For anyone interested, and particularly for Ernesto, I have successfully created a version of the InactivityTimeOutPanel that works even if the user changes to another browser tab. The key, which I became aware of from the jsfiddle links (see previous posts below is to use an absolute time for when a period of inactivity would result in a timeout (timeoutTime,set to now plus an interval,, say 5 minutes), and then when the timer fires, to compare the current time against the timeoutTime. This seems to avoid the problem of the timer slowing down when the user changes to another page. Reset just resets the timeoutTime to a now plus the interval. I have set the timer to go off in one quarter of the interval, thinking if there are some delays this will keep it closer to the mark.
I have tested this and it seems to work fine. I only had to change the JavaScript in Ernesto's example to get everything to work, and I have posted the code here: http://pastebin.com/TWwS5hhE for people to see. Ernesto, It would be great if you could put this code into the JavaScript in your example and try it out. Then we would have a complete example to point people to if it works OK. Regards, Bruce -----Original Message----- From: Ernesto Reinaldo Barreiro [mailto:reier...@gmail.com] Sent: Friday, May 16, 2014 2:42 PM To: users@wicket.apache.org Subject: Re: Ajax timer not counting when tab doesn't have focus? Bruce. Let me see if I can find some time during weekend to play a bit with the links your provided and come up with a more robust solution. On Fri, May 16, 2014 at 3:57 PM, Bruce Lombardi <brlom...@gmail.com> wrote: > First, thank you for your continuing interest. This is the best > support group experience I've ever had. I hope the message come in > the right sequence. Lke Ernesto and other A saw the sequencing problem > cause by the apache server problem. > > I ran the test that Martin suggestion with 6.16.0-SNAPSHOTand got the > same behavior, which is what Ernesto and I suspected would happen. > > I'm sorry, but I don't know a lot about JavaScript so I'm having with > certain things. I read this article in stackoverflow: > > > https://stackoverflow.com/questions/6032429/chrome-timeouts-interval-s > uspended-in-background-tabs > > Which lead to this: > > > http://stackoverflow.com/questions/5927284/how-can-i-make-setinterval- > also-work-when-a-tab-is-inactive-in-chrome > > > Which contained an example of a script that doesn’t work correctly (if > you go to it, you will see that it walks the word "test" across the screen. > If you go to another tab and come back "test" hasn’t moved) > > http://jsfiddle.net/7f6DX/3/ > > > and this example that works (does the same thing but you can see that > "test" moves even when you go to another tab - but don't go away too > long it will reach the end and disappear). > > http://jsfiddle.net/7f6DX/31/ > > Do you think the last (working) script could be adapted for our timeout? > > Regards, > Bruce > > > > -----Original Message----- > From: Ernesto Reinaldo Barreiro [mailto:reier...@gmail.com] > Sent: Tuesday, May 13, 2014 1:59 AM > To: users@wicket.apache.org > Subject: Re: Ajax timer not counting when tab doesn't have focus? > > Martin, > > Thanks for pointing out this. The exact JavaScript enabling this > feature can be found here > > > https://github.com/reiern70/antilia-bits/blob/master/client-sign-out-p > arent/client-sign-out/src/main/java/com/antilia/signout/InactivitySign > OutPanel.js > > Maybe there is a memory leak in here as well... but I do not see the > relation with WICKET-5570 as I do not use any of the AJAX timer > behavior, see > > > https://github.com/reiern70/antilia-bits/blob/master/client-sign-out-p > arent/client-sign-out/src/main/java/com/antilia/signout/InactivitySign > OutPanel.java > > As Bruce describes the problem it sounds like if by switching to other > tab browser give less "priority" to the execution of JavaScript in > inactive tabs. > > > > On Mon, May 12, 2014 at 9:28 PM, Martin Grigorov <mgrigo...@apache.org > >wrote: > > > It could be that https://issues.apache.org/jira/browse/WICKET-5570 > > is the issue. > > Please try with 6.16.0-SNAPSHOT if you can. > > > > Martin Grigorov > > Wicket Training and Consulting > > > > > > On Mon, May 12, 2014 at 8:02 PM, Bruce Lombardi <brlom...@gmail.com> > > wrote: > > > > > Thanks Ernesto. > > > > > > Yes, I do mean another browser tab. I have some new information. > > > The > > timer > > > seems to slow way down, but not stop counting down when the user > > > switches focus to another tab, either opening a new tab, or just > > > clicking on > > another > > > tab. This is happening in both Chrome and Firefox. > > > > > > I thought that my code was never timing out, but further > > > experimentation with a shorter timeout period shows that it is > > > doing the same thing as > > the > > > demo. That is, timing out in about 10 second when you stay on the > > > tab, > > but > > > taking 3 minutes or longer to timeout when if you change focus to > > > a different tab. > > > > > > I would appreciate it if you could look at your demo and see the > > > behavior and then see if you have any ideas. > > > > > > As for my code, I really feel they difference with your code are > > > insignificant especially now that I know the behavior is the same. > > > > > > See below for the differences. > > > > > > Thanks again, > > > > > > Bruce > > > > > > > > > Code. > > > > > > I have your InactivitySignOutPanel.java, > > > InactivitySignOutPanel.html and InactivitySignOutPanel.js from the > > > GitHub site with one slight > > modification > > > in InactivitySignOutPanel.java (see below). > > > > > > I setup my panel in a base class that serves as a superclass for > > > all > > the > > > Web pages in the application (extends WebPage) I have a method > > > that is called to setup the timeout. Here is the method. > > > > > > public void setupTimeout() { > > > > > > // Change the following value to change the > > > session timeout. > > > int numberOfMinutesUntilTimeout = 15; > > > > > > int secondsPerMinute = 60; > > > int secondsToTimeout =numberOfMinutesUntilTimeout > > > * secondsPerMinute; > > > > > > add(new InactivitySignOutPanel("timeout", > > > Duration.seconds(secondsToTimeout)) { > > > > > > private static final long serialVersionUID = 1L; > > > > > > protected String getOnTimeChangedFunction() { > > > return > > > "function(timeRemaining){if(!window.countDown) { window.countDown > > > = $('#countDown'); } $(window.countDown).html(timeRemaining);}"; > > > } > > > }); > > > > > > add( new AjaxLink<Void>("link") { > > > > > > private static final long serialVersionUID > > > = 1L; > > > > > > @Override > > > public void onClick(AjaxRequestTarget target) { > > > // do nothing > > > } > > > }); > > > } > > > > > > Here is the core code in InactivitySignOutPanel I have just > > > modified the respond method to call > > > > > > > > > ((WebDocsSession)WebSession.get()).processLogoutorSessionExpiratio > > > n( > > > ); > > > > > > > > setResponsePage(WebApplication.get().getApplicationSettings().getPag > > eE > > xpiredErrorPage()); > > > > > > so that the application does the right thing. Note: a break point > > > shows that this method is never call when the user is on another tab. > > > > > > > > > public InactivitySignOutPanel(String id, Duration duration) { > > > this(id, Model.of(duration)); > > > } > > > > > > public InactivitySignOutPanel(String id, IModel<Duration> model) { > > > super(id, model); > > > setRenderBodyOnly(true); > > > add(ajaxBehavior = new AbstractDefaultAjaxBehavior() { > > > > > > private static final long serialVersionUID > > > = 1L; > > > > > > @Override > > > protected void respond(AjaxRequestTarget target) { > > > beforeSignOut(target); > > > > > > ((WebDocsSession)WebSession.get()).processLogoutOrSessionExpiratio > > > n( > > > ); > > > > > > > > setResponsePage(WebApplication.get().getApplicationSettings().getPag > > eE > > xpiredErrorPage()); > > > } > > > > > > @Override > > > protected void onBind() { > > > // panel will not be repainted, no need to markup > > > markup id. > > > } > > > > > > @Override > > > public void renderHead(Component component, > > > IHeaderResponse > > > response) { > > > super.renderHead(component, response); > > > response.render(JavaScriptHeaderItem.forReference(JS)); > > > } > > > }); > > > } > > > > > > -----Original Message----- > > > From: Ernesto Reinaldo Barreiro [mailto:reier...