Re: alternative solution for addOrReplace()?
Hi Andrea,
this is the code (a bit pseudo-like...) I am quite sure it is not the
way it should be and full of wicket anti-patterns. But still exists
since my first wicket experiences 2 years ago. ;-)
This panel gets rendered via RefreshingView about 30 times.
public ElementPanel(String id, IModel? extends ElementState
elementState){
super(id, elementState);
container = new WebMarkupContainer(innerElementContainer,
elementState){
private static final long serialVersionUID = 1L;
@Override
protected void onConfigure() {
super.onConfigure();
ElementState s = (ElementState)getDefaultModelObject();
if(s instanceof DefaultState){
container.addOrReplace(new
DefaultStatePanel(REPLACE_CONTAINER_ID, s));
}else if(s instanceof SpecialState){
container.addOrReplace(new
SpecialStatePanel(REPLACE_CONTAINER_ID, s));
}else if(s instanceof AnotherSpecialState){
container.addOrReplace(new
AnotherSpecialStatePanel(REPLACE_CONTAINER_ID, s));
...
}else{
String error = No panel registered for class
+getDefaultModelObject().getClass().getSimpleName();
container.addOrReplace(new
Label(REPLACE_CONTAINER_ID, error));
error(error);
}
}
add(container);
}
Patrick
Am 07.10.2014 19:42, schrieb Andrea Del Bene:
Hi,
could you post the code you use inside onConfigure to addOrReplace inner
container? Maybe you could avoid detaching/attaching overriding method
detachModels in your inner containers.
Hi Paul,
thanx for replying.
Oh, I use Ajax very heavily in this case. Thats why I said: the panel
works quite good for ajax rendering and refreshing use-cases
Hmm... ok... I think I missed some details. I try to explain.
My page contains two sections seperated from each other.
A kind of data-entry and a kind of displaying the results.
The section which displays the results is a quite complicated panel,
having an RefreshingView for about 30 (I will name them) sub-panels.
These sub-panels addOrReplace() an inner container while rendering
itself.
And the replacement depends on my model object type.
So, my sub-panel class has a hugh knowledge about how to render the
results.
On entering some data on the data-entry section I refresh the whole
panel in the display section with all its sub-panels using Ajax. And
this works great.
Some of the sub-panels also rerender/refresh own there own, by listening
to events etc pepe. All this behaviours are working very well.
The only disadvantage is, using addOrReplace(), I have a heavy
detaching/attaching situation and a performance problem.
So, the best solution for now would be, keeping my panel implementation
as it is, but just having my models detached only once, not for
iteration.
And here I dont have any idea how to solve that... :-/
thanx and kind regards
Patrick
-
To unsubscribe, e-mail: users-unsubscr...@wicket.apache.org
For additional commands, e-mail: users-h...@wicket.apache.org
Re: alternative solution for addOrReplace()?
I think you might have over-complicated your design. Yes, I also think so, but unfortunately currently I'm not able to redesign it, and I try to find a quick solution. :-/ (please have a look at my reply to andreas mail. There is some code...) should be simple to update the parent which in turn delegates that responsibility to the children. It is -somehow- the way you described. I think... hmm... but not very good implemented it seems. Take a quick look over how Decebal implemented his Wicket-Dashboards Ok, thanx, I will do... Otherwise I suggest you get rid of the addOrReplace() and move that logic into your onBeforeRender() methods for those panels to update themself. Are markup modifications handled different by onBeforeRender() / add() compared to onConfigure() / addOrReplace()? Or do I also have to use addOrReplace() in onBeforeRender()? Patrick Am 07.10.2014 21:47, schrieb Paul Bors: I think you might have over-complicated your design. Your use-case is quite simple, you have user input that filters the display. It might be that there are a lot of panels to be updated but then again it should be simple to update the parent which in turn delegates that responsibility to the children. Take a quick look over how Decebal implemented his Wicket-Dashboards as in his project the entire dashboard is refreshed when adding a new panel which in turn can be quite complicated. His code and examples are at: https://github.com/decebals/wicket-dashboard Study his design and see how much of that you can apply to your page. Otherwise I suggest you get rid of the addOrReplace() and move that logic into your onBeforeRender() methods for those panels to update themself. On Tue, Oct 7, 2014 at 12:43 PM, Patrick Davids patrick.dav...