tomcat-users mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "CPC Livelink Admin" <cpclv...@fitzpatrick.cc>
Subject RE: non-cookie session tracking?
Date Mon, 13 Nov 2000 19:42:56 GMT
RE: non-cookie session tracking?
Well, for the first one, they don't have a session yet - so the JavaScript
just needs to be smart enough to behave when it's not there.  As sson as
they hit a dynamic page, they will get a session, and then the 'static'
pages can use it.
  -----Original Message-----
  From: Burgess, Jay [mailto:jburgess@digarch.com]
  Sent: Monday, November 13, 2000 02:18 PM
  To: 'tomcat-user@jakarta.apache.org'
  Subject: RE: non-cookie session tracking?


  While our servlets use cookies for session handling, we use this technique
all the time with parameters other than session ID.  I thought it might be
applicable to your problem.

  However, now I'm wondering about the "bootstrap" case i.e. how do you get
the session ID onto the URL for the very FIRST page?  Maybe there's more to
this than my original email implied.

  Jay



  -----Original Message-----
  From: Daniel A. Theobald [mailto:theo@vecnatech.com]
  Sent: Monday, November 13, 2000 12:14 PM
  To: tomcat-user@jakarta.apache.org
  Subject: Re: non-cookie session tracking?



  This approach sounds promising, with much less of a performance hit than
  doing it dynamically on the server side.  Has anyone actually tried this
  approach before?

  The way the site works right now, any page that leaves the site is
  brought up in a separate window anyway.  So by the same token we could
  have any url that needs to maintain session use the addSessionID
  approach.

  This project has many restrictions:
  No cookies,
  No frames,
  Applets only when explicitly authorized.

  All which ultimately hurt performance.  One question I have is if every
  single page in the entire site is generated from our main.jsp page, does
  it do any good to integrate with Apache?  My understanding is that
  Apache would only serve the static pages requested directly.  Is that
  correct?

  theo







  > "Burgess, Jay" wrote:
  >
  > Another option, assuming you know ahead of time which URLs on a page
  > should have the session ID propagated, would be to use Javascript to
  > grab the session ID off of the current page's URL, add it to the new
  > page's URL, and then ask for the new page.
  >
  > For example, each URL on your static HTML page can be changed as
  > follows:
  >
  > (old)
  > <a href="myPage.html">Go to myPage</a>
  >
  > (new)
  > <a href="javascript:addSessionID('myPage.html')">Go to myPage</a>
  >
  > The addSessionID() function can grab the sessionID value from the
  > current URL, append it to the page name passed into the function, and
  > then set "location" to the new URL.
  >
  > Jay


Mime
View raw message