I opened https://issues.apache.org/jira/browse/WICKET-506 on this, but I'm
curious to see if it's just me or if others have had similar problems, and
if so, if there is a known workaround.
Basically when using an AjaxFallbackLink to launch a ModalWindow, IE 6 and
7 both fail to launch the
One other thing from today :)
A few of my users have reported, and I've experienced once myself, getting
an Internal Error page while navigating the new site, but I don't get
anything in my containers stdout or stderr log files.
Is there something I need to do to configure the logging level
Hi Eelco
Thanks for the help.
I've a question however : currently, I put an object containing the
Set as a final attribute of my Panel class, in order to be able to
increase/decrease its values depending of the action in the AjaxLink
onClick.
However, having read about memory/session size, I
Hi
Something's not right with RadioGroup when using
wantOnSelectionChangedNotifications or I am doing something wrong.
When the same Radio are selected for the 2nd time, the radio becomes unselected
and cannot be selected again until another radio has been selected. However the
On 4/25/07, Igor Vaynberg [EMAIL PROTECTED] wrote:
the behavior works, i like it a lot better then a panel. to me:
TextField tf=new TextField(..);
tf.add(new TinyMceEnabler().setAllowLinking(false));
yes, it looks better this way, less boiler plate code.
seems a lot more natural then
On 4/26/07, Nick Johnson [EMAIL PROTECTED] wrote:
One other thing from today :)
A few of my users have reported, and I've experienced once myself, getting
an Internal Error page while navigating the new site, but I don't get
anything in my containers stdout or stderr log files.
Is there
Here is another example log4j properties file (log4j.properties).
Maybe it will help you. It logs everything into a file called
tomcat.log in Tomcat's logs directory. At runtime, log4j will create
tomcat.log if it doesn't exist. It logs Wicket code at an INFO level
and logs code from
Er, that should be as follows (logging Wicket at INFO level). Sorry,
I need to proof-read better next time... :)
#
# Application log level options
#
log4j.logger.org.apache.wicket=INFO
log4j.logger.org.springframework=INFO
log4j.logger.com.example=DEBUG
On Apr 26, 2007, at 1:47 AM, Justin
package com.xyz.xyzcom.wicket.pages.sections;
import com.xyz.xyzcom.wicket.pages.DefaultxyzPage;
public class DefaultHomePage extends DefaultxyzPage {
public static final String HOME = home;
public String getSection() {
return HOME;
}
public String
since when does a logger have state?
I am very interrested to know what it is that a logger isn't thread safe.
You configure it through the constructor (so i guess assign final fields)
you don't set any variables after that on it. And you call only info/error
with 1 or to params that it needs
But if static logs are really a problem:
http://wiki.apache.org/jakarta-commons/Logging/StaticLog
if would do this:
private void doSomething()
{
/// foo
LogFactory.getLog(Xxx.class).debug(blaat)
// bar
}
or
private void doSomething()
{
/// foo
getLog().debug(blaat);
// bar
}
On 4/25/07, Xavier Hanin [EMAIL PROTECTED] wrote:
On 4/25/07, Igor Vaynberg [EMAIL PROTECTED] wrote:
what do you mean by once authenticated
what does your loginpage code look like, eg how do you navigate back?
My login page is really very simple, it doesn't do much except include
a
Hi , I am wondering why the checkbox property model is not updated when the
method wantOnSelectionChangedNotifications is overriden and returns true.
When i call form submit, property is set ok.
DropDownChoice updates the model well on roudtrip , so I am surprise that
checkbox not. I had to
bump anyone for answer?
Fra: [EMAIL PROTECTED] på vegne af Nino Wael
Sendt: to 26-04-2007 09:25
Til: wicket-user@lists.sourceforge.net
Emne: [Wicket-user] RadioGroup Broken(1.2.5+?/when
usingwantOnSelectionChangedNotifications)?
