date:20100413


 re. Is it documented anywhere?

See GWT Presentations page -
http://code.google.com/webtoolkit/media_gallery.html. The Story of Your
Compile presentation explains this, and a lot of other optimizations.

Other than that, I am not aware of any other documentation.

--Sri



On 13 April 2010 15:38, Jon Vaughan jsvaug...@gmail.com wrote:

 aha, that is helpful.  I didn't know about the existence of the
 {App}.rpc.log file (is that documented anywhere?)

 for anyone reading this who isn't aware, this log file is presumably
 created as the compiler works out what types are reachable and whether
 they are serializable and goes something like:

 com.foo.bar.MyClass
   Serialization status
  Instantiable
   Path
  'com.foo.bar.MyClass' is reachable as a subtype of type
 'interface com.bar.whatever.ICommandT'
  Started from 'com.itf.shared.actions.ICommandT'

 and so on.


 On Apr 12, 3:05 pm, kozura koz...@gmail.com wrote:
  Yes, T must extend a serializable type, as you said otherwise GWT
  can't tell what it might be.  Yes serialization discovery issues are a
  pain...right now you just have to stare through the RPC log and try to
  divine what happened.
 
  On Apr 12, 4:46 am, Jon Vaughan jsvaug...@gmail.com wrote:
 
 
 
   I think the problem is that on this example the type of the payload
   can be anything (it is not required to be serializable); given that
   this is possible, GWT must say, OK, then this type itself cannot be
   serialized (I would like this to fail the compile though somehow)
 
   On Apr 12, 10:27 am, Jon Vaughan jsvaug...@gmail.com wrote:
 
Hi,
 
I have a serialization issue to solve, where a class that is passed
 to
the client, and is marked as IsSerializable, with a default no arg
constructor, does not end up in the whitelist.  There are no messages
during the gwt compile (at debug level)
 
1.  Does anyone have any thoughts about how I could automatically
 test
for, or fail a build, if this is the case?
2.  Does anyone know how I can get detailed information from GWT
 about
this type of issue?
3.  Anybody got any ideas what the problem itself is? :) (Are there
issues with generic types and serialization?  Do I have to explicitly
whitelist the variations I want to use somehow?)
 
Thanks if anyone has the time to look
 
Jon
 
My specific problem is as follows: Type
'com.itf.shared.actions.ResponseWithPayload' was not included in the
set of types which can be serialized by this SerializationPolicy
 
With the following command pattern style interface (where
SingleQuestion is itself in the whitelist)
 
1.  The service
 
public interface ApplicationService extends RemoteService {
T extends IResponse T execute(ICommandT action);
 
}
 
2. The command
 
public class SingleQuestionRequest implements
ICommandResponseWithPayloadSingleQuestion, IsSerializable {
public SingleQuestionRequest() {
}
 
}
 
3.  The response class that doesn't end up in the whitelist
 
public class ResponseWithPayloadT implements IResponse,
IsSerializable {
private T payload;
 
@SuppressWarnings(unused)
private ResponseWithPayload() {
}
 
public ResponseWithPayload(T payload) {
super();
this.payload = payload;
}
 
public T getPayload() {
return payload;
}
 
}

 --
 You received this message because you are subscribed to the Google Groups
 Google Web Toolkit group.
 To post to this group, send email to google-web-tool...@googlegroups.com.
 To unsubscribe from this group, send email to
 google-web-toolkit+unsubscr...@googlegroups.comgoogle-web-toolkit%2bunsubscr...@googlegroups.com
 .
 For more options, visit this group at
 http://groups.google.com/group/google-web-toolkit?hl=en.



-- 
You received this message because you are subscribed to the Google Groups 
Google Web Toolkit group.
To post to this group, send email to google-web-tool...@googlegroups.com.
To unsubscribe from this group, send email to 
google-web-toolkit+unsubscr...@googlegroups.com.
For more options, visit this group at 
http://groups.google.com/group/google-web-toolkit?hl=en.

Re: Loading please wait + Animated GIF

Javascript is single threaded, and I presume the entire page as well. If you
are running a javascript loop, the animation will appear sluggish or will
stop, there isn't much you can do about it.

If you want to run heavy javascript, use the IncrementalCommand or
DeferredCommand patterns. This would allow the browser to do other
activities, including animating the gif.

--Sri

On 13 April 2010 16:00, Yogesh yogeshrn...@gmail.com wrote:

Hi,
I am showing a Loading, please wait.. message along with an animated
gif.
For this I have added following in my host page.
div id=loadingimg src=loading.gif/ Loading Please wait../div
I have used http://www.ajaxload.info/ to generate a simple loading
GIF image.
And onModuleLoad() I am am removing this div before adding actual
components into Rool Panel. This is working fine.

The issue is that the animation of the GIF is not displayed. I found
out that when javascript code in mymodule.nocache.js starts running in
browser, the animation of GIF stops.

I even wrote following HTML page to test this. When I click on Click
me button, I just run some big for loop in javascript. When the
javascript is running, the animation of GIF is stopped.

script
function bigJS()
{
for (var i=1; i=1000; i++);
alert('done');
}
/script
BODY
div id=loadingimg src=loading.gif/ Loading Please wait../div
input id=btnClick name=btnClick type=button value=Click me!
onclick=bigJS() /
/BODY

Anyone know how to display the animated GIF properly?

--
You received this message because you are subscribed to the Google Groups
Google Web Toolkit group.
To post to this group, send email to google-web-tool...@googlegroups.com.
To unsubscribe from this group, send email to
google-web-toolkit+unsubscr...@googlegroups.com.
For more options, visit this group at
http://groups.google.com/group/google-web-toolkit?hl=en.

Re: Project Stopped Working (wasn't me, I never touched it!)

Just a hunch .. do you have a HTML element (div/span/table etc) whose id is
the same as your module name?

GWT uses a hidden iframe whose id is the rename-to attribute in your
module.xml.. and if you have another element with the same id.. well, things
break unceremoniously.

--Sri

On 13 April 2010 16:03, Ian Bambury ianbamb...@gmail.com wrote:

Hi,

Thanks for the response. It's nothing like that - like I say, I can rename
the project and it works OK. Rename it back and it stops working.

Something somewhere has a setting that stops it working is my guess.

Ian

http://examples.roughian.com

On 12 April 2010 14:15, Katharina Probst kpro...@google.com wrote:

Write permissions in the war directory? Files being read (do you have the
app running from this directory?) so it can't override them...?

kathrin

On Mon, Apr 12, 2010 at 7:18 AM, Ian Bambury ianbamb...@gmail.comwrote:

I have a bit of a weird problem.

A project stopped working. When I run it, it fails to create the
directory in the /war/ directory for the project (as in rename-to).

I wasn't doing anything to the project structure, just coding when it
happened.

The odd thing is that if I just rename the project, it works fine, name
it back again and it stops working.

Any suggestions?

Ian

http://examples.roughian.com

--
You received this message because you are subscribed to the Google Groups
Google Web Toolkit group.
To post to this group, send email to google-web-toolkit@googlegroups.com
.
To unsubscribe from this group, send email to
google-web-toolkit+unsubscr...@googlegroups.comgoogle-web-toolkit%2bunsubscr...@googlegroups.com
.
For more options, visit this group at
http://groups.google.com/group/google-web-toolkit?hl=en.

--
You received this message because you are subscribed to the Google Groups
Google Web Toolkit group.
To post to this group, send email to google-web-tool...@googlegroups.com.
To unsubscribe from this group, send email to
google-web-toolkit+unsubscr...@googlegroups.com.
For more options, visit this group at
http://groups.google.com/group/google-web-toolkit?hl=en.

Re: Project Stopped Working (wasn't me, I never touched it!)

2010-04-13 Thread Ian Bambury

No, I've found the problem. Me!

Somehow the -war argument got an extra character in it - not deliberately,
just a late night thing when I was putting the -port argument back, I
suppose.

Thanks to both of you for you help

Ian

http://examples.roughian.com

On 13 April 2010 12:47, Sripathi Krishnan sripathikrish...@gmail.comwrote:

Just a hunch .. do you have a HTML element (div/span/table etc) whose id is
the same as your module name?

GWT uses a hidden iframe whose id is the rename-to attribute in your
module.xml.. and if you have another element with the same id.. well, things
break unceremoniously.

--Sri

On 13 April 2010 16:03, Ian Bambury ianbamb...@gmail.com wrote:

Hi,

Thanks for the response. It's nothing like that - like I say, I can rename
the project and it works OK. Rename it back and it stops working.

Something somewhere has a setting that stops it working is my guess.

Ian

http://examples.roughian.com

On 12 April 2010 14:15, Katharina Probst kpro...@google.com wrote:

Write permissions in the war directory? Files being read (do you have
the app running from this directory?) so it can't override them...?

kathrin

On Mon, Apr 12, 2010 at 7:18 AM, Ian Bambury ianbamb...@gmail.comwrote:

I have a bit of a weird problem.

A project stopped working. When I run it, it fails to create the
directory in the /war/ directory for the project (as in rename-to).

I wasn't doing anything to the project structure, just coding when it
happened.

The odd thing is that if I just rename the project, it works fine, name
it back again and it stops working.

Any suggestions?

Ian

http://examples.roughian.com

--
You received this message because you are subscribed to the Google
Groups Google Web Toolkit group.
To post to this group, send email to
google-web-tool...@googlegroups.com.
To unsubscribe from this group, send email to
google-web-toolkit+unsubscr...@googlegroups.comgoogle-web-toolkit%2bunsubscr...@googlegroups.com
.
For more options, visit this group at
http://groups.google.com/group/google-web-toolkit?hl=en.

--
You received this message because you are subscribed to the Google Groups
Google Web Toolkit group.
To post to this group, send email to google-web-tool...@googlegroups.com.
To unsubscribe from this group, send email to
google-web-toolkit+unsubscr...@googlegroups.com.
For more options, visit this group at
http://groups.google.com/group/google-web-toolkit?hl=en.

Serialization of lists

2010-04-13 Thread Jochen Schnaidt

Hi,
I posted this already at the AppEngine forum but I think this is an
interface problem so I post here as well.

I’m building an application with GWT and GAE and have a problem saving
data in the datastore. In my application I generate a list which
consists of lists of strings, looks this way: ListListString
trackList;

It compiles but gives me a warning:

[WARN] Warnings in
generated://D271BF8A9063BFFBB6E2618C9E23EF15/eventManagement/shared/HandleEventService_TypeSerializer.java'
[WARN] Line 31: Referencing deprecated class
'com.google.gwt.user.client.ui.ChangeListenerCollection'
And 3 other deprecated classes

When I run the application and call the method for saving data I get
the following error:

javax.jdo.JDOException: Error creating the MetaDataManager for API
JDO :
NestedThrowables:
java.lang.reflect.InvocationTargetException

I think the problem is that the compiler uses old GWT classes for
serialization and not the classes of the GAE. Right?