@nubologic.com wrote: Hi Paul, thanx for replying. Oh, I use Ajax very heavily in this case. Thats why I said: the panel works quite good for ajax rendering and refreshing use-cases Hmm... ok... I think I missed some details. I try to explain. My page contains two sections seperated from each other. A kind of data-entry and a kind of displaying the results. The section which displays the results is a quite complicated panel, having an RefreshingView for about 30 (I will name them) sub-panels. These sub-panels addOrReplace() an inner container while rendering itself. And the replacement depends on my model object type. So, my sub-panel class has a hugh knowledge about how to render the results. On entering some data on the data-entry section I refresh the whole panel in the display section with all its sub-panels using Ajax. And this works great. Some of the sub-panels also rerender/refresh own there own, by listening to events etc pepe. All this behaviours are working very well. The only disadvantage is, using addOrReplace(), I have a heavy detaching/attaching situation and a performance problem. So, the best solution for now would be, keeping my panel implementation as it is, but just having my models detached only once, not for iteration. And here I dont have any idea how to solve that... :-/ thanx and kind regards Patrick Am 07.10.2014 17:52, schrieb Paul Bors: Why not use Ajax? addOrReplace() is really for when you need the full request cycle and have the page rendered on the server side. With Ajax you get your target for which you can add all the components you want to refresh on the page. Given the day and age we live in, I don't think there is a need to even worry about browsers that do not support Ajax. Than again, that depends on your product's requirements. Otherwise, you can keep on using addOrReplace() but stop using instaceof and start using Generics. If you need an example code-snippet of that let me know and I'll dig some out for you. On Tue, Oct 7, 2014 at 9:51 AM, Patrick Davids patrick.dav...@nubologic.com wrote: Hi all, I have a panel, which renders a inner markupcontainer (different panels for different ways of displaying my model object) depending on its model objects type. I do this by addOrReplace() the inner panel in an onConfigure() having an if instanceof; and it also iterates over a list... I'm not sure, if it is a good solution (I dont like the instanceof approach), but the panel works quite good for ajax rendering and refreshing use-cases. But, the addOrReplace() has one big disadvantage. It forces to detach the model on each iteration, so on each iteration I get a database access to retrieve it again. Is there anyway to keep the model attached until the entire request cycle is finished? Or should I try to find complete other solution for this kind of high dynamic iterating panel? Is it better to use e.g. Fragments to display a model object in different ways depending on its state? Please give some inspirations... I ran out of ideas. Help! kind regards :-) Patrick -
Re: unable to find property UploadProgressBar.starting
I just wanted to report back and say that, after our installer team updated the distribution to maintain the individual JARs (and thus not overwrite the files with clashing filenames), that solved the problem! Thanks to Martin Grigorov and Andrew Geery for helping point me in the direction of the problem. Garret On 10/1/2014 1:04 PM, Garret Wilson wrote: I think I've found the source of the problem (even though I don't understand the internal details). Our installer creates an uber-JAR that has all the dependencies exploded and then placed inside a single JAR file. I looked inside wicket-extensions-7.0.0-M3.jar, and it has a file wicket.properties /in the root of the JAR file!