gmane.comp.java.dwr.user

[dwr-users] Re: Web application architecture using DWR

Mike Wilson — 2012-05-10T22:03:09

One pattern I like a lot is to keep the GUI code entirely in the client and
make a kind of service layer with the classes remoted by DWR. Each page
calls the services it needs for its data and then uses f ex client-side
templating (Moustache and similar) to render views. Everything except DWR
calls may be cached in the browser and pages link to each other in
traditional path+parameters style. This allows you to build a site at any
scale and is also efficient on bandwidth. It is also very easy to refactor
the UI to f ex show something in a popup instead of navigating to a
different page.
 
Best regards
Mike Wilson
 
Dhaval Patel wrote:

Hi David, 

Thank you for your update. 

JQuery is used just for common HTML component manipulation and calling DWR
functions. I have no plan to use JQuery UI widgets. Rather I would try to
stay away from widget if I have to and use plain HTML5 goodies.

I will definitely consider your suggestion about Hibernate and Sprint
security.

Can you tell me abstractly how would DWR fi

[dwr-users] Re: Web application architecture using DWR

David Marginian — 2012-05-08T18:33:48

The strength of DWR is that it allows you to code your business logic
without knowledge of the web client, thus allowing all clients to call into
and re-use that logic.  DWR's Spring integration allows you to use Spring
as a bean container/manager and helps reduce the amount of DWR specific
configuration.

Newer versions of Spring MVC have nice REST features.  Based on what you
need you may choose to follow this approach as an alternative to using
DWR.  There was a discussion on the mailing list regarding this:
http://dwr.2114559.n2.nabble.com/Problem-with-Interface-implementation-in-Spring2-DWR3-td6077025.html#a6078729
.

On Tue, May 8, 2012 at 12:18 PM, Dhaval Patel <dhaval04-Re5JQEeQqe8AvxtiuMwx3w< at >public.gmane.org> wrote:

[dwr-users] Re: Web application architecture using DWR

Dhaval Patel — 2012-05-08T18:18:02

Hi David,

Thank you for your update.

JQuery is used just for common HTML component manipulation and calling DWR
functions. I have no plan to use JQuery UI widgets. Rather I would try to
stay away from widget if I have to and use plain HTML5 goodies.

I will definitely consider your suggestion about Hibernate and Sprint
security.

Can you tell me abstractly how would DWR fit nicely in this architecture?
In my small web app I used DWR Servlet and call the Java objects and
functions directly from Javascript using DWR. It works awesomely on small
scale but going big, I don't think it would be a good choice. I know DWR
has spring integration but how does it fit in architecture?

Thanks.

On Tue, May 8, 2012 at 2:07 PM, David Marginian <david-H55tQ8LTKd1xxXtqJ93Y6g< at >public.gmane.org> wrote:

[dwr-users] Re: Web application architecture using DWR

David Marginian — 2012-05-08T18:07:50

Sounds like you have a good start.  A few comments:

JQuery is good at what it does but I tend to not use it much because of the
weakness of JQuery UI's widget selection and robustness.  I prefer to work
with Ext JS (hefty license) as their widgets are top notch or Dojo.
Obviously this is personal preferences and my decisions are based on the
requirements of the application.

I also like to use Spring along with Hibernate JPA.  For security I would
recommend you take a look at Spring Security (
http://static.springsource.org/spring-security/site/).



On Tue, May 8, 2012 at 8:43 AM, Dhaval Patel <dhaval04-Re5JQEeQqe8AvxtiuMwx3w< at >public.gmane.org> wrote:

[dwr-users] Web application architecture using DWR

Dhaval Patel — 2012-05-08T14:43:07

Hello all,

I just wanted to ask users of DWR about their application architecture. I
mean if you have to make a new web application using DWR, how would you
architect it? Which MVC do you use if you are planning to use one? What
would makes a good sense from growth and expansion point of view?