Any idea how I could fix this problem?

Thanks a lot. Greetings

-- 
You received this message because you are subscribed to the Google Groups 
Google Web Toolkit group.
To post to this group, send email to google-web-tool...@googlegroups.com.
To unsubscribe from this group, send email to 
google-web-toolkit+unsubscr...@googlegroups.com.
For more options, visit this group at 
http://groups.google.com/group/google-web-toolkit?hl=en.

UIBinder with code in earlier versions of GWT

2010-04-13 Thread San

Hi all,

Can I use some of the widgets those are written in earlier versions of
GWT2.0 directly in UIBinder classes.

e.g.can I do something like suppose I am having MyWidget as a old
class and I develop a new class as WidgetContainer with UIBinder and
try to add MyWidget directly to WidgetContainer's ui.xml file without
having ui.xml file for MyWidget?


-Thanks

-- 
You received this message because you are subscribed to the Google Groups 
Google Web Toolkit group.
To post to this group, send email to google-web-tool...@googlegroups.com.
To unsubscribe from this group, send email to 
google-web-toolkit+unsubscr...@googlegroups.com.
For more options, visit this group at 
http://groups.google.com/group/google-web-toolkit?hl=en.

Re: Best (faster) place to include gwt script in html ?

2010-04-13 Thread redlaber

It makes no difference. The html-code generate the page almost
instantly. GWT metod RootPanel.get(id) also work very quickly.
Almost all time of page generation will take the loading of gwt-
module. So place your gwt-application where it is more convenient to
you.

On 13 апр, 13:51, Óscar Frías ofr...@gmail.com wrote:
 Which is the best place to put the gwt script in the HTML page from a
 performance point of view (page speed) ?
 I mean to minimize the load time of a page that contains some HTML
 code in addition to the GWT module.

 Is it better to put it at the end of the head section (immediately
 before the /head tag, or immediately before the analytics code) ?
 Or would it be faster to put it at the end of the html (immediately
 before the /html tag) ?

 Thanks!
 Oscar

-- 
You received this message because you are subscribed to the Google Groups 
Google Web Toolkit group.
To post to this group, send email to google-web-tool...@googlegroups.com.
To unsubscribe from this group, send email to 
google-web-toolkit+unsubscr...@googlegroups.com.
For more options, visit this group at 
http://groups.google.com/group/google-web-toolkit?hl=en.

Re: How to scroll page to element

2010-04-13 Thread redlaber

This is exactly what I was looking for. Thanks much, Jason! I'll try
it tomorrow and write here the result.

On 13 апр, 11:41, Jason Morris lem...@gmail.com wrote:
 I assume you're using GWT 2.0. If so dom.client.Element has a method:
 scrollIntoView which may be what you are looking for.

 To scroll to a Button for example:

 Button button = new Button();
 //...

 button.getElement().scrollIntoView();

 This will scroll any required elements (including ScrollPanels and the
 document itself) to make sure that the Button is visible to the user.

 Hope that helps.
 //Jason

 On Tuesday, 13 April 2010 10:35:21 mram wrote:



  I don't understand well your question, but why don't you use div to
  separate parts from the page, and then call the div class that you
  need??

  On 13 abr, 08:55, redlaber vyalov.e...@gmail.com wrote:
   The next problem: I have finished big gwt-application and I cant add a
   ScrollPanel in it. But i know position of browser scrollbar, where I
   want to jump (scroll). Is there no way to simply scroll through the
   document to a given position?

   On 12 апр, 19:12, t.dave da...@lorgeousdays.com wrote:
use ScrollPanel.setScrollPosition().  that sets the vertical scroll
position of the scrollpanel, which you will need to calculate.  try
something like this:

new ClickHandler() {
  public void onClick( ClickEvent event ) {
    int tableTop = secondFlexTable.getAbsoluteTop();
    scrollPanel.setScrollPosition( tableTop );
  }

}

i suppose in your case it wouldn't be a click handler if you're basing
it off the history token, but hopefully this will point you in the
right direction.

On Apr 12, 7:34 am, redlaber vyalov.e...@gmail.com wrote:
 I want to scroll page, generated with gwt, to some element. Its
 should be simple, but i cannt find the solution. For example: I have
 a vertical panel with two flexables. When i get the history token
 goto2 I want to scroll my page to the second table. (Sorry for the
 terrible english).

-- 
You received this message because you are subscribed to the Google Groups 
Google Web Toolkit group.
To post to this group, send email to google-web-tool...@googlegroups.com.
To unsubscribe from this group, send email to 
google-web-toolkit+unsubscr...@googlegroups.com.
For more options, visit this group at 
http://groups.google.com/group/google-web-toolkit?hl=en.

Re: Serialization of lists

Check out these threads, it's a bug in the RPC serializer, there's a
workaround on the second link.

http://groups.google.com/group/google-web-toolkit/browse_thread/thread/cbf566637f03496a
http://code.google.com/p/google-web-toolkit/issues/detail?id=4438

On Apr 13, 7:27 am, Jochen Schnaidt j.schna...@t-online.de wrote:
Hi,
I posted this already at the AppEngine forum but I think this is an
interface problem so I post here as well.

I’m building an application with GWT and GAE and have a problem saving
data in the datastore. In my application I generate a list which
consists of lists of strings, looks this way: ListListString
trackList;

It compiles but gives me a warning:

[WARN] Warnings in
generated://D271BF8A9063BFFBB6E2618C9E23EF15/eventManagement/shared/HandleEventService_TypeSerializer.java'
[WARN] Line 31: Referencing deprecated class
'com.google.gwt.user.client.ui.ChangeListenerCollection'
And 3 other deprecated classes

When I run the application and call the method for saving data I get
the following error:

javax.jdo.JDOException: Error creating the MetaDataManager for API
JDO :
NestedThrowables:
java.lang.reflect.InvocationTargetException

I think the problem is that the compiler uses old GWT classes for
serialization and not the classes of the GAE. Right?

Any idea how I could fix this problem?

Thanks a lot. Greetings

--
You received this message because you are subscribed to the Google Groups
Google Web Toolkit group.
To post to this group, send email to google-web-tool...@googlegroups.com.
To unsubscribe from this group, send email to
google-web-toolkit+unsubscr...@googlegroups.com.
For more options, visit this group at
http://groups.google.com/group/google-web-toolkit?hl=en.

Re: MouseMoveEvent problem when i move my mouse very fast

You can't get every single mouse position, as browsers (indeed the OS)
don't generate an event for every single pixel the mouse passes over.
In fact this is not expected behavior in any application, including
Paint; if you move fast you'll see it has to extrapolate, as you will
as well.

On Apr 13, 2:35 am, fonghuangyee fonghuang...@gmail.com wrote:
 http://www.farbtube.com

 this is the demo, or should i use a canvas?
 Anyone know how to implement it?

 On Apr 13, 3:45 pm, fonghuangyee fonghuang...@gmail.com wrote:

  addMouseMoveHandler(new MouseMoveHandler() {

          @Override
          public void onMouseMove(MouseMoveEvent event) {
                  System.out.println(X :  + event.getX());

  });

  My problem is, if i move my mouse very fast, i cant receive every
  MouseMoveEvent.
  As the code above, i aspect:
  0,1,2,3,4,5,6

  But when i move very fast, i will get
  0, 5, 15, 20, 50.

  I am doing a simple app as Window Paint.
  I need to detect every MouseMoveEvent so that i can do drawing on the
  page.

  Thanks.

-- 
You received this message because you are subscribed to the Google Groups 
Google Web Toolkit group.
To post to this group, send email to google-web-tool...@googlegroups.com.
To unsubscribe from this group, send email to 
google-web-toolkit+unsubscr...@googlegroups.com.
For more options, visit this group at 
http://groups.google.com/group/google-web-toolkit?hl=en.

Re: File Download in GWT

2010-04-13 Thread Manuel Carrasco Moñino

I'm not sure what you mean.
I think mean you want to send information with the link (no the
iframe, because it is used to receive the response), so you have to
add the parameters to the URL of the link (query-string).

-Manolo

On Sun, Apr 11, 2010 at 3:40 PM, daaSdemahoM m.s...@tu-bs.de wrote:
 Thanks! it works great!
 But i need to send a string to the servlet over this iframe.
 it s possible to pass a string in any attribute from this iframe?

 please, how can i read the sended value??

 --
 You received this message because you are subscribed to the Google Groups 
 Google Web Toolkit group.
 To post to this group, send email to google-web-tool...@googlegroups.com.
 To unsubscribe from this group, send email to 
 google-web-toolkit+unsubscr...@googlegroups.com.
 For more options, visit this group at 
 http://groups.google.com/group/google-web-toolkit?hl=en.



-- 
You received this message because you are subscribed to the Google Groups 
Google Web Toolkit group.
To post to this group, send email to google-web-tool...@googlegroups.com.
To unsubscribe from this group, send email to 
google-web-toolkit+unsubscr...@googlegroups.com.
For more options, visit this group at 
http://groups.google.com/group/google-web-toolkit?hl=en.

Re: Serialization of lists

2010-04-13 Thread Jochen Schnaidt

Thank you very much. First problem solved.

The javax.jdo.JDOException still occurs :-(

On Apr 13, 4:06 pm, kozura koz...@gmail.com wrote:
Check out these threads, it's a bug in the RPC serializer, there's a
workaround on the second link.

http://groups.google.com/group/google-web-toolkit/browse_thread/threa...http://code.google.com/p/google-web-toolkit/issues/detail?id=4438

On Apr 13, 7:27 am, Jochen Schnaidt j.schna...@t-online.de wrote:

Hi,
I posted this already at the AppEngine forum but I think this is an
interface problem so I post here as well.

It compiles but gives me a warning:

When I run the application and call the method for saving data I get
the following error:

javax.jdo.JDOException: Error creating the MetaDataManager for API
JDO :
NestedThrowables:
java.lang.reflect.InvocationTargetException

I think the problem is that the compiler uses old GWT classes for
serialization and not the classes of the GAE. Right?

Any idea how I could fix this problem?

Thanks a lot. Greetings

--
You received this message because you are subscribed to the Google Groups
Google Web Toolkit group.
To post to this group, send email to google-web-tool...@googlegroups.com.
To unsubscribe from this group, send email to
google-web-toolkit+unsubscr...@googlegroups.com.
For more options, visit this group at
http://groups.google.com/group/google-web-toolkit?hl=en.

