Brien, I guess I'm not sure why you say both windows will have the same timestamp. The first time the user submits the form, the hidden field that would contain the timestamp id would be blank. It would then get set and the returned page would have the timestamp id in a hidden form field for use on subsequent form submit actions. If a new browser window was spawned, I don't see why it would copy the contents of the web page along with the values of the form fields. However if the user was set on creating such a scenario, and there was a way for them to do so, then they would end up sharing the same id reference. If the user likes the confused experience this provides him he will continue to do this. If he doesn't I suspect he will determine it isn't a good idea and avoid doing so. Beyond that there's not much else I can see you can reasonably be expected to do. Dan > ---------- > From: Brien Voorhees[SMTP:[EMAIL PROTECTED]] > Sent: Wednesday, November 17, 1999 12:02 PM > To: Kirkdorffer, Daniel; [EMAIL PROTECTED] > Subject: Re: Re: Tracking sessions in browser windows(also > nonrepeatable transactions) > > I'm a little confused as to how you detect when the user has spawned a new > window in the middle of using your web app. The new window will have the > same timestamp as the original one so both windows will be hitting the > same > "web application object", no? Or does your application assume the user > will > begin a new session (i.e. window) from the login page? > > By the way, the October issue of Dr. Dobb's Journal has an interesting > article on implementing reliable nonrepeatable transactions on websites > (i.e. preventing the user from hitting "submit" twice, or refreshing, > thereby causing duplicate orders). Essentially it entails embedding a > token in the initial page in the transaction sequence (e.g. "choose items > page", "order items page", "confirm order page") and incrementing it (and > checking it) in each subsequent page. > > Ensuring nonrepeatable transactions and identifying multiple app instances > should really be covered in every good Servlet book yet, strangely, it > isn't > covered in any of them. > > Thanks, > Brien Voorhees > > > ----- Original Message ----- > From: "Kirkdorffer, Daniel" <[EMAIL PROTECTED]> > To: <[EMAIL PROTECTED]> > Sent: Tuesday, November 16, 1999 5:56 PM > Subject: Re: Tracking sessions in browser windows > > [nice detailed explanation snipped] > > =========================================================================== To unsubscribe: mailto [EMAIL PROTECTED] with body: "signoff JSP-INTEREST". FAQs on JSP can be found at: http://java.sun.com/products/jsp/faq.html http://www.esperanto.org.nz/jsp/jspfaq.html