I have used DWR but on very small scale application like 5-6 page. I am
thinking about writing a small web application. It may grow bigger but my
main concern is about technology I used. I want to make it as light weight
as possible. So far this is what I have in my mind:
- Front end: HTML, CSS, JQuery, DWR
- Back end: Spring MVC, DWR

Front end will be as much decoupled as possible so that I can replace
backend if I needed to.

Lot of things are missing from this picture like authentication,
authorization etc. But I just wanted to reach out to DWR users about their
experience in their work and if they have to change something, what would
they change to make the application better?

Also I want to ask authors of DWR

[dwr-users] Re: Question on SessionScript handling

David Marginian — 2012-05-07T11:28:18

Yes, you may use it A.J.

On 05/07/2012 12:58 AM, newbie wrote:

[dwr-users] Re: Question on SessionScript handling

newbie — 2012-05-07T06:58:25

Hi David,

We have planned for a release of our product by next month. I would like to
know if its permitted to bundle the version of DWR which you had attached in
this thread, the one with the temporary fix for the discussed issue. Later
when Mike & team comes up with a proper fix in upcoming GA r RC whatever, I
would migrate to the same. Please confirm if I can bundle this test fix DWR
jar without any license complications.

Regards,
A.J.Sudharshan.

--
View this message in context: http://dwr.2114559.n2.nabble.com/Question-on-SessionScript-handling-tp7409664p7534175.html
Sent from the DWR - Users mailing list archive at Nabble.com.

[dwr-users] Re: ScheduledThreadPoolExecutor won't run after intialized

Garfi61416 — 2012-05-04T18:04:59

Okay, I'll tackle it internally.  Thanks.

--
View this message in context: http://dwr.2114559.n2.nabble.com/ScheduledThreadPoolExecutor-won-t-run-after-intialized-tp7525522p7528106.html
Sent from the DWR - Users mailing list archive at Nabble.com.

[dwr-users] Re: ScheduledThreadPoolExecutor won't run after intialized

David Marginian — 2012-05-04T17:45:54

1) Simplify what you have.  You said it worked originally so what has
changed?  Slowly remove pieces to find out what is wrong.  It wouldn't hurt
to start from our official example.
2) Learn how to debug through code running on your server.

On Fri, May 4, 2012 at 11:27 AM, Garfi61416 <garfi61416-/E1597aS9LQAvxtiuMwx3w< at >public.gmane.org> wrote:

[dwr-users] Re: ScheduledThreadPoolExecutor won't run after intialized

Garfi61416 — 2012-05-04T17:27:51

#Tue Jun 28 15:30:30 PDT 2011
bamboo.build.number=312
major=3
revision=0
minor=0
title=RC2-final


--
View this message in context: http://dwr.2114559.n2.nabble.com/ScheduledThreadPoolExecutor-won-t-run-after-intialized-tp7525522p7527924.html
Sent from the DWR - Users mailing list archive at Nabble.com.

[dwr-users] Re: ScheduledThreadPoolExecutor won't run after intialized

David Marginian — 2012-05-04T17:09:05

What version of DWR are you using?

On Fri, May 4, 2012 at 10:49 AM, Garfi61416 <garfi61416-/E1597aS9LQAvxtiuMwx3w< at >public.gmane.org> wrote:

[dwr-users] Re: ScheduledThreadPoolExecutor won't run after intialized

Garfi61416 — 2012-05-04T16:49:30

Looks like it didn't like my gt; lt; tags, trying again:

DWR XML looks like this:

&lt;dwr&gt;
  &lt;allow&gt;
*    &lt;create creator="new" scope="application"&gt;
      &lt;param name="class" value="com.jfs.util.AlertRunner"/&gt;
    &lt;/create&gt;*    
    &lt;create creator="new" scope="application"&gt;
      &lt;param name="class" value="com.jfs.util.JFSInfo"/&gt;
    &lt;/create&gt;
    &lt;convert converter="bean" match="com.jfs.util.JFSInfo" /&gt;

    &lt;create creator="new" scope="application"&gt;
      &lt;param name="class" value="com.jfs.ws.JFSWSClient"/&gt;
    &lt;/create&gt;

    &lt;convert match="java.lang.Exception" converter="exception"&gt;
  &lt;param name='include' value='message,lineNumber'/&gt;
    &lt;/convert&gt;      
  &lt;/allow&gt;
&lt;/dwr&gt;

--
View this message in context: http://dwr.2114559.n2.nabble.com/ScheduledThreadPoolExecutor-won-t-run-after-intialized-tp7525522p7527778.html
Sent from the DWR - Users mailing list archive at Nabble.com.