Re: firebug aborted *.cache.html

2010-04-13 Thread mmoossen

From time to time, i have the same problem here... right now with
the hosted.html
and i think only in dev mode, but not sure...

Michael

On Apr 12, 7:26 pm, Simon Botting simon.bott...@enbuenosaires.com
wrote:
 Hi,
 In firefox firebug a red message is showing in the net tab that the
 request for the *.cache.html file is being aborted then re-requested
 and reloaded - fresh from the server.   Here is a screen shot of
 firebug:

 http://www.enbuenosaires.com/images/firebug-aborted-cache-html.png

 my apache (v 2.2.9) config as it relates to GWT (as far as I'm aware)
 is as follows :

    # cache and compression
    Files *.cache.*
       SetOutputFilter DEFLATE
       ExpiresActive On
       ExpiresDefault modification plus 2 years
    /Files

    Files *.nocache.*
       SetOutputFilter DEFLATE
       ExpiresActive On
       ExpiresActive Off
    /Files

 Does anyone know why I'm getting these aborted messages, I imagine it
 must to be something to do with my apache.conf

 Any help greatly appreciated,

 Cheers
 Simon

-- 
You received this message because you are subscribed to the Google Groups 
Google Web Toolkit group.
To post to this group, send email to google-web-tool...@googlegroups.com.
To unsubscribe from this group, send email to 
google-web-toolkit+unsubscr...@googlegroups.com.
For more options, visit this group at 
http://groups.google.com/group/google-web-toolkit?hl=en.

Re: Non-Anonymous UIBinding without @UIField(provided=true), How?

2010-04-13 Thread Eric



On Apr 12, 2:35 pm, Brian Reilly brian.irei...@gmail.com wrote:
 There is one convenience that this appears to leave out, though:
 @UiHandler. These are things that MVP tells us should be in the
 presenter, yet the annotation has to live in the view class instead. I
 haven't fully worked out my thoughts on this yet, but having already
 done MVP without UiBinder, I don't feel burdened by not using
 @UiHandler. Some have talked about having the presenter register
 callbacks with the view that the view can then call from @UiHandler
 annotated methods, which seems cumbersome to me. Others have talked
 about letting the view hold a reference to the presenter to call
 specific methods (maybe defined in a presenter interface) from
 @UiHandler methods. This may not be a bad approach, but I haven't
 tried it myself yet.

I have a form that includes two date widgets and a Save button. One
date
represents a start date and the other represents an end date.  I
therefore
have a business requirement that the end date not precede the start
date.
I would have been willing to use two @UiHandler annotations in the
view
class to disable 'illegal' dates on each DatePicker because that's a
matter
purely internal to the panel. It turned out that I could not use
@UiHandler
for that because there are two methods in DatePicker that take a
single
ShowRangeEvent argument. At least, I couldn't do that.

On the other hand, the save button has its event handler attached
within
the presenter class.  The view should not know how it is used.

Respectfully,
Eric Jablow

-- 
You received this message because you are subscribed to the Google Groups 
Google Web Toolkit group.
To post to this group, send email to google-web-tool...@googlegroups.com.
To unsubscribe from this group, send email to 
google-web-toolkit+unsubscr...@googlegroups.com.
For more options, visit this group at 
http://groups.google.com/group/google-web-toolkit?hl=en.

Re: Best (faster) place to include gwt script in html ?

2010-04-13 Thread Óscar Frías

But this document suggests putting scripts at the bottom of the HTML
to improve page download speed:
http://developer.yahoo.com/performance/rules.html#js_bottom

Is this true also for GWT script ?

Thanks,
Oscar


On Apr 13, 3:34 pm, redlaber vyalov.e...@gmail.com wrote:
 It makes no difference. The html-code generate the page almost
 instantly. GWT metod RootPanel.get(id) also work very quickly.
 Almost all time of page generation will take the loading of gwt-
 module. So place your gwt-application where it is more convenient to
 you.

 On 13 апр, 13:51, Óscar Frías ofr...@gmail.com wrote:

  Which is the best place to put the gwt script in the HTML page from a
  performance point of view (page speed) ?
  I mean to minimize the load time of a page that contains some HTML
  code in addition to the GWT module.

  Is it better to put it at the end of the head section (immediately
  before the /head tag, or immediately before the analytics code) ?
  Or would it be faster to put it at the end of the html (immediately
  before the /html tag) ?

  Thanks!
  Oscar

-- 
You received this message because you are subscribed to the Google Groups 
Google Web Toolkit group.
To post to this group, send email to google-web-tool...@googlegroups.com.
To unsubscribe from this group, send email to 
google-web-toolkit+unsubscr...@googlegroups.com.
For more options, visit this group at 
http://groups.google.com/group/google-web-toolkit?hl=en.

Re: MouseMoveEvent problem when i move my mouse very fast

2010-04-13 Thread ben fenster

i had the same problem and i dont think it has a solution since the
events are sent from the browser and if they are wrong then theirs
nothing to do about it

On 13 אפריל, 09:35, fonghuangyee fonghuang...@gmail.com wrote:
 http://www.farbtube.com

 this is the demo, or should i use a canvas?
 Anyone know how to implement it?

 On Apr 13, 3:45 pm, fonghuangyee fonghuang...@gmail.com wrote: 
 addMouseMoveHandler(new MouseMoveHandler() {

          @Override
          public void onMouseMove(MouseMoveEvent event) {
                  System.out.println(X :  + event.getX());

  });

  My problem is, if i move my mouse very fast, i cant receive every
  MouseMoveEvent.
  As the code above, i aspect:
  0,1,2,3,4,5,6

  But when i move very fast, i will get
  0, 5, 15, 20, 50.

  I am doing a simple app as Window Paint.
  I need to detect every MouseMoveEvent so that i can do drawing on the
  page.

  Thanks.

-- 
You received this message because you are subscribed to the Google Groups 
Google Web Toolkit group.
To post to this group, send email to google-web-tool...@googlegroups.com.
To unsubscribe from this group, send email to 
google-web-toolkit+unsubscr...@googlegroups.com.
For more options, visit this group at 
http://groups.google.com/group/google-web-toolkit?hl=en.

Re: Best (faster) place to include gwt script in html ?

If you have just a simple html page that just loads the GWT app and
maybe a spinning Loading gif, then it makes no difference.  That
page is more about when you have a regular html page that happens to
also load a script, that you want the script at the bottom so all the
other content - images etc - will load first and be visible to the
user, before loading the script.

On Apr 13, 8:59 am, Óscar Frías ofr...@gmail.com wrote:
 But this document suggests putting scripts at the bottom of the HTML
 to improve page download 
 speed:http://developer.yahoo.com/performance/rules.html#js_bottom

 Is this true also for GWT script ?

 Thanks,
 Oscar

 On Apr 13, 3:34 pm, redlaber vyalov.e...@gmail.com wrote:

  It makes no difference. The html-code generate the page almost
  instantly. GWT metod RootPanel.get(id) also work very quickly.
  Almost all time of page generation will take the loading of gwt-
  module. So place your gwt-application where it is more convenient to
  you.

  On 13 апр, 13:51, Óscar Frías ofr...@gmail.com wrote:

   Which is the best place to put the gwt script in the HTML page from a
   performance point of view (page speed) ?
   I mean to minimize the load time of a page that contains some HTML
   code in addition to the GWT module.

   Is it better to put it at the end of the head section (immediately
   before the /head tag, or immediately before the analytics code) ?
   Or would it be faster to put it at the end of the html (immediately
   before the /html tag) ?

   Thanks!
   Oscar

-- 
You received this message because you are subscribed to the Google Groups 
Google Web Toolkit group.
To post to this group, send email to google-web-tool...@googlegroups.com.
To unsubscribe from this group, send email to 
google-web-toolkit+unsubscr...@googlegroups.com.
For more options, visit this group at 
http://groups.google.com/group/google-web-toolkit?hl=en.

Re: Extending the 'Contact: MVP' example

2010-04-13 Thread Sean C.

The problem is that only one presenter can be active at a given time.
If a presenter is controlling the left, right and center column, even
handing the center column container to another presenter will not
result in the original presenter giving up control or you get the
effect of one view being drawn followed by the second view quickly
replacing it.

I ended up implementing the left and right columns as pure HTML+CSS
and used JSNI or the History mechanism to trigger the appropriate View
+Presenter to affect the center column. This way no presenter is
controlling the left and right columns.

Sean C.

On Apr 12, 2:13 pm, Brian Reilly brian.irei...@gmail.com wrote:
 I think you should be able to do this pretty easily by putting
 something that implements HasWidgets (such as SimplePanel, FlowPanel,
 etc.) in your center column. You can then pass that container into
 your presenter as the place where its content should live. That should
 leave the other things that you've added to the RootPanel in place
 when you switch presenters for the center column.

 -Brian

 On Apr 10, 3:24 pm, Sean C. sean.cra...@gmail.com wrote:

  Having a similar problem with creating Composite Views backed by their
  respective Presenters.

  The bug in DockPanelLayoutPanel for Java 5 does not help and I don't
  have the option to switch to Java 6 but that is a separate 
  discussion:http://code.google.com/p/google-web-toolkit/issues/detail?id=4254

  Back to the problem: the Home Page is composed of three columns with
  the East and West columns being static HTML + CSS as well as some
  dynamic content (Vertical Panels), that affects the content of the
  Center column. Yes, JSNI is being used here.

  E.g: clicking the Login button in the West column presents a Login
  form in the Center column. The Home Page is one View + Presenter but
  the Login form is a separate View + Presenter. They do not play
  together and the Home View + Presenter always override the other Views
  + Presenters. Even tried placing separate DIV hooks in the HTML page
  to get the different views to attach to different nodes in the DOM but
  to no avail.

  Several options that are being tried and tested:
  1. Play directly with the DOM (class) to swap widgets but then you
  just end up having one View + Presenter and a lot of widgets.
  2. Creating a parent class for all Views and Presenters that contains
  the logic and content for the East and West Panels and then attach the
  new Views directly to the RootPanel. This is defeating the purpose of
  AJAX as it is equivalent to redrawing the page with the exception of
  the static parts.
  3. Go back to the drawing board by looking at the options 
  available:http://martinfowler.com/eaaDev/uiArchs.html

  Derek Greer has something that might 
  inspire:http://www.aspiringcraftsman.com/2007/08/interactive-application-arch...

  A .NET centric post about 
  this:http://bradwilson.typepad.com/blog/2008/06/composite-views.html

  Dolphin SmallTalk based discussion on Composite 
  MVP:http://www.mimuw.edu.pl/~sl/teaching/00_01/Delfin_EC/Patterns/Composi

  Sean C.

  On Mar 18, 1:29 pm, Fabio Kaminski fabiokamin...@gmail.com wrote:

   As far as i understand you have windows that contain widgets... like:

   MainWidgetContainer (which owns) LoginWidget, HeaderWidget and so on..
   LogoutWidgetContainer (owns) LogoutWidget, ComeBackSoonMessageWidget, 
   etc..

   if is that the case, you would only call the go() which means
   RootPanel.get().add()
   when you wanna change from one widget container to another...
   MainWidgetContainer - LogoutWidgetContainer
   not LoginWidget  HeaderWidget or LogoutWidget...

   as soon you establish this rule (you can even restrict go() method to
   allowing only widget containers)
   if you dont do it yet.. the only thing you need ins to assemble the
   Containers.. like:

   public class MainView extends Composite implements Display {

      final LoginView loginView;
      final HeaderView headerView;

     (add those to this widget)
     (...)

   }

   Fabio Kaminski

   On Thu, Mar 18, 2010 at 2:06 PM, Brian Reilly 
   brian.irei...@gmail.comwrote:

It seems to me like you need a supervising presenter to coordinate
interactions among each of the panes of the application, in a
DockPanel as you suggest. The AppController would then bootstrap this
instead of managing what is shown in the root panel. Finally, the
logic that is in the AppController of the Contacts example would need
to be pushed down into this presenter or something that sits between
that and the contacts widget.

Doing this, though, removes a lot of responsibility from the
AppController. I don't know that the AppController is a bad idea. It
may be useful for a simple example. However, I think it would be more
useful to have a more self-contained contacts widget that could be
reused without having to worry about the AppController.

I haven't tried any

HTML5 Video in Mobile Safari

2010-04-13 Thread KevMo

'm trying to code a site in GWT that plays videos with HTML5.
Everything works great on the desktop, but mobile Safari on both the
iPhone and iPad do not play the video.

I can play a video using Video for Everybody. I've even copied the
code to my own plain HTML page, and it works flawlessly. If I serve
that same code via a GWT widget, mobile safari will not play the
video. On the iPhone I see a gray box with a prohibitory sign around
the play button, and on the iPad it shows up as a black box.

I've made sure my doctype is !DOCTYPE html, but I don't know where
else to start debugging. Perhaps it it because the code is injected
via javascript? Any pointers on where to start looking would be
greatly appreciated.

Here is the exact code I am using for the video:

!-- Video For Everybody by Kroc Camen. see camendesign.com/code/
video_for_everybody for documented code
 
===
--
video width=640 height=360 poster=poster.jpg controls autoplay
source src=http://clips.vorwaerts-gmbh.de/big_buck_bunny.mp4;
type=video/mp4/source
source src=http://clips.vorwaerts-gmbh.de/big_buck_bunny.ogv;
type=video/ogg/source
!--[if gt IE 6]
object width=640 height=375 classid=clsid:02BF25D5-8C17-4B23-
BC80-D3488ABDDC6B!
[endif]--!--[if !IE]!--
object width=640 height=375 type=video/quicktime
data=http://clips.vorwaerts-gmbh.de/big_buck_bunny.mp4;
!--![endif]--
param name=src value=http://clips.vorwaerts-gmbh.de/
big_buck_bunny.mp4 /
param name=autoplay value=true /
param name=showlogo value=false /
object width=640 height=384 type=application/x-shockwave-
flash
data=player.swf?
autostart=trueamp;image=poster.jpgamp;file=http://clips.vorwaerts-
gmbh.de/big_buck_bunny.mp4
param name=movie value=player.swf?
autostart=trueamp;image=poster.jpgamp;file=http://clips.vorwaerts-
gmbh.de/big_buck_bunny.mp4 /
!-- fallback image --
img src=poster.jpg width=640 height=360 alt=Big Buck
Bunny
 title=No video playback capabilities, please download
the video below /
/object!--[if gt IE 6]!--
/object!--![endif]--
/video

-- 
You received this message because you are subscribed to the Google Groups 
Google Web Toolkit group.
To post to this group, send email to google-web-tool...@googlegroups.com.
To unsubscribe from this group, send email to 
google-web-toolkit+unsubscr...@googlegroups.com.
For more options, visit this group at 
http://groups.google.com/group/google-web-toolkit?hl=en.

Re: Multi-tenancy in GWT

2010-04-13 Thread Muz

Kyle, I built a similar architecture, but with only one instance of
the app serving all users. When a user authenticates, I look them up
in a global store to figure out which tenant they belong to, and then
use their tenant-specific database for all data persistence for the
session. So all visitors use one instance of the app but each tenant
has their own isolated database...multi-tenancy.

The problem is, true multi-tenancy isn't supported by GAE...they've
been talking about it for almost 2 years now, but it doesn't seem to
be a priority for them. They want you to use a different domain for
every tenant, which means multiple app instances (multiple versions of
your app to deploy  maintain). It also means your customer/tenant has
to be in the loop for all domain administration tasks, which to me
kind of defeats the purpose of customer-focused cloud computing.

I set my multi-tenant app up on Amazon EC2, using 1 tomcat server and
6 mysql databases (6 tenants...each tenant has on average 10 users).
The beauty is if I need to scale or load balance the app, add more
tenants/database instances or just backup one tenant's data, this is
pretty much just a mouse-click or 6 in the EC2 developer console.

Murray

On Mar 25, 8:15 am, Kyle Baley kyle.ba...@gmail.com wrote:
 Is that possible with GAE?

 In any case, I found something that works. I created a JSP page and
 dumped the contents of the MyApp.html into it. Then added the
 following to my web.xml:

         servlet
                 servlet-nametest/servlet-name
                 jsp-file/MyApp.jsp/jsp-file
         /servlet

         servlet-mapping
                 servlet-nametest/servlet-name
                 url-pattern/myApp/clients/*/url-pattern
         /servlet-mapping

 I added /clients in there for now because we're using gwt-dispatch and
 I haven't overridden its default of submitting things to /myApp/
 dispatch.

 Still some kinks to work out (like resource references) but early
 results are encouraging.

 On Mar 24, 10:54 am, Uros Trebec uros.tre...@gmail.com wrote:

  I think the best way to do this is to deploy new instance of client
  application for every customer and have one instance of server
  application that all client applications work with to get their data,
  non-client-side business logic, etc.

  We're doing it this way when we have different versions of client
  application, which tends to change more often than server one. Also
  works when you have a development version of client that you need to
  test with the same server backend.

  Regards,
  Uros

  On Mar 24, 2:24 am, Kyle Baley kyle.ba...@gmail.com wrote:

   Not sure I understand. I plan to filter data on the server side of
   things. But I don't know how to configure GWT to allow me to navigate
   to specific client for the entire app.

   For example, Acme Hair Salon would navigate to mysite.com/
   acmeHairSalon/myapp.html and Generic Dog Collars would navigate to
   mysite.com/genericDogCollars/myapp.html. Both URLs should map to the
   same instance of the application but in it, I'd check the URL to see
   which company was being used and filter all the data in it
   accordingly. I know how to do the filtering and can probably figure
   out how to check the URL for a company token. But I don't know how to
   configure GWT and/or GAE to use a single instance of the application
   for multiple (generic) URLs that aren't pre-defined. It sounds to me
   like something that should be done in web.xml but again, I don't know
   how.

   On Mar 23, 1:01 pm, dolcra...@gmail.com dolcra...@gmail.com wrote:

So you can just built the ui once and always include it in the login
page for which ever client as the server should be where you filter/
prevent access to data that's not for the current client.

On Mar 22, 10:41 am, Kyle Baley kyle.ba...@gmail.com wrote:

 I'm new to GWT and the Java world having been in the .NET space for
 about 10 years. Our application will be a multi-tenant one (at least,
 to the degree that I understand the definition). We'd like customers
 to be able to navigate towww.mysite.com/customerName, then log in
 from there. I think I have a handle on how to manage this from the GAE
 datastore side but am wondering how to manage this URL mapping in
 web.xml for GWT.

 Also, would like to get some opinions on how to manage the
 authentication once it's set up. Here's how I see it working:
 - User navigates towww.mysite.com/customerName
 - System retrieves company data and stores in session and displays
 login page
 - User logs in. System authenticates user
 - For all RPC calls, system verifies user is logged in *and* checks
 that the URL matches the company info in the session

 Basically, I want to guard against a user logging into one client
 site, then navigating to another.

 Thanks
 Kyle



-- 
You received this message because you are subscribed to the Google

An implementation of Selection/Cursor for RichTextArea

So I've written a complete selection/range implementation, which
allows much more sophisticated Rich Text Editor operations.  You can
obtain and manipulate the cursor or current selected range - with
this, I was able to trivially create a proper Link functionality to
deal with existing links in the editor.  Inspired from the partial
solution in gwt-rocket, it uses a structural selection model like the
W3C spec, with emulation to make it work the same in cursed IE*.  And
except for wrappers around js calls, everything is written in Java to
allow full GWT optimization.

I really think this functionality should be part of the GWT library,
as I think it's the main missing core feature needed to implement full
RTE capabilities without having to go import external JS RTE
libraries.  I could always post it as a separate library like gwt-
rocket, but it's already fairly annoying the number of these little
libraries with random bits of functionality, which quickly become
unsupported or bloated.  Additionally, due to an IE bug it would be
best to have a getSelection/getRange method directly as part of the
RichTextArea widget.

But that's just me, do others think this is a really need-to-have for
GWT?  If others feel it'd be a good addition to GWT, I'd be willing to
spend the time to try to get it in, maybe starting with getting it in
the incubator.  And through that would love to see any critiques or
better ways to do any of it.