Hi
Something's not right with
Yep, that was it. Here's the very primitive log4j.properties I'm
including now (src/main/resources for Maven 2), just in case this thread
ever turns up in a Google search or it helps somebody else:
log4j.rootLogger=WARN, stdout
log4j.appender.stdout=org.apache.log4j.ConsoleAppender
myapplication app=nnew myapplication();
wickettester tester=new wicketteseter(app);
-igor
On 4/26/07, Akshat [EMAIL PROTECTED] wrote:
package com.xyz.xyzcom.wicket.pages.sections;
import com.xyz.xyzcom.wicket.pages.DefaultxyzPage;
public class DefaultHomePage extends DefaultxyzPage {
hrm. it does seem rather strange. we have a similar authenetication stuff
going on in a few projects and it works.
but i wonder
is the page being intercepted a stateless page?
is the login page a stateless page?
-igor
On 4/26/07, Xavier Hanin [EMAIL PROTECTED] wrote:
On 4/25/07, Xavier
open a bug please
-igor
On 4/26/07, kubino [EMAIL PROTECTED] wrote:
Hi , I am wondering why the checkbox property model is not updated when
the
method wantOnSelectionChangedNotifications is overriden and returns true.
When i call form submit, property is set ok.
DropDownChoice updates the
This is going to paste in an ugly way I'm sure, but I'll try to tidy up
the spacing just a bit.
span id=fragment_forum_tree_entry_1_fragment_nav_replyOrNot
a
href=/tally-ho/article?wicket:interface=:5:12::#fragment_forum_tree_entry_1_fragment_nav_replyOrNot
onclick=var
On 4/26/07, Igor Vaynberg [EMAIL PROTECTED] wrote:
hrm. it does seem rather strange. we have a similar authenetication stuff
going on in a few projects and it works.
but i wonder
is the page being intercepted a stateless page?
is the login page a stateless page?
I don't think so, because I
you could just make sure that your deployed app has no bugs, that way there
is no need to bother setting up all this logging mambojumbo
-igor
On 4/26/07, Nick Johnson [EMAIL PROTECTED] wrote:
Yep, that was it. Here's the very primitive log4j.properties I'm
including now (src/main/resources
I'm sorry, I've just noticed that I have an exception[1] before
getting the Page Expired message.
The exception occurs during deserialization of the page from which I
come, and is related to a field injected with via Spring using the
wicket-spring-annot. Then this exception is catched and simply
ah well that explains it :)
1.3 is using our experimental serialization code to save the state which
results in a much smaller size then the default serialization. currently it
doesnt handle all cases, such as serializing proxies. that is why there are
instructions in that error message that
Not a single line of code has been changed between the two versions. The
version with Wicket 1.2.5 jars displays property web pages using css files
in the css subdirectory of the web application root directory, and the
version with Wicket 1.2.6 jars cannot find the css files that are in the
same
maybe it is because wicket now always adds a / to the end of the servlet
mapping. this was inconsistent in the past, sometimes urls were
/context/mapping?wicket:id... and sometimes they were
/context/mapping/?wicket:id=...
the problem was that /context/mapping?wicket:id urls did not always get
On 4/26/07, Igor Vaynberg [EMAIL PROTECTED] wrote:
ah well that explains it :)
1.3 is using our experimental serialization code to save the state which
results in a much smaller size then the default serialization. currently it
doesnt handle all cases, such as serializing proxies. that is why
servlet-mapping
servlet-namedmWicketApplication/servlet-name
url-pattern/app/*/url-pattern
/servlet-mapping
--
View this message in context:
http://www.nabble.com/from-1.2.5-to-1.2.6-tf3647980.html#a10203607
Sent from the Wicket - User mailing list
Yes, it seems that this is the problem:
http://localhost:8081/dmWicket/app?wicket:bookmarkablePage=:course.wicket.app.AboutPage
But I do not know how to correct it since my servlet mapping is:
url-pattern/app/*/url-pattern
igor.vaynberg wrote:
maybe it is because wicket now always adds a
well, at first i didnt want to believe we had a printstacktrace() somewhere.
but yeah i found it. i changed it to throw an rte, hopefully there was a
very good reason for swallowing it like that and someone will revert my
change with an explanation.
