value of "orange". If you don't set a path for the cookie it gets set for you so to stop this from happening I always set the path explicitly.
Regards, Randy Burgess Sr. Software Architect D5 Systems, LLC > From: Pierre Thibaudeau <[EMAIL PROTECTED]> > Reply-To: Struts Users Mailing List <user@struts.apache.org> > Date: Sat, 9 Aug 2008 00:16:11 -0400 > To: Struts Users Mailing List <user@struts.apache.org> > Subject: [S2] general question about cookies and namespace > > Struts: 2.1.2 > browser: Firefox 3.0, with Firebug add-on > > Strange phenomenon involving cookies. I don't know who to blame: my code, > Struts or Firefox... > > I am running a copy of my application locally on my own local Tomcat server > (on localhost). After several operations on the webapplication (among which > would be logging in), four cookies are set in the browser: > * JSESSIONID (dealt with by tomcat; not my doing, of course) > * language (the user's last selection of language) > * password > * username > > However, when I study the net requests through firebug, I notice that > Firefox includes only the first of these cookies in the following requests' > header. Accordingly, if I start a new session, the language of the previous > session is read as it should from the "language" cookie, but the auto-login > interceptor fails to find the "username" and "password" cookies (even though > they are present to the browser). > > On the other hand, things work fine on Safari. > > This seems strange to me. The only formal distinction I can find between, > say, the "language" cookie and the "username" cookie is the "path": > * Firefox report the path of the "language" cookie as: /mywebapp/ > * but it reports the path of the "username" cookie as: /mywebapp/users/ > > Is it possible that Firefox is overstrict with which cookie to provide back > to a given action, and the different namespace of the "username" cookie > becomes problematic? The actions that register my auto-login interceptor > are all in the "/" namespace (that is, the same namespace as reported by the > "language" cookie). Has anyone come accross anything like this? --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