Here's the API, purely structural and doesn't include any of the
position-based methods that IE has:


// Current selection is based on a document/window, so there is a
browser
// global selection, and a selection for any iframe windows.
class Selection
{
// Get the browser global selection
static Selection getSelection();

// Get selection associated with a particular window, ie an RTE
iframe
static Selection getSelection(JavaScriptObject window);

// Get a range containing current selection.  Changes to the range
do not
// affect the selection, and vice versa
Range getRange();

// Set the selection to what is contained in the range.
Subsequent
// changes to this range object will not affect the selection
void setRange(Range range);

void clear();
Document getDocument()
}

// Represents a range in an html document, as a start and end text
node,
// and the offset in to them, ie between the two | marks here:
//
// bA r|ange athat spans/a/bi some elem|ents./i
class Range
{
// Constructors
Range(Document doc);
Range(Element element);
Range(RangeEndPoint cursorPoint);
Range(RangeEndPoint startPoint, RangeEndPoint endPoint);

// Get the start/end points of this range, or the cursor if no
selection
RangeEndPoint getCursor();
RangeEndPoint getStart();
RangeEndPoint getEnd();

// Get the document this range is contained within
Document getDocument();

// Set the range to be a cursor, ie start == end
void setCursor(RangeEndPoint cursorPoint);

// Set the range to surround the given element
void setRange(Element element);

// Set the endpoints of the range
void setRange(RangeEndPoint startPoint, RangeEndPoint endPoint);

// Set the start or end point of the range
void setStartPoint(RangeEndPoint startPoint);
void setEndPoint(RangeEndPoint endPoint);


// Returns true if start == end
boolean isCursor();

// Collapses the range to a cursor, either to the start or end
point
void collapse(boolean start);

// Compare the start or end point of this range with the start or
end point
// of another.  how determines which endpoints, ie
Range.START_TO_END
int compareBoundaryPoint(Range compare, short how);

// Get just the text selected, ie ange that spans some elem
String getText();

// Get the html selected, including any tags needed to make it
complete, ie:
// bange athat spans/a/bi some elemi
String getHtmlText();

// Returns the lowest level element that completely contains this
range
Element getCommonAncestor();


// Returns an ordered list of all the text nodes in this range
ListText getSelectedTextNodes();

// Copy the contents of this range into the given element, leaving
the
// document undisturbed
void copyContents(Element copyInto);

// Delete the contents of this range from the document
void deleteContents();

// Extracts the contents of this range into the given element,
removing them
// from the document
void extractContents(Element copyInto);
Element extractContents(); // same but creates a span element for
you

// Surrounds the contents of this range with the given element,
putting it
// in their place in the document and setting the range to contain
it.
void surroundContents(Element copyInto);
Element surroundContents(); // same but creates a span element for
you


// A static utility function that should be part of Text or Node
class
//

I'm getting error The web.xml file does not exist

2010-04-13 Thread riahut.com

Obviously because I moved it from war folder and deleted it. But how
to tell Google Web App that web.xml is in different folder?

I use Eclipse 3.5
GWT 2.0.2
App Engine 1.3.1

Deploying Under Tomcat 6
JDK 1.5

-- 
You received this message because you are subscribed to the Google Groups 
Google Web Toolkit group.
To post to this group, send email to google-web-tool...@googlegroups.com.
To unsubscribe from this group, send email to 
google-web-toolkit+unsubscr...@googlegroups.com.
For more options, visit this group at 
http://groups.google.com/group/google-web-toolkit?hl=en.

Re: MouseMoveEvent problem when i move my mouse very fast

2010-04-13 Thread fonghuangyee

but when i refer the http://www.farbtube.com and http://canvaspaint.org,
event i move very fast, they browser still can capture every single
pixel,
so i believe that this is GWT MouseEvent behavior or limitation.


On Apr 13, 11:02 pm, ben fenster fenster@gmail.com wrote:
 i had the same problem and i dont think it has a solution since the
 events are sent from the browser and if they are wrong then theirs
 nothing to do about it

 On 13 אפריל, 09:35, fonghuangyee fonghuang...@gmail.com wrote:



 http://www.farbtube.com

  this is the demo, or should i use a canvas?
  Anyone know how to implement it?

  On Apr 13, 3:45 pm, fonghuangyee fonghuang...@gmail.com wrote: 
  addMouseMoveHandler(new MouseMoveHandler() {

           @Override
           public void onMouseMove(MouseMoveEvent event) {
                   System.out.println(X :  + event.getX());

   });

   My problem is, if i move my mouse very fast, i cant receive every
   MouseMoveEvent.
   As the code above, i aspect:
   0,1,2,3,4,5,6

   But when i move very fast, i will get
   0, 5, 15, 20, 50.

   I am doing a simple app as Window Paint.
   I need to detect every MouseMoveEvent so that i can do drawing on the
   page.

   Thanks.

-- 
You received this message because you are subscribed to the Google Groups 
Google Web Toolkit group.
To post to this group, send email to google-web-tool...@googlegroups.com.
To unsubscribe from this group, send email to 
google-web-toolkit+unsubscr...@googlegroups.com.
For more options, visit this group at 
http://groups.google.com/group/google-web-toolkit?hl=en.

Re: MouseMoveEvent problem when i move my mouse very fast

I can guarantee you they don't, the OS doesn't even provide that.  If
you move fast in a curve, you'll notice there's straight segments
creating the curve.  Thus they extrapolate.  Smarter paint programs
actually create a spline of the points to make it look better, but
they still have to work with non-contiguous mouse points.

On Apr 13, 10:40 am, fonghuangyee fonghuang...@gmail.com wrote:
 but when i refer thehttp://www.farbtube.comandhttp://canvaspaint.org,
 event i move very fast, they browser still can capture every single
 pixel,
 so i believe that this is GWT MouseEvent behavior or limitation.

 On Apr 13, 11:02 pm, ben fenster fenster@gmail.com wrote:

  i had the same problem and i dont think it has a solution since the
  events are sent from the browser and if they are wrong then theirs
  nothing to do about it

  On 13 אפריל, 09:35, fonghuangyee fonghuang...@gmail.com wrote:

  http://www.farbtube.com

   this is the demo, or should i use a canvas?
   Anyone know how to implement it?

   On Apr 13, 3:45 pm, fonghuangyee fonghuang...@gmail.com wrote: 
   addMouseMoveHandler(new MouseMoveHandler() {

        @Override
        public void onMouseMove(MouseMoveEvent event) {
                System.out.println(X :  + event.getX());

});

My problem is, if i move my mouse very fast, i cant receive every
MouseMoveEvent.
As the code above, i aspect:
0,1,2,3,4,5,6

But when i move very fast, i will get
0, 5, 15, 20, 50.

I am doing a simple app as Window Paint.
I need to detect every MouseMoveEvent so that i can do drawing on the
page.

Thanks.

-- 
You received this message because you are subscribed to the Google Groups 
Google Web Toolkit group.
To post to this group, send email to google-web-tool...@googlegroups.com.
To unsubscribe from this group, send email to 
google-web-toolkit+unsubscr...@googlegroups.com.
For more options, visit this group at 
http://groups.google.com/group/google-web-toolkit?hl=en.

Re: Multi-tenancy in GWT


 In any case, I found something that works. I created a JSP page and
 dumped the contents of the MyApp.html into it. Then added the
 following to my web.xml:

servlet
servlet-nametest/servlet-name
jsp-file/MyApp.jsp/jsp-file
/servlet
servlet-mapping
servlet-nametest/servlet-name
url-pattern/myApp/clients/*/url-pattern
/servlet-mapping

 I added /clients in there for now because we're using gwt-dispatch and
 I haven't overridden its default of submitting things to /myApp/
 dispatch.


That is the best way to move forward. Since its unlikely that one client
will access the website for another, you don't want the /#customer-name kind
of urls. Also, in future, if you want to support a separate domain for each
client, it would be easily doable.


-Sri
http://blog.530geeks.com


On 13 April 2010 21:23, Muz murray.bo...@gmail.com wrote:

 Kyle, I built a similar architecture, but with only one instance of
 the app serving all users. When a user authenticates, I look them up
 in a global store to figure out which tenant they belong to, and then
 use their tenant-specific database for all data persistence for the
 session. So all visitors use one instance of the app but each tenant
 has their own isolated database...multi-tenancy.

 The problem is, true multi-tenancy isn't supported by GAE...they've
 been talking about it for almost 2 years now, but it doesn't seem to
 be a priority for them. They want you to use a different domain for
 every tenant, which means multiple app instances (multiple versions of
 your app to deploy  maintain). It also means your customer/tenant has
 to be in the loop for all domain administration tasks, which to me
 kind of defeats the purpose of customer-focused cloud computing.

 I set my multi-tenant app up on Amazon EC2, using 1 tomcat server and
 6 mysql databases (6 tenants...each tenant has on average 10 users).
 The beauty is if I need to scale or load balance the app, add more
 tenants/database instances or just backup one tenant's data, this is
 pretty much just a mouse-click or 6 in the EC2 developer console.

 Murray

 On Mar 25, 8:15 am, Kyle Baley kyle.ba...@gmail.com wrote:
  Is that possible with GAE?
 
  In any case, I found something that works. I created a JSP page and
  dumped the contents of the MyApp.html into it. Then added the
  following to my web.xml:
 
  servlet
  servlet-nametest/servlet-name
  jsp-file/MyApp.jsp/jsp-file
  /servlet
 
  servlet-mapping
  servlet-nametest/servlet-name
  url-pattern/myApp/clients/*/url-pattern
  /servlet-mapping
 
  I added /clients in there for now because we're using gwt-dispatch and
  I haven't overridden its default of submitting things to /myApp/
  dispatch.
 
  Still some kinks to work out (like resource references) but early
  results are encouraging.
 
  On Mar 24, 10:54 am, Uros Trebec uros.tre...@gmail.com wrote:
 
   I think the best way to do this is to deploy new instance of client
   application for every customer and have one instance of server
   application that all client applications work with to get their data,
   non-client-side business logic, etc.
 
   We're doing it this way when we have different versions of client
   application, which tends to change more often than server one. Also
   works when you have a development version of client that you need to
   test with the same server backend.
 
   Regards,
   Uros
 
   On Mar 24, 2:24 am, Kyle Baley kyle.ba...@gmail.com wrote:
 
Not sure I understand. I plan to filter data on the server side of
things. But I don't know how to configure GWT to allow me to navigate
to specific client for the entire app.
 
For example, Acme Hair Salon would navigate to mysite.com/
acmeHairSalon/myapp.html and Generic Dog Collars would navigate to
mysite.com/genericDogCollars/myapp.html. Both URLs should map to the
same instance of the application but in it, I'd check the URL to see
which company was being used and filter all the data in it
accordingly. I know how to do the filtering and can probably figure
out how to check the URL for a company token. But I don't know how to
configure GWT and/or GAE to use a single instance of the application
for multiple (generic) URLs that aren't pre-defined. It sounds to me
like something that should be done in web.xml but again, I don't know
how.
 
On Mar 23, 1:01 pm, dolcra...@gmail.com dolcra...@gmail.com
 wrote:
 
 So you can just built the ui once and always include it in the
 login
 page for which ever client as the server should be where you
 filter/
 prevent access to data that's not for the current client.
 
 On Mar 22, 10:41 am, Kyle Baley kyle.ba...@gmail.com wrote:
 
  I'm new to GWT and the Java world having been in the .NET space
 for
  about 10

Re: Need to explicitly remove event handlers manually?