-igor
On 4/26/07, Xavier Hanin [EMAIL
Many thanks to Igor and Matej, the problem was found this morning.
It turns out that while Firefox and Opera let you get away with it, MSIE
will not allow you to use span as the placeholder for a ModalWindow. It
has to be div.
Also found an obscure bug in IE after that one got fixed. The
Hi Eelco
Thanks for the help.
I've a question however : currently, I put an object containing the
Set as a final attribute of my Panel class, in order to be able to
increase/decrease its values depending of the action in the AjaxLink
onClick.
However, having read about memory/session
Yes, that was my point earlier. Or make it a transient with lazy creation as
that would be more efficient.
Eelco
On 4/26/07, Johan Compagner [EMAIL PROTECTED] wrote:
But if static logs are really a problem:
http://wiki.apache.org/jakarta-commons/Logging/StaticLog
if would do this:
private
is that the 1.2.6 or the 1.2.5 url?
-igor
On 4/26/07, dzenanr [EMAIL PROTECTED] wrote:
Yes, it seems that this is the problem:
http://localhost:8081/dmWicket/app?wicket:bookmarkablePage=:course.wicket.app.AboutPage
But I do not know how to correct it since my servlet mapping is:
Sorry,
It is the 1.2.6 url and the url is:
http://localhost:8081/dmWicket/app/?wicket:bookmarkablePage=:course.wicket.app.AboutPage
When the page is displayed there is no CSS formatting for the page. In the
log there is a warning message:
the resource requested by request [method = GET,
try changing this:
link rel=stylesheet type=text/css media=screen href=css/box.css /
to
link rel=stylesheet type=text/css media=screen href=../css/box.css
/
notice the ../ in href. that should fix that.
also note that this isnt the greatest way to include css because it will
break if you use
I need to work with a servlet from another group and they need to see the user
as an attribute on the HttpSession.It seems as though setAttribute on my
wicket.Session doesn't actually set it on the HttpSession. Is that correct?
I've found that the following works, but it just doesn't seem
On my page, I use a DataTable with custom IColumn implementtions that use
fragments to make the cells editable. If I change that values in the cells
and click submit (which just submits the Form object) then it preoperly
persists my data and refreshes with the correct data when there are no
I need to work with a servlet from another group and they need to see the user
as an attribute on the HttpSession.It seems as though setAttribute on my
wicket.Session doesn't actually set it on the HttpSession. Is that correct?
Typicaly it does, but it depends on your configuration.
To be more precise:
It works in the home page when it is
link rel=stylesheet type=text/css media=screen href=css/app.css /
It does not work in the home page when it is
link rel=stylesheet type=text/css media=screen href=../css/app.css
/
It works in the about page (link from the home page)
your model reuse strategy must not be working correctly.
the input values are lost because they bound to the component, and if that
component is recreated by the repeater they are then gone. thus the
disappearing values.
-igor
On 4/26/07, ChuckDeal [EMAIL PROTECTED] wrote:
On my page, I
I have found out what the problem is:
In HomePage.html I use:
link rel=stylesheet type=text/css media=screen href=css/app.css /
style type=text/css media=screen
@import css/box.css;
@import css/cite.css;
@import css/code.css;
@import css/def.css;
link and import have not changed behavior
what has changed is that wicket now always appends a / after the servlet
mapping, where as before it only did so sometimes. so now it is more
consistent but is messing with your context path.
-igor
On 4/26/07, dzenanr [EMAIL PROTECTED] wrote:
I
Well, I changed imports to links and the same definition in both pages works
well without using ../:
link rel=stylesheet type=text/css media=screen href=css/app.css /
link rel=stylesheet type=text/css media=screen
href=css/box.css /
link rel=stylesheet type=text/css media=screen
but not for Wicket components, an instance log field in Component (or other
base wicket class)
will be a memory slot for every wicket component instance...
johan
On 4/26/07, Eelco Hillenius [EMAIL PROTECTED] wrote:
Yes, that was my point earlier. Or make it a transient with lazy creation
as
Yeah, I had thought that was the problem, that's I why I switched to
ReuseIfModelsEqualStrategy (actually, a custom impl based upon that class
because I could figure out how to use a custom model for the cells;
AbstractDataGridView.getPopulatorsIterator is private and it doesn't supply
an
yeah, so it either more efficient in terms of processing (reusing the member
var) or in memory (not keep the member var).
Eelco
On 4/26/07, Johan Compagner [EMAIL PROTECTED] wrote:
but not for Wicket components, an instance log field in Component (or
other base wicket class)
will be a memory
what does your output markup look like though? did wicket rewrite
css/app.css to /contextpath/css/app.css for you?
-igor
On 4/26/07, dzenanr [EMAIL PROTECTED] wrote:
Well, I changed imports to links and the same definition in both pages
works
well without using ../:
link rel=stylesheet
On Thursday 26 April 2007 1:02 pm, Eelco Hillenius escreveu:
I need to work with a servlet from another group and they need to see the
user as an attribute on the HttpSession.It seems as though
setAttribute on my wicket.Session doesn't actually set it on the
HttpSession. Is that
On 4/26/07, ChuckDeal [EMAIL PROTECTED] wrote:
Yeah, I had thought that was the problem, that's I why I switched to
ReuseIfModelsEqualStrategy (actually, a custom impl based upon that class
because I could figure out how to use a custom model for the cells;
why not just make them static?
you want the nonstatic loggers so you can get better context, but we dont
need better context! for us classname:linenumber is enough. so in the
framework we can use the static instances. if a user wants better context
then their loggers can be non-static.
-igor
On 4/26/07, Igor Vaynberg [EMAIL PROTECTED] wrote:
why not just make them static?
http://wiki.apache.org/jakarta-commons/Logging/StaticLog
-
This SF.net email is sponsored by DB2 Express
Download DB2 Express C - the FREE
link rel=stylesheet type=text/css href=/dmWicket/css/app.css
media=screen/
link rel=stylesheet type=text/css href=/dmWicket/css/box.css
media=screen/
The answer is yes.
igor.vaynberg wrote:
what does your output markup look like though? did wicket rewrite
css/app.css to
I've run into a problem with how wicket handles requests in
PageParameters. I'm seeing this particular behaviour with BookmarkablePage
urls that contain a string with non ascii, utf-8 characters, like this:
pages/J%C3%A1nos
This doesn't work. The following does:
pages/J%E1nos
First I
yeah yeah yeah.
and this is a problem only for things that are deployed into a SHARED
classpath.
no body (at least i hope not) is sharing wicket.jar between multiple web
applications. so there is no problem.
-igor
On 4/26/07, Johan Compagner [EMAIL PROTECTED] wrote:
On 4/26/07, Igor
so it was never the link element that was the problem, it was the @imports
because wicket doesnt process those.
-igor
On 4/26/07, dzenanr [EMAIL PROTECTED] wrote:
link rel=stylesheet type=text/css href=/dmWicket/css/app.css
media=screen/
link rel=stylesheet type=text/css
true
Is there a reason why Wicket does not process imports?
igor.vaynberg wrote:
so it was never the link element that was the problem, it was the @imports
because wicket doesnt process those.
-igor
On 4/26/07, dzenanr [EMAIL PROTECTED] wrote:
link rel=stylesheet type=text/css
why not just make them static?
Earlier this thread:
(me) I would prefer to either have log fields as static in components...
We were just discussing what people could do if they didn't want those
loggers as static variables. Personally, I have no problem with them being
static.
Eelco
what they would do is simple
they would do the lazylookup you showed, but instead of using a field they
would use the metadata store.
that way you get a non-static logger, and do not incur a memory footprint
hit.
-igor
On 4/26/07, Eelco Hillenius [EMAIL PROTECTED] wrote:
why not just make
Here is a quickstart[1] that attempts to illustrate this. Maybe you can tell
me where I went wrong. It doesn't use my classes per se, but it does
simulate the way in which I construct and use a DataTable.
http://www.nabble.com/file/8104/quickstart-datatable.zip
quickstart-datatable.zip
[1]
I am experimenting with Wicket to use it for my dissertation. I am
developing a basic book library application. I am creating an edit page
which edits boolen values of a borrower object (isInDebt, isAccountExpired,
etc.). I tried to map it with the following code, which is an inner-class:
static
Igor Vaynberg wrote:
the model you return from dataprovider has to have its equals and
hashcode implemented, if you want those can forward to the modelobject.
If you're using a DatabinderProvider, Chuck, then it's a
HibernateObjectModel; I'll add those overrides to it. Something like
this
what they would do is simple
they would do the lazylookup you showed, but instead of using a field they
would use the metadata store.
that way you get a non-static logger, and do not incur a memory footprint
hit.
The memory foot print would still be there if you use the Component's meta
not really. you would remove the logger in ondetach. it isnt a permanent
footprint - that is it doesnt take up a slot in all subclasses.
-igor
On 4/26/07, Eelco Hillenius [EMAIL PROTECTED] wrote:
what they would do is simple
they would do the lazylookup you showed, but instead of using a
* Janos Cserep:
I've run into a problem with how wicket handles requests in
PageParameters. I'm seeing this particular behaviour with BookmarkablePage
urls that contain a string with non ascii, utf-8 characters, like this:
Is it this one?
Parameters of nice URL's pages with 'sensitive'
on which class is it then bound?
we use private static loggers now so they are really bind to the class.
But if you are storing them in the meta data then they are not really bind
to the class but more the the resulting class/instance
so never Component itself. Don't know if that really matters
like i said in my previous response, ditto. for wicket private static is
fine.
-igor
On 4/26/07, Johan Compagner [EMAIL PROTECTED] wrote:
on which class is it then bound?
we use private static loggers now so they are really bind to the class.
But if you are storing them in the meta data then
Hi Erik
The translation is there :
http://zedros.developpez.com/articles/java/wicket/ajax/
The article will be checked by fellow members of developpez.com and
then announce on the news of this website. I'll then add it to the
wicket. In between, every comment is welcome ;) (or afterwards lol)
great, a neat way to make logging even cheaper...
instantiate all the time for each and every message :-)
Am 26.04.2007 um 13:28 schrieb Johan Compagner:
But if static logs are really a problem:
http://wiki.apache.org/jakarta-commons/Logging/StaticLog
if would do this:
private void
great, a neat way to make logging even cheaper...
instantiate all the time for each and every message :-)
It doesn't get instantiated every time. For instance Log4jLoggerFactory
creates one instance lazily and then does a map lookup. Of course, it is
still more expensive then just referring
update wicket and it will work
-igor
On 4/26/07, ChuckDeal [EMAIL PROTECTED] wrote:
Here is a quickstart[1] that attempts to illustrate this. Maybe you can
tell
me where I went wrong. It doesn't use my classes per se, but it does
simulate the way in which I construct and use a DataTable.
Hi everyone!
I'm having trouble with an AutoCompleteTextField. I want to add an ajax
behavior to it (specifically, for onblur) like can I do with other wicket
components, but somehow it doesn't work.
I can add an AjaxFormComponentUpdatingBehavior to a TextField and it works.
I cannot add it
Unless I am missing something, why not do it like this:
private transient Logger logger;
private Logger getLogger()
{
if (logger == null) {
logger = LoggerFactory.getLogger(Foo.class);
}
}
public void myMethod()
{
...
getLogger().debug(Something here);
}
On Thu,
because if we put that into Component for example, all components will have
a logger take up a memory slot
-igor
On 4/26/07, Philip A. Chapman [EMAIL PROTECTED] wrote:
Unless I am missing something, why not do it like this:
private transient Logger logger;
private Logger getLogger()
{
I am trying to create an edit form which models the hashset of roles in a
CheckBoxMultipleChoice; so many roles can be selected and displayed. How can
I achieve this? Here is the class that that I am modelling and the EditForm
page.
*Borrower.java**
package
Hello,
We've run into a small issue tare trying to position a ModalWindow on
the screen relative to the link that spawned it. Does anyone know of a
solution for this?
Thanks in advance
Reuben
-
This SF.net email is
Thanks a lot Steven. Forwarded this to Wicket users for anyone who is
interested in playing with it.
Eelco
On 4/26/07, Steven Harris [EMAIL PROTECTED] wrote:
Our first version of a wicket config module is now in trunk (will be in
tomorrow's nightly). we would
love feedback so if anyone gets
It makes sense to me to do this. Does anyone object?
Eelco
On 4/25/07, Peter Ertl [EMAIL PROTECTED] wrote:
I have a login form which persists username and password using
FormComponent.setPersistent(boolean)
On page load I want to set the focus on
- username, if username is empty
-
Can someone share the code for me? I couldn't get the example from internet.
thanks~
-
This SF.net email is sponsored by DB2 Express
Download DB2 Express C - the FREE version of DB2 express and take
control of your XML.
Is this the right place to post for contrib-related stuff??
I noticed that the yui code used in wicket-contrib-yui is very old.
Current version is 2.2.2, while wicket seems to be using 0.11.4.
I'm still in the investigation stage, but I was thinking about having
the yahoo code served by yahoo
Hello!
I have overridden Page with PlainTextPage that outputs text/plain.
I'm wondering: what's the best way to be able to add the plain-text
equivalent of panels to my template?
wicket-contrib-velocity seems to be no good, since it appears to be for
html templates.
Should I implement my own
Is this the right place to post for contrib-related stuff??
Sure.
I noticed that the yui code used in wicket-contrib-yui is very old.
Current version is 2.2.2, while wicket seems to be using 0.11.4.
Yeah, it's been a while since anyone seriously worked on that.
I'm still in the
Hello!
I have overridden Page with PlainTextPage that outputs text/plain.
Hmmm. Does that work? Wicket still needs tags to operate...
I'm wondering: what's the best way to be able to add the plain-text
equivalent of panels to my template?
Resources. See for instance TextTemplate and
Hi Igor,
Thanks for your reply. Here is the thing:
(using wicket 1.2.5)
I could not find the constructor that takes the Application object. So I
tried the following.
tester = new WicketTester();
tester.set(new xyzApplication());
But this gives me the
David, if you send us your sourceforge id, (not the literal id, but
the name, like mine is eelco12), we can give you commit rights, and
you can decide for yourself whether and when you play with it.
My id = dleangen
Thanks, Eelco!
Dave
Just to be sure we're talking about the same thing here... ;-)
I have overridden Page with PlainTextPage that outputs text/plain.
Hmmm. Does that work? Wicket still needs tags to operate...
Since there's so much nice stuff already available through Wicket and
my Wicket framework is already
You know that you can remove all tag from output, do you? To remove
wicket:xx is simply a matter of settings and all other tags (e.g.
span wicket:id=..) can be removed by subclassing onComponentTag
and not output anything.
Juergen
On 4/27/07, David Leangen [EMAIL PROTECTED] wrote:
Just to be
86 matches
Mail list logo