[dwr-users] Re: ScheduledThreadPoolExecutor won't run after intialized

Garfi61416 — 2012-05-04T16:47:27

This is my DWR file:

<dwr>
  <allow>
*    <create creator="new" scope="application">
      
    </create>*    
    <create creator="new" scope="application">
      
    </create>
    <convert converter="bean" match="com.jfs.util.JFSInfo" />

    <create creator="new" scope="application">
      
    </create>

    <convert match="java.lang.Exception" converter="exception">
  
    </convert>      
  </allow>
</dwr>

And I read the FAQ so I know that engine will only initiailize application
level when it is first called, which is what I want because I will be having
lots of people accessing the alertrunner and it will only come into affect
when they are on that specific landingpage.  However, when they are on the
landing page, I want alertrunner to continue to update a table on that page
for them.

Thanks for continuing to look at this.

--
View this message in context: http://dwr.2114559.n2.nabble.com/ScheduledThreadPoolExecutor-won-t-run-after-intialized-tp7525522p7527772.html
Sent from the DWR - Users mai

[dwr-users] Re: ScheduledThreadPoolExecutor won't run after intialized

David Marginian — 2012-05-04T16:31:07

"I could never quite figure out how to debug a runnable within eclipse, but
adding System.out.printlns at different points in the code, I see the
following behavior:"

If you don't know how to debug through server code in Eclipse you are
missing out on an essential debugging tool.

"So on page load, it looks like initializes the executor twice, then calls
the update display function, which is expected from the javascript onload
method, followed by executing each initializer once and then that's it.  Any
ideas why this would be happening?"

What is the scope of your AlertRunner creator?  You are making two calls to
it in the JavaScript you posted so the behavior you are seeing may be
expected based on the scope you have specified:
http://directwebremoting.org/dwr/documentation/server/configuration/dwrxml/creators/index.html
.


On Fri, May 4, 2012 at 10:21 AM, Garfi61416 <garfi61416-/E1597aS9LQAvxtiuMwx3w< at >public.gmane.org> wrote:

[dwr-users] Re: ScheduledThreadPoolExecutor won't run after intialized

Garfi61416 — 2012-05-04T16:21:25

I know and I apologize for all the code but when I initially wrote the
runnable it would continue to run even when I navigated away from the page,
which was the exact opposite behavior (but closer to the desired outcome). 
I could never quite figure out how to debug a runnable within eclipse, but
adding System.out.printlns at different points in the code, I see the
following behavior:

Initialize ScheduledThreadPoolExecutor
Initialize ScheduledThreadPoolExecutor
In updateTableDisplay
In browser runnable.
At main runnable.
In updateTableDisplay
In browser runnable.
At main runnable.
In updateTableDisplay
In browser runnable.

So on page load, it looks like initializes the executor twice, then calls
the update display function, which is expected from the javascript onload
method, followed by executing each initializer once and then that's it.  Any
ideas why this would be happening?

One other note, I created a custom scriptsessionlistener in a servlet that
calls the landingpage so that I could pass custom vari

[dwr-users] Re: ScheduledThreadPoolExecutor won't run after intialized

David Marginian — 2012-05-04T13:36:48

I suggest debugging a bit more.  It is pretty difficult to look at 200
lines of code and configuration and be able to tell exactly what is going
on.