2010-04-13 Thread Brian Reilly

I'd say that cleaning up no longer needed handlers is just as
important as closing open files when you're done or freeing memory
that you've allocated. You won't notice the effects at first, but
eventually you will. How soon depends on the cases under which your
application creates handlers). Eventually, you'll see event delivery
taking longer and longer because your event bus is busy notifying
handlers that you don't care about anymore.

There's a reason why the handler manager goes to the trouble of
returning a handler registration. There's no way for it to know that
you're done with the handler. This is one of those cases where garbage
collection wouldn't even help you because the handler manager has to
hold a reference to the handler (unless you use weak references, but
I'm pretty sure we don't have those in the JDK emulation layer). 

It's not too difficult to keep track of handler registrations. I often
make a collection to hold them. Just add them all to the collection in
your bind method and unregister everything (and clear the collection)
in your unbind method.

-Brian

On Apr 12, 8:14 pm, rjcarr rjc...@gmail.com wrote:
 If you no longer need the handler then yeah, you should remove it,
 because you don't want to execute code blocks (onClick()) that are no
 longer needed or used.

 But it certainly isn't required.

 On Apr 12, 2:32 pm, macagain rgk...@gmail.com wrote:



  Is it required or even good practice to manually remove event handlers
  (say, on detach or unload)?  I.e. keep the handlerRegistration around
  when returned from addHandler, then later call
  handlerManager.removeHandler().

  Of course in the old world of manually managing listener collections,
  one had to do it, but in the new world of HandlerManagers?

-- 
You received this message because you are subscribed to the Google Groups 
Google Web Toolkit group.
To post to this group, send email to google-web-tool...@googlegroups.com.
To unsubscribe from this group, send email to 
google-web-toolkit+unsubscr...@googlegroups.com.
For more options, visit this group at 
http://groups.google.com/group/google-web-toolkit?hl=en.

Re: GWT calendar

2010-04-13 Thread Manuel Carrasco Moñino

Take a look to an article I wrote time ago about this subject:
http://code.google.com/p/gwtchismes/wiki/Tutorial_ExportingGwtLibrariesToJavascript_en

Cheers
-Manolo

On Tue, Apr 13, 2010 at 9:59 AM, Naveen r.naveens...@gmail.com wrote:
 I am still waiting for some reply on this...
 Any one who has integrated GWT component with a normal project can
 share what you did..

 Thanks

 On Apr 8, 8:15 pm, Naveen r.naveens...@gmail.com wrote:
 Hi

 I need to implementcalendarfunctionality to an already existing 
 nonGWTproject. I would like to know if it is possible to integrate 
 theGWTcalendarto my nonGWTproject. I am aware that we need to convert
 the java code into java script, so my question is can i run this as an
 separateGWTproject and then implement that java script in my
 project.. Plz let me know if this will work also any other info will
 be of great help..

 Thanks

 --
 You received this message because you are subscribed to the Google Groups 
 Google Web Toolkit group.
 To post to this group, send email to google-web-tool...@googlegroups.com.
 To unsubscribe from this group, send email to 
 google-web-toolkit+unsubscr...@googlegroups.com.
 For more options, visit this group at 
 http://groups.google.com/group/google-web-toolkit?hl=en.



-- 
You received this message because you are subscribed to the Google Groups 
Google Web Toolkit group.
To post to this group, send email to google-web-tool...@googlegroups.com.
To unsubscribe from this group, send email to 
google-web-toolkit+unsubscr...@googlegroups.com.
For more options, visit this group at 
http://groups.google.com/group/google-web-toolkit?hl=en.

Deploy GWT 2.0 in a jetty

2010-04-13 Thread Julian!

Hi
im trying to deploy my GWT app in a jetty servlet container, im
developing with eclipse with the google plugin. and when i run with
the plugin http://127.0.0.1:/index.html?gwt.codesvr=127.0.0.1:9997,
works fine. but when I copy the war directory and configure to jetty
server the application doesnt work.

Is a javascript error:
Error: $doc.getElementById(Hsc) is null
Archivo de origen: 
http://127.0.0.1:/planviajegwt/4A217B0203D1C3793F30E5DE0A69D1CC.cache.html
Línea: 3601

function XY(){var a;!!$stats$stats({moduleName:$moduleName,sessionId:
$sessionId,subSystem:Dsc,evtGroup:Esc,millis:(new
Date).getTime(),type:Fsc,className:Gsc});parseInt($doc.getElementById(Hsc)
[zrc])||0;a=UTb(new STb);yyb(PNb(Isc),a)}

Some idea?


Thanks !!!

-- 
You received this message because you are subscribed to the Google Groups 
Google Web Toolkit group.
To post to this group, send email to google-web-tool...@googlegroups.com.
To unsubscribe from this group, send email to 
google-web-toolkit+unsubscr...@googlegroups.com.
For more options, visit this group at 
http://groups.google.com/group/google-web-toolkit?hl=en.

Static and dynamic loading of images

2010-04-13 Thread Juan Quito

Hello, my application has some standard icons like save, new, edit,
etc... but the user can customize the icons and specify the images he
wants the application to display at runtime. How can I accomplish this
using ClientBundle interface (the standard images would be specified
in the extended interface) and in runtime the system gets the custom
image instead of the standard one by loading from an URL?
I don't mind getting all the standard icons even if they won't be used
at all. Today my system loads all images by the URL, so the startup of
the application is too slow and most of the images are standard.

I would like to do something like this:

Image w = new Image(ResourceLoader.INSTANCE.background());

ResourceLoader looks if there is a customized image for background,
and if there is... loads from an URL, if not... calls the Resources
interface loading the standard one.

interface Resources extends ClientBundle {
Resources INSTANCE = GWT.create(Resources.class);
@Source(images/background.png)
ImageResource background();
}



Thanks!
Juan José Quito

-- 
You received this message because you are subscribed to the Google Groups 
Google Web Toolkit group.
To post to this group, send email to google-web-tool...@googlegroups.com.
To unsubscribe from this group, send email to 
google-web-toolkit+unsubscr...@googlegroups.com.
For more options, visit this group at 
http://groups.google.com/group/google-web-toolkit?hl=en.

Re: __listener tags getting added in IE8 using GWT 2.0

2010-04-13 Thread Brandon Turner

Usually the attribute is so long (from my brief inspection it looked like
the elements innerHtml) that the IE8 developer tool is just bugged out from
it. At least that is what I always assumed.

-Brandon

On Mon, Apr 12, 2010 at 4:55 PM, myapplicationquestions
parag.bhag...@cgi.com wrote:

Hi All,

I am having a very hard time understanding why each of my div when
viewed in IE8 ( using its crappy developer tools) show me __listener
tag which is incomplete..

div class=portlet-column style=height: 100%; __listener=
DIV style=HEIGHT: 100% class=portlet-columnDIV class=mss-
DockableWidget
DIV class=mss-PortletContainer __eventBits=1

Has anyone faced this issue with IE8 and GWT 2.0?

Thanks,

--
You received this message because you are subscribed to the Google Groups
Google Web Toolkit group.
To post to this group, send email to google-web-tool...@googlegroups.com.
To unsubscribe from this group, send email to
google-web-toolkit+unsubscr...@googlegroups.com.
For more options, visit this group at
http://groups.google.com/group/google-web-toolkit?hl=en.

TabLayoutPanel at 100% height

2010-04-13 Thread Jérôme

Hi there,

I'm trying to use the newly (version 2.0) introduced TabLayoutPanel.
What I also want is that the embedded widget in each panel's tab has a
height of 100% (take the whole vertical place). In order to do so I
just made a very simple example (no CSS, no customisation, no noting)
that look like that:

public void onModuleLoad() {
TabLayoutPanel p = new TabLayoutPanel(30, Unit.PX);
// do not work
// p.setHeight(100%);
// work
p.setHeight(700px);
p.add(new HTML(this), [this]);
p.add(new HTML(that), [that]);
p.add(new HTML(the other), [the other]);

// Add the nameField and sendButton to the RootPanel
// Use RootPanel.get() to get the entire body element
RootPanel.get().add(p);
}

The html is trivial and only contains a div id=main/div. As you
can guess I have a problem. When I say that the height should be 100%
I see nothing. I can click on the various tabs no panel appears /
changes. But when I say that the height should be XXXpx (any height
will do the trick) then it works like a charm.

Of course what I really want is the max height available in the
browser window that is a 100%.

Did any one already succeed in doing such a thing? What am I missing?

Thanks for helping.

CU Jerome

-- 
You received this message because you are subscribed to the Google Groups 
Google Web Toolkit group.
To post to this group, send email to google-web-tool...@googlegroups.com.
To unsubscribe from this group, send email to 
google-web-toolkit+unsubscr...@googlegroups.com.
For more options, visit this group at 
http://groups.google.com/group/google-web-toolkit?hl=en.

UIBinder and CSS resources obfuscation

2010-04-13 Thread david.herv...@gmail.com

Hi,
I'm currently searching around the web to found out how to tell
UIBinder not to perform obfuscation (at-all) on css class. I've read
codes samples in which it is tell to put inside the .gwt.xml the
property CssResource.obfuscationPrefix  to empty. But the css class
name doesn't have any links with original name. I've tried to combined
with the CssResource.style to pretty but the css class are still
prefixed with the package name in which the css resources are located.

Prefixing each class with @external in css is some kind of solution,
but doing it for all class in all css I want to test is not a real
solution.

So, my question is : Is there any options for telling UIBinder not to
obfuscate css class ? (like a global @external)

Thanks in advance

-- 
You received this message because you are subscribed to the Google Groups 
Google Web Toolkit group.
To post to this group, send email to google-web-tool...@googlegroups.com.
To unsubscribe from this group, send email to 
google-web-toolkit+unsubscr...@googlegroups.com.
For more options, visit this group at 
http://groups.google.com/group/google-web-toolkit?hl=en.

Re: How to scroll page to element

2010-04-13 Thread Brandon Turner

You can get away without a ScrollPanel if you want to as well.  Something
like:

verticalPanel.getElement().setScrollTop(secondFlexTable.getElement().getAbsoluteTop());

-Brandon

On Mon, Apr 12, 2010 at 12:12 PM, t.dave da...@lorgeousdays.com wrote:

 use ScrollPanel.setScrollPosition().  that sets the vertical scroll
 position of the scrollpanel, which you will need to calculate.  try
 something like this:

 new ClickHandler() {
  public void onClick( ClickEvent event ) {
int tableTop = secondFlexTable.getAbsoluteTop();
scrollPanel.setScrollPosition( tableTop );
  }
 }

 i suppose in your case it wouldn't be a click handler if you're basing
 it off the history token, but hopefully this will point you in the
 right direction.


 On Apr 12, 7:34 am, redlaber vyalov.e...@gmail.com wrote:
  I want to scroll page, generated with gwt, to some element. Its should
  be simple, but i cannt find the solution. For example: I have a
  vertical panel with two flexables. When i get the history token
  goto2 I want to scroll my page to the second table. (Sorry for the
  terrible english).

 --
 You received this message because you are subscribed to the Google Groups
 Google Web Toolkit group.
 To post to this group, send email to google-web-tool...@googlegroups.com.
 To unsubscribe from this group, send email to
 google-web-toolkit+unsubscr...@googlegroups.comgoogle-web-toolkit%2bunsubscr...@googlegroups.com
 .
 For more options, visit this group at
 http://groups.google.com/group/google-web-toolkit?hl=en.



-- 
You received this message because you are subscribed to the Google Groups 
Google Web Toolkit group.
To post to this group, send email to google-web-tool...@googlegroups.com.
To unsubscribe from this group, send email to 
google-web-toolkit+unsubscr...@googlegroups.com.
For more options, visit this group at 
http://groups.google.com/group/google-web-toolkit?hl=en.

Re: Gwt 2 dev mode, refresh time very slow, any hints ?

2010-04-13 Thread dparish

I'm seeing this as well.  When I switch from embedded jetty to
noserver, everything is MUCH slower, despite being all local on the
same machine with PLENTY of RAM and CPU to spare.  It's bad enough
that development time is severely impacted.

On Apr 1, 6:11 am, Skal pasvinc...@gmail.com wrote:
 Hi all,

 We are developping an application with many panels and remote services
 (about an hundred) and we are facing a very annoying issue with the
 oophm dev mode, the browser refresh time takes up 30 000 to 90 000 ms.
 The -Dgwt.perflog jvm option shows that most of the time is spent in
 the Oophm.SessionHandler.loadModule call (all the app is in a single
 module). Also, we don't use the embedded jetty server (-noServer, we
 use an another jetty server).

 I notice that in dev mode, *.gwt.rpc and *.cache.png files are still
 generated in the war directory after each refresh, even if the code
 modification doesn't have an impact on this ressources (for example
 just a label change), why that ?

 I've heard about a rpc.blacklist property config extension, does it
 make sense to use this property to block some remote service
 compilation thus accelerating the refresh time ?

 Thanks a lot for any clue.

-- 
You received this message because you are subscribed to the Google Groups 
Google Web Toolkit group.
To post to this group, send email to google-web-tool...@googlegroups.com.
To unsubscribe from this group, send email to 
google-web-toolkit+unsubscr...@googlegroups.com.
For more options, visit this group at 
http://groups.google.com/group/google-web-toolkit?hl=en.

Can we Integrate reCaptcha with GWT FormPanel

2010-04-13 Thread GWT Groups

Hello Guys,
 First of i want to say thanks to all you for help me
a lot. THANKS ONES AGAIN.

Now i have got stuck in an another  problem,

I want to reCaptcha into my GWT FormPanel, I don't know how to do
this. I have done lot of efforts but i fail to do this.

I succeed to run reCaptcha in GWT Project by  configure the HTML file,
but this reCaptcha not integrate with the GWT FormPanel.



Please Help me ones again

your  friend
Ankit

-- 
You received this message because you are subscribed to the Google Groups 
Google Web Toolkit group.
To post to this group, send email to google-web-tool...@googlegroups.com.
To unsubscribe from this group, send email to 
google-web-toolkit+unsubscr...@googlegroups.com.
For more options, visit this group at 
http://groups.google.com/group/google-web-toolkit?hl=en.

Re: An implementation of Selection/Cursor for RichTextArea

2010-04-13 Thread Manuel Carrasco Moñino

This is a nice functionality which I think should be present in gwt-core.

FYI: Some weeks ago I sent a patch to enable/disable the RichTextArea,
and now it is in the core (http://gwt-code-reviews.appspot.com/131802)

The steps you have to follow are:
1.- Subscribe to gwt-contributors mailing list
2.- Open a ticket with the issue.
3.- Add a patch to the cod-review application (don't forget to add
enough code to test the new feature, and follow the style conventions)
4.- Wait for feedback.
5.- Finally if the patch is accept you must sign a CLA

Take a look to this page:
http://code.google.com/webtoolkit/makinggwtbetter.html#submittingpatches

Cheers
Manolo




On Tue, Apr 13, 2010 at 5:54 PM, kozura koz...@gmail.com wrote:
 So I've written a complete selection/range implementation, which
 allows much more sophisticated Rich Text Editor operations.  You can
 obtain and manipulate the cursor or current selected range - with
 this, I was able to trivially create a proper Link functionality to
 deal with existing links in the editor.  Inspired from the partial
 solution in gwt-rocket, it uses a structural selection model like the
 W3C spec, with emulation to make it work the same in cursed IE*.  And
 except for wrappers around js calls, everything is written in Java to
 allow full GWT optimization.

 I really think this functionality should be part of the GWT library,
 as I think it's the main missing core feature needed to implement full
 RTE capabilities without having to go import external JS RTE
 libraries.  I could always post it as a separate library like gwt-
 rocket, but it's already fairly annoying the number of these little
 libraries with random bits of functionality, which quickly become
 unsupported or bloated.  Additionally, due to an IE bug it would be
 best to have a getSelection/getRange method directly as part of the
 RichTextArea widget.

 But that's just me, do others think this is a really need-to-have for
 GWT?  If others feel it'd be a good addition to GWT, I'd be willing to
 spend the time to try to get it in, maybe starting with getting it in
 the incubator.  And through that would love to see any critiques or
 better ways to do any of it.


 Here's the API, purely structural and doesn't include any of the
 position-based methods that IE has:


 // Current selection is based on a document/window, so there is a
 browser
 // global selection, and a selection for any iframe windows.
 class Selection
 {
    // Get the browser global selection
    static Selection getSelection();

    // Get selection associated with a particular window, ie an RTE
 iframe
    static Selection getSelection(JavaScriptObject window);

    // Get a range containing current selection.  Changes to the range
 do not
    // affect the selection, and vice versa
    Range getRange();

    // Set the selection to what is contained in the range.
 Subsequent
    // changes to this range object will not affect the selection
    void setRange(Range range);

    void clear();
    Document getDocument()
 }

 // Represents a range in an html document, as a start and end text
 node,
 // and the offset in to them, ie between the two | marks here:
 //
 // bA r|ange athat spans/a/bi some elem|ents./i
 class Range
 {
    // Constructors
    Range(Document doc);
    Range(Element element);
    Range(RangeEndPoint cursorPoint);
    Range(RangeEndPoint startPoint, RangeEndPoint endPoint);

    // Get the start/end points of this range, or the cursor if no
 selection
    RangeEndPoint getCursor();
    RangeEndPoint getStart();
    RangeEndPoint getEnd();

    // Get the document this range is contained within
    Document getDocument();

    // Set the range to be a cursor, ie start == end
    void setCursor(RangeEndPoint cursorPoint);

    // Set the range to surround the given element
    void setRange(Element element);

    // Set the endpoints of the range
    void setRange(RangeEndPoint startPoint, RangeEndPoint endPoint);

    // Set the start or end point of the range
    void setStartPoint(RangeEndPoint startPoint);
    void setEndPoint(RangeEndPoint endPoint);


    // Returns true if start == end
    boolean isCursor();

    // Collapses the range to a cursor, either to the start or end
 point
    void collapse(boolean start);

    // Compare the start or end point of this range with the start or
 end point
    // of another.  how determines which endpoints, ie
 Range.START_TO_END
    int compareBoundaryPoint(Range compare, short how);

    // Get just the text selected, ie ange that spans some elem
    String getText();

    // Get the html selected, including any tags needed to make it
 complete, ie:
    // bange athat spans/a/bi some elemi
    String getHtmlText();

    // Returns the lowest level element that completely contains this
 range
    Element getCommonAncestor();


    // Returns an ordered list of all the text nodes in this range
    ListText getSelectedTextNodes();

    // Copy the contents of this range

Re: TabLayoutPanel at 100% height

You need to add it to the RootLayoutPanel vs the RootPanel, as
specified here:
http://code.google.com/webtoolkit/doc/latest/DevGuideUiPanels.html#LayoutPanels

You may also want to put it inside one of the other *LayoutPanels like
a DockLayoutPanel.

On Apr 13, 8:09 am, Jérôme jerome.pasch...@gmail.com wrote:
 Hi there,

 I'm trying to use the newly (version 2.0) introduced TabLayoutPanel.
 What I also want is that the embedded widget in each panel's tab has a
 height of 100% (take the whole vertical place). In order to do so I
 just made a very simple example (no CSS, no customisation, no noting)
 that look like that:

 public void onModuleLoad() {
         TabLayoutPanel p = new TabLayoutPanel(30, Unit.PX);
         // do not work
         // p.setHeight(100%);
         // work
         p.setHeight(700px);
         p.add(new HTML(this), [this]);
         p.add(new HTML(that), [that]);
         p.add(new HTML(the other), [the other]);

         // Add the nameField and sendButton to the RootPanel
         // Use RootPanel.get() to get the entire body element
         RootPanel.get().add(p);

 }

 The html is trivial and only contains a div id=main/div. As you
 can guess I have a problem. When I say that the height should be 100%
 I see nothing. I can click on the various tabs no panel appears /
 changes. But when I say that the height should be XXXpx (any height
 will do the trick) then it works like a charm.

 Of course what I really want is the max height available in the
 browser window that is a 100%.

 Did any one already succeed in doing such a thing? What am I missing?

 Thanks for helping.

 CU Jerome

-- 
You received this message because you are subscribed to the Google Groups 
Google Web Toolkit group.
To post to this group, send email to google-web-tool...@googlegroups.com.
To unsubscribe from this group, send email to 
google-web-toolkit+unsubscr...@googlegroups.com.
For more options, visit this group at 
http://groups.google.com/group/google-web-toolkit?hl=en.

Re: Unresponsive Script Warning - Need help on Incremental Command

2010-04-13 Thread abickford

To clarify, are you getting to your 'onSuccess()' of your callback?
It could be the case that GWT can't deserialize the objects and there
isn't much you can do about it except grab your data in smaller
chunks.  The deRpc may help but it has it's own issues, large payloads
in FireFox throw an exception saying it couldn't evaluate the payload.

If you are successfully getting your data and causing that error in
your code, such as looping over that large collection creating a
table, you will want to investigate IncrementalCommand as you suspect.