@gmail.com] > > > Sent: Monday, May 12, 2014 11:39 AM > > > To: users@wicket.apache.org > > > Subject: Re: Ajax timer not counting when tab doesn't have focus? > > > > > > Hi Bruce, > > > > > > > > > On Mon, May 12, 2014 at 4:52 PM, Bruce Lombardi > > > <brlom...@gmail.com> > > > wrote: > > > > > > > This message is related to an original thread "Link to Wicket > > > > page from Javascript" but the focus is changed so much that I > > > > though a new title would be better. In the earlier thread I > > > > asked about linking to a Wicket page, but the real need was for > > > > help in implementing a session timeout that would send the user > > > > to a timeout page after a period of inactivity and would reset > > > > whenever a user clicked any button on the page. > > > > > > > > > > > > > > > > Ernesto Reinaldo Barreiro sent me links(see below) to a working > > > > demo that did exactly what I required. I used that code and > > > > everything seemed to work perfectly. However, I have since > > > > discovered that my timer does not seem to function correctly if > > > > the user goes to another tab. It seems like the timer stops > > > > timing down or slows way down until the user comes back to the original > > > > tab. > > > > > > > > Do you mean another browser tab? > > > > > > > > > > > > > > > I checked Ernesto's demo. If I go to the demo page and stay > > > > there it times out after roughly the defined 10 second timeout > > > > period, but if I go to the demo page, then click on another tab, > > > > it eventually times out, but after more than 3 minutes. My code, > > > > however, never seems to seem to timeout the page if I go to > > > > another > tab. > > > > > > > > Does this happens in a all browsers? Or in particular one? > > > > > > Mind that this is just a quick "something" I build to illustrate > > > the idea... So, it might not be bullet proof. > > > > > > > > > > > > > > I really need this to work to fulfill a requirement and it is > > > > preventing the launch of our software. It all works fine except > > > > if the user goes to a different tab. > > > > > > > > > > > > > > > > Ernesto, or anyone, do you know why this is happening. Is the > > > > ajax timer stopping when the tab is not selected? Why is the > > > > demo eventually timing out while application is not (can't see > > > > any real difference in the code)? > > > > > > > > > > Without seeing your code or a quick-start app illustrating the > > > problem > > it > > > might be difficult to point out what's going on. > > > > > > > > > > > Is there any way to get a better handle on this (e.g., firebug)? > > > > If the ajax timer is stopping, could something be done, perhaps > > > > using absolute elapsed time, to expire the session immediately > > > > if the user goes back to the tab? > > > > The latter is not ideal but might be acceptable. > > > > > > > > > > > > > > > > (Note: My session timeout in web.xml is 30 minutes, client > > > > timeout is > > > > 15 minutes, but for testing I set it to 10 seconds). > > > > > > > > > > > > > > > > > > > > > > > > // Links > > > > > > > > > > > > > > > > Demo of Example > > > > > > > > > > > > > > > > http://107.170.4.97:9080/client-sign-out-demo/?0 > > > > > > > > > > > > > > > > > > > > > > > > Code that sets up example > > > > > > > > > > > > > > > > https://github.com/reiern70/antilia-bits/blob/master/client-sign > > > > -o > > > > ut-p > > > > aren > > > > > > > > t/client-sign-out-demo/src/main/java/com/antilia/HomePage.java > > > > > > > > > > > > > > > > > > > > > > > > Code that provides the implementation (We have this code in > > > > package > > > > > > > > com.xyzco.web.docs.presentation.helpers.timeout) > > > > > > > > > > > > > > > > > > > > > > > > https://github.com/reiern70/antilia-bits/tree/master/client-sign > > > > -o > > > > ut-p > > > > aren > > > > > > > > t/client-sign-out-demo/src/main/java/com/antilia > > > > > > > > > > > > > > > > Thanks, > > > > > > > > Bruce > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > -----Original Message----- > > > > > > > > From: Bruce Lombardi [mailto:brlom...@gmail.com] > > > > > > > > Sent: Monday, March 17, 2014 8:24 AM > > > > > > > > To: users@wicket.apache.org > > > > > > > > Subject: Re: Link to Wicket page from Javascript > > > > > > > > > > > > > > > > Thanks, this is very helpful. > > > > > > > > > > > > > > > > Bruce > > > > > > > > > > > > > > > > Sent from my iPad > > > > > > > > > > > > > > > > > On Mar 17, 2014, at 4:52 AM, Ernesto Reinaldo Barreiro > > > > > > > > <reier...@gmail.com> wrote: > > > > > > > > > > > > > > > > > > You can see it at work here > > > > > > > > > > > > > > > > > > http://107.170.4.97:9080/client-sign-out-demo/ > > > > > > > > > > > > > > > > > > > > > > > > > > > On Mon, Mar 17, 2014 at 9:18 AM, Ernesto Reinaldo Barreiro < > > > > > > > > > reier...@gmail.com> wrote: > > > > > > > > > > > > > > > > > >> Martin, > > > > > > > > >> > > > > > > > > >> What I mean is something that triggers the sign out before > > > > >> serve side > > > > > > > > >> session expires... and trigger this expire event. > > > > > > > > >> > > > > > > > > >> > > > > > > > > >> https://github.com/reiern70/antilia-bits/blob/master/client-s > > > > >> ig > > > > >> n-ou > > > > >> t- > > > > > > > > >> parent/client-sign-out/src/main/java/com/antilia/signout/Inac > > > > >> ti > > > > >> vity > > > > >> Si > > > > > > > > >> gnOutPanel.java > > > > > > > > >> > > > > > > > > >> https://github.com/reiern70/antilia-bits/blob/master/client-s > > > > >> ig > > > > >> n-ou > > > > >> t- > > > > > > > > >> parent/client-sign-out/src/main/java/com/antilia/signout/Inac > > > > >> ti > > > > >> vity > > > > >> Si > > > > > > > > >> gnOutPanel.js > > > > > > > > >> > > > > > > > > >> The above seem to work. You use it as in > > > > > > > > >> > > > > > > > > >> > > > > > > > > >> https://github.com/reiern70/antilia-bits/blob/master/client-s > > > > >> ig > > > > >> n-ou > > > > >> t- > > > > > > > > >> parent/client-sign-out-demo/src/main/java/com/antilia/HomePage. > > > > >> java > > > > > > > > >> > > > > > > > > >> > > > > > > > > >> > > > > > > > > >> > > > > > > > > >> On Mon, Mar 17, 2014 at 9:11 AM, Martin Grigorov > > > > > > > > < <mailto:mgrigo...@apache.org> mgrigo...@apache.org>wrote: > > > > > > > > >> > > > > > > > > >>> Hi, > > > > > > > > >>> > > > > > > > > >>> On Fri, Mar 14, 2014 at 9:25 PM, Ernesto Reinaldo Barreiro < > > > > > > > > >>> reier...@gmail.com> wrote: > > > > > > > > >>> > > > > > > > > >>>> Hi, > > > > > > > > >>>> > > > > > > > > >>>> I would implement this as.. > > > > > > > > >>>> > > > > > > > > >>>> 1- A JS function that counts how much time has passed since > > > > >>>> last > > > > > > > > >>>> AJAX request. Once a certain threshold is passed an AJAX > > > > >>>> request is > > > > > > > > >>>> sent to > > > > > > > > >>> the > > > > > > > > >>>> page. > > > > > > > > >>>> 2- An AbstractAjaxBehavior that is added to the page. I > > > > >>>> would use > > > > > > > > >>>> it as context for AJAX request on 1). > > > > > > > > >>> > > > > > > > > >>> at that time the session should be already expired. > > > > > > > > >>> > > > > > > > > >>> But this can be used to do what is needed here - Wicket will > > > > > > > > >>> redirect to the configured PageExpiredErrorPage. > > > > > > > > >>> > > > > > > > > >>> > > > > > > > > >>>> 3- On the respond of AJAXBehavior you redirect to > > > > > > > > >>>> SessionTimeoutPage > > > > > > > > >>> and > > > > > > > > >>>> invalidate the session. > > > > > > > > >>>> > > > > > > > > >>>> > > > > > > > > >>>> > > > > > > > > >>>> On Fri, Mar 14, 2014 at 8:07 PM, Bruce Lombardi > > > > > > > > >>>> <brlom...