Hi,
Sorry it took some time to respond - I sat down and wrote a somewhat simple
test program that I can redistribute if required.
ok here is my first pass. i cannot reproduce the problem, so i am kinda
flying blind here - all i know for sure is that it compiles
it seems to me that as soon as you change the page the browser dumps the old
javascript, in my case looks like it also dumps any running ajax requests. i
tried both ie and ffox.
i added a self upating counter that keeps updating every 1 sec and a link to
go to another page, havent been able to catch the right timing i guess
I did some more testing also. As I mentioned in my first e-mail, our setup
is *similar* to self updating timer. The difference to timer is that we have an
applet receiving events from server. The applet then executes javascript
function to fire ajax update when needed.
It looks like that the threads in applet continue to run for some time even
after link is clicked (ie. applet is not stopped immediately)
and seem to be able to run javascript on page also.
At least using an applet here is the easiest way to reproduce the situation.
(And I understand that we are bending wicket a lot here :) - but compared
to what we had earlier wicket+ajax is very elegant and simple)
anyways, i am attaching the patch, let me know if it works for you and we
can take it from here
The patch mostly works; however, there seems to be a small window where
two threads (one from click and other one from ajax) can be executing at
DefaultRequestTargetResolverStrategy which still results in "page expired".
My test application is available at:
http://download.syncrontech.com/public/wicket-ajax-race.zip
How to test:
- load ExpireTestPage1
- click on "page 2 link"
- click on "to page 1". Without your Igor's patch, you'll get "page expired"
almost always
With Igor's patch, it requires some rather mad and rapid clicking of links to
get
the requests generated so that error still occurs. But it is still there.
Ari S.
----- Original Message -----
On 5/23/06, Igor Vaynberg <[EMAIL PROTECTED]> wrote:
i got up extra early to take care of this, so by the end of the day
-igor
On 5/23/06, Ari Suutari <[EMAIL PROTECTED]> wrote:
>
> This is so great !
>
> TIA,
>
> Ari S.
>
> ----- Original Message -----
> From: "Johan Compagner" <[EMAIL PROTECTED]>
> To: < wicket-user@lists.sourceforge.net>
> Sent: Tuesday, May 23, 2006 11:37 AM
> Subject: Re: [Wicket-user] pagemap problem with late arriving ajax
> request
>
>
> We will fix it in the 1.2 branch.
>
> I am currently a bit full of work for at least 2 days. So i can only
> look at
> it by the end of this week.
>
> johan
>
>
> On 5/23/06, Ari Suutari <[EMAIL PROTECTED]> wrote:
> >
> > Hi,
> >
> > >best thing todo currently is to have that -1 check in the pagemap and
>
> > make
> > >your own version of wicket 1.2
> >
> > We already decided to put my modified version to production and
> > if there are no side effects from my change we can keep it for
> some
> > time.
> >
> > >in trunk (2.0) this is already fixed in the default setting because
> we
> > >refactored the pagemap
> > >and have now a different default pagemap and httpsessionstore.
> >
> > You mean that if I would use 2.0 from trunk the problem would go
> > away ? If so, would it be possible to back-port the new pagemap
> > to 1.2 branch ? Not as default, but something that I could turn on
>
> > from
> > application settings ?
> >
> > Another point: when 1.2 is released, someone else might also
> > hit this problem so it might be necessary to implement some kind
> > of fix anyway (at least for us, some pages were almost unusable
> > because of this as other pages worked quite ok - it seems that
> YMMV).
> >
> > Ari S.
> >
> > On 5/23/06, Ari Suutari <[EMAIL PROTECTED] > wrote:
> > >
> > > Hi,
> > >
> > > It is very important for us to get this fixed, preferably for 1.2.
> > > If any wicket committer is willing to help us, we will pay
> > > for the work (please contact me directly by e-mail for this).
> > >
> > > Ari S.
> > >
> > > ----- Original Message -----
> > > From: "Igor Vaynberg" <[EMAIL PROTECTED]>
> > > To: < wicket-user@lists.sourceforge.net>
> > > Sent: Monday, May 22, 2006 6:21 PM
> > > Subject: Re: [Wicket-user] pagemap problem with late arriving ajax
> > request
> > >
> > >
> > > problem is, not every ajax behavior uses version -1. i guess we need
> to
> > > have
> > > the ability to tag a request as process-only-if-page-is-active
> > >
> > > -Igor
> > >
> > >
> > > On 5/22/06, Ari Suutari <[EMAIL PROTECTED] > wrote:
> > > >
> > > > If I modify the PageMap.java like this:
> > > >
> > > > final Page get(final int id, int versionNumber)
> > > > {
> > > > final IPageMapEntry entry =
> > > > (IPageMapEntry)session.getAttribute(attributeForId(id));
> > > > if (entry != null)
> > > > {
> > > > if (versionNumber == -1) {
> > > > Access a = peekAccess();
> > > > if (a.getId() != id)
> > > > return null;
> > > > }
> > > >
> > > > So this kludgery piece of the code checks that if request is
> coming
> > from
> > > > ajax
> > > > (ie. versionNumber == -1) and the topmost page is not the one
> being
> > > > requested
> > > > -> return null.
> > > >
> > > > After adding this it looks like I cannot reproduce the problem any
> > more.
> > > >
> > > > Ari S.
> > > >
> > > > ----- Original Message -----
> > > > From: "Johan Compagner" <[EMAIL PROTECTED]>
> > > > To: < wicket-user@lists.sourceforge.net>
> > > > Sent: Monday, May 22, 2006 2:37 PM
> > > > Subject: Re: [Wicket-user] pagemap problem with late arriving ajax
> > > request
> > > >
> > > >
> > > > So what request comes first?
> > > > The ajax request should always do -1 (so latest release)
> > > > So if the request from the new page did come first then the ajax
> > request
> > > > shouild work on that new page.
> > > > If the ajax request did come a bit earlier then there shouldn't be
> a
> > > > problem
> > > > at all because
> > > > the last reqest will be the normal link click for a new page.
> > > >
> > > > johan
> > > >
> > > >
> > > > On 5/22/06, Ari Suutari <[EMAIL PROTECTED]> wrote:
> > > > >
> > > > > I cannot figure out any workaround myself.
> > UnversionedBehaviorListener
> > > > > seems to use -1 as
> > > > > page version number, would it make any sense to try to alter
> PageMap
> > > so
> > > > > that
> > > > > if version is -1 the page stack is not cleared ?
> > > > >
> > > > > Ari S.
> > > > >
> > > > > ----- Original Message -----
> > > > > From: "Matej Knopp" <[EMAIL PROTECTED]>
> > > > > To: <wicket-user@lists.sourceforge.net >
> > > > > Sent: Monday, May 22, 2006 11:44 AM
> > > > > Subject: Re: [Wicket-user] pagemap problem with late arriving
> ajax
> > > > request
> > > > >
> > > > >
> > > > > >I guess this is because in 1.2 pagemap behaves like stack.
> Johan?
> > > > > >
> > > > > > -Matej
> > > > > >
> > > > > > Ari Suutari wrote:
> > > > > >> Hi,
> > > > > >>
> > > > > >> I'm having a problem with wicket's ajax stuff where an ajax
> > request
> > > > > >> fired by previous page hits the web server after user has
> clicked
> > a
> > > > > link
> > > > > >> which has already transferred him to next page.
> > > > > >>
> > > > > >> This might be a problem in my application, but I'm seeking
> for
> > > ideas
> > > > > >> to solve this.
> > > > > >>
> > > > > >>> From web server's access log:
> > > > > >>
> > > > > >> First there are some normal ajax requests arriving from
> browser:
> > > > > >>
> > > > > >> 192.168.5.102 - - [22/May/2006:07:46:43 +0000] "GET
> > > > > >>
> > > > >
> > > >
> > >
> >
>
/wicket?wicket:interface=:0:uiPanel:viewPanel:normalView:panels:0:panel:listTableSpan:-1:IUnversionedBehaviorListener&wicket:behaviorId=0&random=
> > > >
> > > > > 0.35930677427195573
> > > > > >> HTTP/1.1" 200 824
> > > > > >> 192.168.5.102 - - [22/May/2006:07:46:43 +0000] "GET
> > > > > >>
> > > > >
> > > >
> > >
> >
>
/wicket?wicket:interface=:0:uiPanel:viewPanel:normalView:panels:0:panel:listTableSpan:-1:IUnversionedBehaviorListener&wicket:behaviorId=0&random=
> > > >
> > > > > 0.11935459751117722
> > > > > >> HTTP/1.1" 200 824
> > > > > >>
> > > > > >> Here, user clicks a link which transfers application to help
> > page,
> > > it
> > > > > is displayed ok.
> > > > > >>
> > > > > >> 192.168.5.102 - - [22/May/2006:07:46:43 +0000] "GET
> > > > >
> > /wicket?wicket:interface=:0:uiPanel:topPanel:showHelp:34:ILinkListener
> > > > > >> HTTP/1.1" 302 -
> > > > > >> 192.168.5.102 - - [22/May/2006:07:46:43 +0000] "GET
> > > > > /wicket?wicket:interface=:34:: HTTP/1.1" 200 3191
> > > > > >>
> > > > > >> Now, for some reason a ajax request arrives from previous
> page,
> > > which
> > > >
> > > > > causes to pagemap stack to be
> > > > > >> popped so the page user's browser is on is removed from page
> map.
> > > > > >>
> > > > > >> 192.168.5.102 - - [22/May/2006:07:46:43 +0000] "GET
> > > > > >>
> > > > >
> > > >
> > >
> >
>
/wicket?wicket:interface=:0:uiPanel:viewPanel:normalView:panels:0:panel:listTableSpan:-1:IUnversionedBehaviorListener&wicket:behaviorId=0&random=
>
> > > > > 0.2873453536331084
> > > > > >> HTTP/1.1" 200 824
> > > > > >>
> > > > > >> Now, user presses a link on the page but gets 'page expired'
> > > message,
> > > > > because the page
> > > > > >> is no longer in pagemap.