/ Unfortunately, wicket-core-7.0.0-M3.jar (and likely other Wicket JARs) also have a wicket.properties file in their root as well. As you might guess, these have conflicting values: initializer=org.apache.wicket.extensions.Initializer initializer=org.apache.wicket.Initializer So when we create our uber-JAR only one of these wicket.properties files wins and gets placed in the uber-JAR. The one we happen to have now contains initializer=org.apache.wicket.Initializer. I will check with the installer team to see if we can distribute the application with all its dependencies as separate JARs rather than an uber-JAR. But on Wicket's side, is it really a good practice to stick some file in the root directory of a JAR, outside of any package, with a name you expect to be identical across JARs but with different contents? I naively would imagine that some better approach exists. Garret On 10/1/2014 1:22 PM, Martin Grigorov wrote: The .properties file is packed inside wicket-extensions.jar, not in his application. I have no other ideas but to attach a remote debugger to org.apache.wicket.resource.loader.InitializerStringResourceLoader#loadStringResource(java.lang.Class?, java.lang.String, java.util.Locale, java.lang.String, java.lang.String) and set condition on the key parameter to be equal to the missing resource key. Martin Grigorov Wicket Training and Consulting https://twitter.com/mtgrigorov On Wed, Oct 1, 2014 at 5:55 PM, Andrew Geery andrew.ge...@gmail.com wrote: As a sanity check, is the property file with the property UploadProgressBar.starting in the jar file? Perhaps it didn't get copied over by the Maven build process into the jar but the IDE was properly copying it over... Andrew On Wed, Oct 1, 2014 at 11:38 AM, Garret Wilson gar...@globalmentor.com wrote: On 10/1/2014 12:33 PM, Martin Grigorov wrote: Hi, Do you by chance manipulate the list of IStringResourceLoader's in DEPLOYMENT mode ? I don't think I touched anything related to IStringResourceLoader. The only thing I've done relating to modes is this: //turn on Wicket development mode if in debug mode final String wicketConfiguration = Debug.isDebug() ? development : deployment; filterHolder.setInitParameter(configuration, wicketConfiguration); In other words, in embedded Jetty, if we've started up specifying a debug mode (using a system variable), then I set the Wicket configuration to development; otherwise, I set it to deployment. Garret - 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
Re: Demonstrate End-to-End Security Enforcement using Open Source Software Wicket
On 08/20/2014 10:08 AM, Shawn McKinney wrote: Notably missing from the material is theory or why these types of complex security mechanisms are necessary. I'm working on that now and will publish it back here when ready. Hello again, just now getting back to this thread *** The fortressdemo2 web app tutorial shows an apache wicket web app deployed inside of a tomcat container using both an ldap and db server. It recommends various security layers for end-to-end security which is a 'defense in depth' approach. The fortressdemo2 source code is here: https://github.com/shawnmckinney/fortressdemo2 The fortress demo2 tutorial page has been moved to a new location: https://symas.com/kb/demonstrate-end-to-end-security-enforcement-using-open-source/ and on this page are more links to: a. static html javadoc (hosted on same server) containing instructions for actual fortressdemo2 tutorial installation. The overview page of the javadoc describes how to download the example source code and how to generate documentation locally. b. link to presentation given last week at JavaOne The J1 deck contains two parts: 1. Overview of the security controls used within the fortressdemo2 web app. 2. Description of how to drop the fortressdemo2 (and its associated infrastructure) into a cloud foundry PaaS (presented by John Field) Finally there is an abbreviated version of the slides containing the rationale for each layer by comparing to everyday situations: https://symas.com/javadocs/fortressdemo2/doc-files/AnatomyOfSecureWebApp.pdf We are donating this material to help others learn the proper way to security inside of web app envs. So there will be less violations and breaches of our personal and business data - events that are seemingly commonplace today. Suggestions or comments are welcome. Thanks for your attention, Shawn - To unsubscribe, e-mail: users-unsubscr...@wicket.apache.org For additional commands, e-mail: users-h...@wicket.apache.org
Re: Could not clear select2Choice component model value.