On Thu, May 3, 2012 at 5:48 PM, Garfi61416 <garfi61416-/E1597aS9LQAvxtiuMwx3w< at >public.gmane.org> wrote:

[dwr-users] ScheduledThreadPoolExecutor won't run after intialized

Garfi61416 — 2012-05-03T23:48:38

I took the PeopleTable method and built my own runner method.  It works
perfectly fine with a direct call.  However, after seeing that
ScheduledThreadPoolExecutor gets initialized, the
ScheduledThreadPoolExecutor run method is never called.  Can anyone tell me
what I'm doing wrong?

Here is my runnable class:

public class AlertRunner implements Runnable
{

// Create default logging
static Logger log = Logger.getLogger(AlertRunner.class);

ScheduledThreadPoolExecutor executor = null;

/**
 * Constructor - Creates a thread pool that runs every minute.
 */
public AlertRunner()
{
System.out.println("Initialize ScheduledThreadPoolExecutor");
executor = new
ScheduledThreadPoolExecutor(1, new DaemonThreadFactory());
executor.scheduleAtFixedRate(this, 1, 10, TimeUnit.SECONDS);

int queued = executor.getQueue().size(); 
int active = executor.getActiveCount(); 
int notCompleted = queued + active; // approximate
System.out.println(String.format("\nBeginning Executor Statistics:\nqueued
%1$

[dwr-users] Re: dwr creator, non static and non new?

David Marginian — 2012-05-01T19:55:21

I think our docs may have confused you.  Do you have the following in your
web.xml?

<servlet>
  <servlet-name>dwr-invoker</servlet-name>
  <servlet-class>org.directwebremoting.server.tomcat.DwrCometProcessor</servlet-class>
  ...
</servlet>

If so you should remove this The DWR integration with Tomcat Async is
not working correctly.  As I said
earlier you can still use DWR's reverse Ajax features with Tomcat but
we will not take advantage of their
Async APIs.  In the configuration that works you should NOT be using
the DwrCometProcessor.



On Tue, May 1, 2012 at 1:36 PM, David Marginian <david-H55tQ8LTKd1xxXtqJ93Y6g< at >public.gmane.org> wrote:

[dwr-users] Re: dwr creator, non static and non new?

David Marginian — 2012-05-01T19:36:56

http://directwebremoting.org/dwr-demo/reverseajax/clock.html

The clock seems to work fine here.  I would recommend downloading the war
and running the clock example as is in your environment I don't think you
will see the behavior you describe.

On Tue, May 1, 2012 at 1:33 PM, stigmund <steve.ville-XZoyATsUNX5Wk0Htik3J/w< at >public.gmane.org> wrote:

[dwr-users] Re: dwr creator, non static and non new?

stigmund — 2012-05-01T19:33:25

sorry, i was talking about the clock example. i was just testing some other
stuff, but...using the clock example exactly as it is: i press start/stop,
and it displays the time once and only once, untill i press it again... i
tested it to see if the thread was still running by adding a print line. 

cant remember exactly the code, im not there atm, but i tried it with all 3
clock examples (seconds, miliseconds etc...) each used a slightly different
method of obtaining the session, but all fail to update the browser after
the first push.
 
Browser.withAllPages..etc(..new runnable()
{
  public void run()
  {
    Util.setValue("clockDisplay", output);
    System.out.println("TEST");
  }
}

when i press play, TEST is spammed to my console once every 1 second, (or
whatever i set it too) but the output is only updated on the browser once.

The web.xml and dwr.xml i posted are what i've been using, the code is from
dwr svn repository. The only other thing i can think of is that the page
itself is a transformed XSLT

[dwr-users] Re: dwr creator, non static and non new?

David Marginian — 2012-05-01T15:08:14

I am not following you here.  The clock example does exactly what you
want.  It updates the same element over and over again.  What you are doing
with the multiple calls to setValue with the sleeps makes no sense.  My
guess is (you haven't sent complete code for me to determine) is that they
are being sent to the browser in one push which means the last one wins.
You can prove this in Firebug (or whatever tool you prefer to examine the
push response).  Once again I ask you to look at the Clock demos code.

On Tue, May 1, 2012 at 8:48 AM, stigmund <steve.ville-XZoyATsUNX5Wk0Htik3J/w< at >public.gmane.org> wrote: