Re: Page instantiated twice

Wed, 21 Oct 2009 12:26:34 -0700

Any comments about this? could It be a bug? If there's no other way of achieving the multiwindow feature that doesn't include instanciating the page again, maybe something could be done to avoid that the first instance consumes the session messages, because as we see this first page instance isn't the one that will be seen in the browser. 

what do you think?

Esteban


Esteban Masoero escribió:
That is correct. I'm openning the page in a new tab just because of the process registration I'm doing: 
- user registers
- app sends him an email with a generated link
- user clicks that link from his email page, which it's oppened in a new window/tab (there's no way It can be openned in its previous app page, where he recently registered)
And as I said in another email, I've found out that setting getPageSettings().setAutomaticMultiWindowSupport(false); solves the problem. However, I'm lossing the multi-window feature and I'm not sure I want that.
Isn't a way that the new page can use a copy of the session used by the old page? (although I guess a copy of the session should be made before its feeedback messages are consumed) Because as I see, there's no way that I can leave a feedback message to the session and so that message can be shown in a new window, because of this feature. 

Thanks,

Esteban



Igor Vaynberg escribió:

are you opening the page in a new tab? this script is part of wicket
auto-multi-window support, it detects if the page has been opened in a
new tab and redirects to a url that will point to the same page but in
a new pagemap.

-igor

On Tue, Oct 20, 2009 at 4:23 PM, Esteban Masoero
<emaso...@getsense.com.ar> wrote:

First of all, thanks to all of you for your help.
I haven't solved the problem, but I'm getting closer to it.
(something like: new PageLink(new MyPage()); <= I checked it and I'm not 
doing that. )

Here's what I've found out:
After debugging a lot, I find out that TWO REQUESTS are made. The first one 
as a consecuence of the user asking for that page, and the second one
because in that page the following script is added at the end of <head> tag: 
<script type="text/javascript" ><!--/*--><![CDATA[/*><!--*/
if (window.name=='' || (window.name.indexOf('wicket') > -1 &&
window.name!='wicket-wicket:default')) {
window.location="?wicket:bookmarkablePage=wicket-6:ar.com.getsense.latamvalley.pages.LoginPage" 
+ (window.location.hash != null ? window.location.hash : ""); }
/*-->]]>*/</script>

Can anyone tell me why is Wicket doing that? or what am I doing wrong?
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||| 
Here's the complete detailed description about what I am doing:
A UserStatusPanel that is always shown does *new Link("register")* 
with:
       @Override
               public void onClick() {
                       throw new
RestartResponseAtInterceptPageException(UserRegistrationPage.class);
               }
When the user gets to the page, submmits its registration data and 
the page sends an email with a link like:
http://localhost:8080/ua?ua=2baab43f784b5b530b5347a50490bb0a


       In my webApp, I've done:
       this.mount(new QueryStringUrlCodingStrategy("/ua",
UserActivationPage.class));
When the user goes to the previous mentioned link by a new tab, the 
following happens:

       _______________________
       GET /ua?ua=2e129db15b6d6db5342ba5d328642262 HTTP/1.1
       Host: localhost:8080
       ....
       Connection: keep-alive
       Cookie: JSESSIONID=78ehtk0o3y0w
       _______________________________
=== IN org.apache.wicket.RequestCycle.processEventsAndRespond() === 
       --processor.processEvents(this);
---- UserActivationPage is constructed, and the page activates the 
user
       ---- UserActivationPage does: setResponsePage(LoginPage.class);
       --processor.respond(this);
       ----LoginPage is constructed (messages exists in session and are
properly rendered to somewhere)
=================================================================== [I can see in firebug that the correct response reaches the browser!] 

       Strangely, In the page, wicket adds:        <script
type="text/javascript" ><!--/*--><![CDATA[/*><!--*/
if (window.name=='' || (window.name.indexOf('wicket') > -1 &&
window.name!='wicket-wicket:default')) {
window.location="?wicket:bookmarkablePage=wicket-6:ar.com.getsense.latamvalley.pages.LoginPage" 
+ (window.location.hash != null ? window.location.hash : ""); }
/*-->]]>*/</script>

       [So a new request is made!]
       ___________________________________________
       GET
/ua?wicket:bookmarkablePage=wicket-6:ar.com.getsense.latamvalley.pages.LoginPage 
HTTP/1.1
       Host: localhost:8080
       ...
       Connection: keep-alive
Referer: http://localhost:8080/ua?ua=2e129db15b6d6db5342ba5d328642262 
       Cookie: JSESSIONID=78ehtk0o3y0w
       ________________________________________________
=== IN org.apache.wicket.RequestCycle.processEventsAndRespond() === 
       --processor.processEvents(this);
---- LoginPage is constructed (not UserActivationPage like it did 
before) (this time, messages from the session are gonne)
       --processor.respond(this); finally this page is shown without my
messages :(
=================================================================== 

Thanks in advance,


Esteban






Martin Grigorov escribió:

El mar, 20-10-2009 a las 15:04 -0700, Douglas Ferguson escribió:


You can also check your html to make sure there aren't any urls that
 would get loaded by the browser.

i.e. <link>, <img>, etc tags..


yeah, <img src="" ... /> will make a second request to the page. quite
nasty!
fortunately there is a checker for these kind of problems in
wicket-devutils project


D/


On Oct 20, 2009, at 4:39 PM, Jeremy Thomerson wrote:



Have you followed the suggestions in this thread?
First, put a breakpoint to see what's instantiating the page - the most 
likely suspect for an unintended additional instantiation is a bad
 link,
something like: new PageLink(new MyPage());   <--- shouldn't do that
If it's Wicket code, what code is it? Why? It might be stateless as 
Martijn suggested.

--
Jeremy Thomerson
http://www.wickettraining.com



On Tue, Oct 20, 2009 at 4:23 PM, Esteban Ignacio Masoero <
emaso...@getsense.com.ar> wrote:



Hi there. I'm bringing to life this old thread to ask whether this
 issue
was
solved and how, because I'm experiencing something alike, which makes 
session messsages to get lost ("consummed" by the first instance of
 the
page), and consecuently not shown by the second page instance.

Thanks,

Esteban



Vinayak Borkar wrote:


Martijn,

Ok. I set the getStatelessHint() to return false.

Now I get java.lang.IllegalStateException: No SessionHandler or
SessionManager
What is the reason that wicket needs to instantiate the page twice? I 
can understand the second instantiation, but my guess is that the
 first
instantiation is not to render the HTML.
If I am correct about my analysis above, is there some way I can know that the instantiation does not need to populate all the components? 

Thanks,

Vinayak

Martijn Dashorst wrote:
Your page is stateless, which is a holy grail for most to attend. Be 
happy :)

Since it is stateless, Wicket has to construct the page with each
request, until it is no longer stateless.
If you have a form, override its getstatelesshint method and return 
false.

Martijn
On Tue, Mar 31, 2009 at 8:36 AM, Vinayak Borkar <vbo...@yahoo.com> 


wrote:


Jeremy,

I did that. The first time it is

at
org.apache.wicket.request.target.component.BookmarkableListenerInterfaceRequestTarget.processEvents 
(BookmarkableListenerInterfaceRequestTarget.java:127)


and the second time it is

at
org.apache.wicket.request.target.component.BookmarkablePageRequestTarget.processEvents 
(BookmarkablePageRequestTarget.java:223)


Does that help?

Thanks,
Vinayak


Jeremy Thomerson wrote:
Run it in debug mode, put a breakpoint in your constructor and see 
what's
instantiating each.  A lot of times it is your own code (a bad
 link,
etc).

--
Jeremy Thomerson
http://www.wickettraining.com
On Tue, Mar 31, 2009 at 1:30 AM, Vinayak Borkar <vbo...@yahoo.com> 
wrote:



Hello,


I have a form that performs a search and shows a page with
 results.
The
result page also has the same form to do a repeat search. I
 noticed
that
the
SearchPage is instantiated twice when I do a search from the form 
on
the
result page. Is there a way to make sure that does not happen?
 The
search
page ends up doing the search twice.

Thanks,
Vinayak
--------------------------------------------------------------------- 
To unsubscribe, e-mail: users-unsubscr...@wicket.apache.org
For additional commands, e-mail: users-h...@wicket.apache.org
--------------------------------------------------------------------- 
To unsubscribe, e-mail: users-unsubscr...@wicket.apache.org
For additional commands, e-mail: users-h...@wicket.apache.org
--------------------------------------------------------------------- 
To unsubscribe, e-mail: users-unsubscr...@wicket.apache.org
For additional commands, e-mail: users-h...@wicket.apache.org





--
View this message in context:
http://www.nabble.com/Page-instantiated-twice- tp22799162p25982806.html 
Sent from the Wicket - User mailing list archive at Nabble.com.
--------------------------------------------------------------------- 
To unsubscribe, e-mail: users-unsubscr...@wicket.apache.org
For additional commands, e-mail: users-h...@wicket.apache.org




---------------------------------------------------------------------
To unsubscribe, e-mail: users-unsubscr...@wicket.apache.org
For additional commands, e-mail: users-h...@wicket.apache.org




---------------------------------------------------------------------
To unsubscribe, e-mail: users-unsubscr...@wicket.apache.org
For additional commands, e-mail: users-h...@wicket.apache.org




---------------------------------------------------------------------
To unsubscribe, e-mail: users-unsubscr...@wicket.apache.org
For additional commands, e-mail: users-h...@wicket.apache.org




---------------------------------------------------------------------
To unsubscribe, e-mail: users-unsubscr...@wicket.apache.org
For additional commands, e-mail: users-h...@wicket.apache.org






---------------------------------------------------------------------
To unsubscribe, e-mail: users-unsubscr...@wicket.apache.org
For additional commands, e-mail: users-h...@wicket.apache.org

Reply via email to