> > > > > >>
> > > > > >> 192.168.5.102 - - [22/May/2006:07:46:44 +0000] "GET
> > > > > /wicket?wicket:interface=:34:back::ILinkListener HTTP/1.1" 404
> 651
> > > > > >>
> > > > > >> Any ideas what I could do to fix this ? The ajax requests are
>
> > fired
> > > > > asynchronously by a system which is very
> > > > > >> similar to AjaxSelfUpdatingTimerBehaviour stuff.
> > > > > >>
> > > > > >> Ari S.
> > > > > >>
> > > > > >>
> > > > > >>
> > > > > >> -------------------------------------------------------
> > > > > >> Using Tomcat but need to do more? Need to support web
> services,
> > > > > security?
> > > > > >> Get stuff done quickly with pre-integrated technology to make
> > your
> > > > job
> > > > > easier
> > > > > >> Download IBM WebSphere Application Server v.1.0.1 based on
> Apache
> > > > > Geronimo
> > > > > >>
> > > > >
> > > http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642
>
> > > >
> > > > > >> _______________________________________________
> > > > > >> Wicket-user mailing list
> > > > > >> Wicket-user@lists.sourceforge.net
> > > > > >> https://lists.sourceforge.net/lists/listinfo/wicket-user
> > > > > >>
> > > > > >
> > > > > >
> > > > > >
> > > > > > -------------------------------------------------------
> > > > > > Using Tomcat but need to do more? Need to support web
> services,
> > > > > security?
> > > > > > Get stuff done quickly with pre-integrated technology to make
> your
> > > job
> > > > > easier
> > > > > > Download IBM WebSphere Application Server v.1.0.1 based on
> Apache
> > > > > Geronimo
> > > > > >
> > > http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642
>
> > > >
> > > > > > _______________________________________________
> > > > > > Wicket-user mailing list
> > > > > > Wicket-user@lists.sourceforge.net
> > > > > > https://lists.sourceforge.net/lists/listinfo/wicket-user
> > > > > >
> > > > >
> > > > >
> > > > >
> > > > > -------------------------------------------------------
> > > > > Using Tomcat but need to do more? Need to support web services,
> > > > security?
> > > > > Get stuff done quickly with pre-integrated technology to make
> your
> > job
> > > > > easier
> > > > > Download IBM WebSphere Application Server v.1.0.1 based on
> Apache
> > > > Geronimo
> > > > >
> > > http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642
>
> > > > > _______________________________________________
> > > > > Wicket-user mailing list
> > > > > Wicket-user@lists.sourceforge.net
> > > > > https://lists.sourceforge.net/lists/listinfo/wicket-user
> > > > >
> > > >
> > > >
> > > >
> > > > -------------------------------------------------------
> > > > Using Tomcat but need to do more? Need to support web services,
> > > security?
> > > > Get stuff done quickly with pre-integrated technology to make your
> job
> > > > easier
> > > > Download IBM WebSphere Application Server v.1.0.1 based on Apache
> > > Geronimo
> > > >
> > > >
> >
> http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642
> > > > _______________________________________________
> > > > Wicket-user mailing list
> > > > Wicket-user@lists.sourceforge.net
> > > > https://lists.sourceforge.net/lists/listinfo/wicket-user
> > > >
> > >
> > >
> > >
> > > -------------------------------------------------------
> > > Using Tomcat but need to do more? Need to support web services,
> > security?
> > > Get stuff done quickly with pre-integrated technology to make your
> job
> > > easier
> > > Download IBM WebSphere Application Server v.1.0.1 based on Apache
> > Geronimo
> > >
> http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642
> > > _______________________________________________
> > > Wicket-user mailing list
> > > Wicket-user@lists.sourceforge.net
> > > https://lists.sourceforge.net/lists/listinfo/wicket-user
> > >
> >
> >
> >
> > -------------------------------------------------------
> > Using Tomcat but need to do more? Need to support web services,
> security?
> > Get stuff done quickly with pre-integrated technology to make your job
> > easier
> > Download IBM WebSphere Application Server v.1.0.1 based on Apache
> Geronimo
> >
> http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642
> > _______________________________________________
> > Wicket-user mailing list
> > Wicket-user@lists.sourceforge.net
> > https://lists.sourceforge.net/lists/listinfo/wicket-user
> >
>
>
>
> -------------------------------------------------------
> Using Tomcat but need to do more? Need to support web services,
> security?
> Get stuff done quickly with pre-integrated technology to make your job
> easier
> Download IBM WebSphere Application Server v.1.0.1 based on Apache
> Geronimo
> http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642
>
> _______________________________________________
> Wicket-user mailing list
> Wicket-user@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/wicket-user
>
-------------------------------------------------------
All the advantages of Linux Managed Hosting--Without the Cost and Risk!
Fully trained technicians. The highest number of Red Hat certifications in
the hosting industry. Fanatical Support. Click to learn more
http://sel.as-us.falkag.net/sel?cmd=lnk&kid=107521&bid=248729&dat=121642
_______________________________________________
Wicket-user mailing list
Wicket-user@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/wicket-user