@gmail.com> > > > > > > > > >>>> wrote: > > > > > > > > >>>> > > > > > > > > >>>>> Hi, > > > > > > > > >>>>> > > > > > > > > >>>>> > > > > > > > > >>>>> > > > > > > > > >>>>> An earlier question I asked about session timeouts was > > > > >>>>> answered by > > > > > > > > >>> Martin > > > > > > > > >>>>> Grigorov. I am implementing his suggestion and have a > > > > >>>>> JQuery > > > > > > > > >>>>> script > > > > > > > > >>> what > > > > > > > > >>>>> sets a timer that times out after a certain period of > inactivity. > > > > > > > > >>>>> The timeout triggers a function call. Currently, for a > > > > >>>>> quick > > > > > > > > >>>>> test, I > > > > > > > > >>> have > > > > > > > > >>>>> that function use location.href to go to an arbitrary > > > > >>>>> static web > > > > > > > > >>>>> page > > > > > > > > >>>> like > > > > > > > > >>>>> this: > > > > > > > > >>>>> > > > > > > > > >>>>> > > > > > > > > >>>>> > > > > > > > > >>>>> function whenUserIdle(){ > > > > > > > > >>>>> > > > > > > > > >>>>> location.href = > > > > > > > > >>>>> 'http://www.natural-environment.com'; > > > > > > > > >>>>> > > > > > > > > >>>>> } > > > > > > > > >>>>> > > > > > > > > >>>>> > > > > > > > > >>>>> > > > > > > > > >>>>> That works fine, but now I need to make it go to a Wicket > > > > >>>>> page in > > > > > > > > >>>>> my application, similar to what would happen if I had an > > > > >>>>> anchor > > > > > > > > >>>>> tag > > > > > > > > like: > > > > > > > > >>>>> > > > > > > > > >>>>> > > > > > > > > >>>>> > > > > > > > > >>>>> <a wicket:id=" SessionTimeoutPage" > > > > >>>>> href="#">SessionTimeoutPage</a> > > > > > > > > >>>>> > > > > > > > > >>>>> > > > > > > > > >>>>> > > > > > > > > >>>>> But I can't figure out how to do this in the javascript > function. > > > > > > > > >>>>> Can anyone tell me how to call the Wicket page from the > > > > >>>>> javascript > > > > > > > > >>>>> function? > > > > > > > > >>> > > > > > > > > >>> You can use <script wicket:id="expirationUrl"></script> and > > > > >>> a > > > > > > > > >>> Label("expirationUrl", "window.expirePageUrl ='" + > > > > > > > > >>> urlFor(SessionTimeoutPage.class) + "';"); Then use it with: > > > > > > > > >>> location.href = window.expirePageUrl'; > > > > > > > > >>> > > > > > > > > >>> > > > > > > > > >>>>> > > > > > > > > >>>>> > > > > > > > > >>>>> > > > > > > > > >>>>> Thanks, > > > > > > > > >>>>> > > > > > > > > >>>>> > > > > > > > > >>>>> > > > > > > > > >>>>> Bruce > > > > > > > > >>>> > > > > > > > > >>>> > > > > > > > > >>>> -- > > > > > > > > >>>> Regards - Ernesto Reinaldo Barreiro > > > > > > > > >> > > > > > > > > >> > > > > > > > > >> > > > > > > > > >> -- > > > > > > > > >> Regards - Ernesto Reinaldo Barreiro > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > -- > > > > > > > > > Regards - Ernesto Reinaldo Barreiro > > > > > > > > > > > > > > > > > > > > > > > > > -- > > > Regards - Ernesto Reinaldo Barreiro > > > > > > > > > ------------------------------------------------------------------ > > > -- > > > - To unsubscribe, e-mail: users-unsubscr...@wicket.apache.org > > > For additional commands, e-mail: users-h...@wicket.apache.org > > > > > > > > > > > > -- > Regards - Ernesto Reinaldo Barreiro > > > --------------------------------------------------------------------- > To unsubscribe, e-mail: users-unsubscr...@wicket.apache.org > For additional commands, e-mail: users-h...@wicket.apache.org > > -- Regards - Ernesto Reinaldo Barreiro --------------------------------------------------------------------- To unsubscribe, e-mail: users-unsubscr...@wicket.apache.org For additional commands, e-mail: users-h...@wicket.apache.org