Thankyou very much for yours reply. I call formComponent.clearInput() in my ClearFormVisitor class. But It does not work. I have created issue-96 https://github.com/ivaynberg/wicket-select2/issues/96 in wicket-select2 project. On Wed, Oct 8, 2014 at 12:13 AM, Maxim Solodovnik solomax...@gmail.com wrote: Done, thanks for pointing this out On 7 October 2014 23:08, Paul Bors p...@bors.ws wrote: Can you update this issue then? https://github.com/ivaynberg/wicket-select2/issues/93 On Tue, Oct 7, 2014 at 12:00 PM, Maxim Solodovnik solomax...@gmail.com wrote: we recently moved this component to wicketstuff: https://github.com/wicketstuff/core/tree/master/jdk-1.7-parent/select2-parent to get Wicket7 compatible version On 7 October 2014 22:56, Paul Bors p...@bors.ws wrote: You can also ask the developer via: https://github.com/ivaynberg/wicket-select2/issues You should probably take a look over the open issues so that you are familiar with that other developers faced. Maybe one of those issues might be a road block for you? Although you will get a faster reply from Igor or another Wicket developer via this list :) On Tue, Oct 7, 2014 at 4:23 AM, Martin Grigorov mgrigo...@apache.org wrote: Hi, You should call formComponent.clearInput() too. Martin Grigorov Wicket Training and Consulting https://twitter.com/mtgrigorov On Tue, Oct 7, 2014 at 9:13 AM, MadasamySankarapandian madas...@mcruncher.com wrote: This is regarding wicket-select2 project. I could not find any mailing list for this project. That is why sending here. Please some one help me. I have created clear link to clear the form input values. After click the clear link, all components values are cleared except Select2Choice component. Here with attached quickstart to recreate this problem. Follow below steps to recreate this problem * Download attachment and extract it. * cd select2-quickstart. * Then execute mvn clean package jetty:run * Type localhost:8080 in browser * Give the input to name and country field. * Click clear link Now name(*TextField*) field value is cleared and country(*Select2Choice*) field value is not cleared. -- Thanks and regards Madasamy - To unsubscribe, e-mail: users-unsubscr...@wicket.apache.org For additional commands, e-mail: users-h...@wicket.apache.org -- WBR Maxim aka solomax -- WBR Maxim aka solomax -- Thanks and regards Madasamy
Re: Could not clear select2Choice component model value.
I also noticed this issue in our project, Will try to take a look at it On 9 October 2014 07:33, MadasamySankarapandian madas...@mcruncher.com wrote: Thankyou very much for yours reply. I call formComponent.clearInput() in my ClearFormVisitor class. But It does not work. I have created issue-96 https://github.com/ivaynberg/wicket-select2/issues/96 in wicket-select2 project. On Wed, Oct 8, 2014 at 12:13 AM, Maxim Solodovnik solomax...@gmail.com wrote: Done, thanks for pointing this out On 7 October 2014 23:08, Paul Bors p...@bors.ws wrote: Can you update this issue then? https://github.com/ivaynberg/wicket-select2/issues/93 On Tue, Oct 7, 2014 at 12:00 PM, Maxim Solodovnik solomax...@gmail.com wrote: we recently moved this component to wicketstuff: https://github.com/wicketstuff/core/tree/master/jdk-1.7-parent/select2-parent to get Wicket7 compatible version On 7 October 2014 22:56, Paul Bors p...@bors.ws wrote: You can also ask the developer via: https://github.com/ivaynberg/wicket-select2/issues You should probably take a look over the open issues so that you are familiar with that other developers faced. Maybe one of those issues might be a road block for you? Although you will get a faster reply from Igor or another Wicket developer via this list :) On Tue, Oct 7, 2014 at 4:23 AM, Martin Grigorov mgrigo...@apache.org wrote: Hi, You should call formComponent.clearInput() too. Martin Grigorov Wicket Training and Consulting https://twitter.com/mtgrigorov On Tue, Oct 7, 2014 at 9:13 AM, MadasamySankarapandian madas...@mcruncher.com wrote: This is regarding wicket-select2 project. I could not find any mailing list for this project. That is why sending here. Please some one help me. I have created clear link to clear the form input values. After click the clear link, all components values are cleared except Select2Choice component. Here with attached quickstart to recreate this problem. Follow below steps to recreate this problem * Download attachment and extract it. * cd select2-quickstart. * Then execute mvn clean package jetty:run * Type localhost:8080 in browser * Give the input to name and country field. * Click clear link Now name(*TextField*) field value is cleared and country(*Select2Choice*) field value is not cleared. -- Thanks and regards Madasamy - To unsubscribe, e-mail: users-unsubscr...@wicket.apache.org For additional commands, e-mail: users-h...@wicket.apache.org -- WBR Maxim aka solomax -- WBR Maxim aka solomax -- Thanks and regards Madasamy -- WBR Maxim aka solomax