On Apr 12, 4:39 am, Anand Mohan Singh andy.bi...@gmail.com wrote:
 Hi,
    I am using GWT 2.0.3 along with Gilead 1.3 and Hibernate 3.2 for
 developing an application. Some of the bean classes have 20+ entities
 because of which when I am trying to bring list of such beans to
 client side the browser is throwing 'Warning: Unresponsive Script'.
 This hold true even when instead of Gilead, DTO is used. So problem is
 basically large chunk of data. However, because of requirement I will
 have to use such big beans.
   I browsed through various forums and after quite a work came along
 'Incremental Command'. However, there is no comprehensive
 documentation/tutorial for it and I don't know how to proceed further
 with it.Hence, if possible can someone help me by explaining how
 'Incremental Command' works and how to use it (If possible any
 comprehensive tutorials). I have already tried to implement on the
 line of the example given in the link below but it was not of much
 use. So request you to please help me through this.

 http://code.google.com/webtoolkit/doc/latest/DevGuideCodingBasicsDela...

-- 
You received this message because you are subscribed to the Google Groups 
Google Web Toolkit group.
To post to this group, send email to google-web-tool...@googlegroups.com.
To unsubscribe from this group, send email to 
google-web-toolkit+unsubscr...@googlegroups.com.
For more options, visit this group at 
http://groups.google.com/group/google-web-toolkit?hl=en.

ext-gwt

2010-04-13 Thread Debjani Roy

Hi,

I would like to know about any commercial implementation with ext-gwt
- how buggy is it etc. Also does it have support for UIBinder

Thanks

-- 
You received this message because you are subscribed to the Google Groups 
Google Web Toolkit group.
To post to this group, send email to google-web-tool...@googlegroups.com.
To unsubscribe from this group, send email to 
google-web-toolkit+unsubscr...@googlegroups.com.
For more options, visit this group at 
http://groups.google.com/group/google-web-toolkit?hl=en.

Re: Need to explicitly remove event handlers manually?

2010-04-13 Thread Tsukasa

This really is one more useful point which would make closures nice
but we should wait for JDK 7 or later for this.

I already ran into some of issues according handlers I missed to
unbind. Therefore it would be nice to find a way to avoid this
destructor-like (java uncommon) coding pattern. I think there should
be an easy way, but actually I haven't had enough time to figure it
out yet and provide an idea for a soultion.

-- 
You received this message because you are subscribed to the Google Groups 
Google Web Toolkit group.
To post to this group, send email to google-web-tool...@googlegroups.com.
To unsubscribe from this group, send email to 
google-web-toolkit+unsubscr...@googlegroups.com.
For more options, visit this group at 
http://groups.google.com/group/google-web-toolkit?hl=en.

Re: How to scroll page to element

2010-04-13 Thread t.dave

I've learned some things reading this thread.  nice.  :)

looks like there are some easier ways of doing things than how i'm
doing my scrolling, and this would probably be not the easiest, but
i'll mention it for completeness.  it's also possible to do the
browser window scrolling in javascript, and make a JSNI call to the
javascript function to do it.  i'm currently using this method to
simulate a page refresh:

function scrollToTop() {
window.scroll( 0, 0 );
}

and the corresponding GWT JSNI method:

public native void jsScrollWindowToTop() /*-{
$wnd.scrollToTop();
}-*/;

just one more option.


On Apr 12, 11:19 am, Brandon Turner boringbran...@gmail.com wrote:
 You can get away without a ScrollPanel if you want to as well.  Something
 like:

 verticalPanel.getElement().setScrollTop(secondFlexTable.getElement().getAbsoluteTop());

 -Brandon

 On Mon, Apr 12, 2010 at 12:12 PM, t.dave da...@lorgeousdays.com wrote:
  use ScrollPanel.setScrollPosition().  that sets the vertical scroll
  position of the scrollpanel, which you will need to calculate.  try
  something like this:

  new ClickHandler() {
   public void onClick( ClickEvent event ) {
     int tableTop = secondFlexTable.getAbsoluteTop();
     scrollPanel.setScrollPosition( tableTop );
   }
  }

  i suppose in your case it wouldn't be a click handler if you're basing
  it off the history token, but hopefully this will point you in the
  right direction.

  On Apr 12, 7:34 am, redlaber vyalov.e...@gmail.com wrote:
   I want to scroll page, generated with gwt, to some element. Its should
   be simple, but i cannt find the solution. For example: I have a
   vertical panel with two flexables. When i get the history token
   goto2 I want to scroll my page to the second table. (Sorry for the
   terrible english).

  --
  You received this message because you are subscribed to the Google Groups
  Google Web Toolkit group.
  To post to this group, send email to google-web-tool...@googlegroups.com.
  To unsubscribe from this group, send email to
  google-web-toolkit+unsubscr...@googlegroups.comgoogle-web-toolkit%2bunsubscr...@googlegroups.com
  .
  For more options, visit this group at
 http://groups.google.com/group/google-web-toolkit?hl=en.

-- 
You received this message because you are subscribed to the Google Groups 
Google Web Toolkit group.
To post to this group, send email to google-web-tool...@googlegroups.com.
To unsubscribe from this group, send email to 
google-web-toolkit+unsubscr...@googlegroups.com.
For more options, visit this group at 
http://groups.google.com/group/google-web-toolkit?hl=en.

Re: UIBinder and CSS resources obfuscation

See the file Resources.gwt.xml .. It has a bunch of properties that you can
specify to control how CSS is generated..
I haven't tried it, but putting this in your module.gwt.xml should do the
trick.

set-configuration-property name=CssResource.style value=pretty /



--Sri


On 13 April 2010 13:32, david.herv...@gmail.com david.herv...@gmail.comwrote:

 Hi,
 I'm currently searching around the web to found out how to tell
 UIBinder not to perform obfuscation (at-all) on css class. I've read
 codes samples in which it is tell to put inside the .gwt.xml the
 property CssResource.obfuscationPrefix  to empty. But the css class
 name doesn't have any links with original name. I've tried to combined
 with the CssResource.style to pretty but the css class are still
 prefixed with the package name in which the css resources are located.

 Prefixing each class with @external in css is some kind of solution,
 but doing it for all class in all css I want to test is not a real
 solution.

 So, my question is : Is there any options for telling UIBinder not to
 obfuscate css class ? (like a global @external)

 Thanks in advance

 --
 You received this message because you are subscribed to the Google Groups
 Google Web Toolkit group.
 To post to this group, send email to google-web-tool...@googlegroups.com.
 To unsubscribe from this group, send email to
 google-web-toolkit+unsubscr...@googlegroups.comgoogle-web-toolkit%2bunsubscr...@googlegroups.com
 .
 For more options, visit this group at
 http://groups.google.com/group/google-web-toolkit?hl=en.



-- 
You received this message because you are subscribed to the Google Groups 
Google Web Toolkit group.
To post to this group, send email to google-web-tool...@googlegroups.com.
To unsubscribe from this group, send email to 
google-web-toolkit+unsubscr...@googlegroups.com.
For more options, visit this group at 
http://groups.google.com/group/google-web-toolkit?hl=en.

Re: Client side error

2010-04-13 Thread Joe Cole

It happened again on IE8.
Is that where you had it happen Michael?

GWT team - does this error even make sense - the global event array
should never be null should it?

On Mar 24, 4:01 am, Michael W mwang_2...@yahoo.com wrote:
 I got similar error.
 Any solution of it?

 On Mar 10, 5:01 am, Joe Cole profilercorporat...@gmail.com wrote:

  I saw an error today which has me puzzled:

  com.google.gwt.core.client.JavaScriptException: (TypeError):
  '$wnd.__gwt_globalEventArray.length' is null or not an object
   number: -2146823281
   description: '$wnd.__gwt_globalEventArray.length' is null or not an
  object

  This doesn't really make sense to me that it could be null - does
  anyone have any idea how this could happen? The application is fully
  initialised and has been in use for quite a time, but idle. Perhaps
  garbage collection?

  The browser config is:
  Browser Mozilla/4.0 (compatible; MSIE 8.0; Windows NT 6.0; Trident/
  4.0; SLCC1; .NET CLR 2.0.50727; Media Center PC 5.0; .NET CLR
  3.0.30729; .NET CLR 3.5.30729)

-- 
You received this message because you are subscribed to the Google Groups 
Google Web Toolkit group.
To post to this group, send email to google-web-tool...@googlegroups.com.
To unsubscribe from this group, send email to 
google-web-toolkit+unsubscr...@googlegroups.com.
For more options, visit this group at 
http://groups.google.com/group/google-web-toolkit?hl=en.

Re: Need to explicitly remove event handlers manually?

2010-04-13 Thread Paul Stockley

I've stepped through the code and there is no requirement to
unregister handlers explicitly. Each widget has a handler manager (as
member variable) that manages its handlers. If a widget is removed
from its parent or its parent explicitly clears it then the registered
handlers will get cleaned up by normal garbage collection.

This is not the case if you are using a global handler manager for say
an event bus. If handlers are added that are specific to a particular
presenter then the presenter needs to remove the handlers when it is
destroyed. I implemented a Place Manager that takes care of all of
this automatically along with history support.

-- 
You received this message because you are subscribed to the Google Groups 
Google Web Toolkit group.
To post to this group, send email to google-web-tool...@googlegroups.com.
To unsubscribe from this group, send email to 
google-web-toolkit+unsubscr...@googlegroups.com.
For more options, visit this group at 
http://groups.google.com/group/google-web-toolkit?hl=en.

[gwt-contrib] Re: Comment on TroubleshootingOOPHM in google-web-toolkit


Comment by vrinda.pujare:

Currently I am using GWT 2.0.3 with tomcat5.5 as a server, IE6 and  
WindowsXP 64bit.
I am trying to debug my application in client debug mode. Also I have  
followed all the steps to set up debug mode. And started tomcat and  
application using DevMode in debug mode. Debug mode launches the browser  
launcher in client debug mode. But when I start debugging application,  
breakpoints does not get invoked when I run application in debug mode.
Could any one help me out to figure our what cud be problem wid GWT 2.0.3  
debugging?


Thanks in advance.


For more information:
http://code.google.com/p/google-web-toolkit/wiki/TroubleshootingOOPHM

--
http://groups.google.com/group/Google-Web-Toolkit-Contributors

To unsubscribe, reply using remove me as the subject.

[gwt-contrib] Have JsInliner refuse to interpose a new expression (issue344801)


Reviewers: scottb,

Description:
Have JsInliner refuse to interpose a new expression
before a substituted argument.


Please review this at http://gwt-code-reviews.appspot.com/344801/show

Affected files:
  M dev/core/src/com/google/gwt/dev/js/JsInliner.java
  M dev/core/test/com/google/gwt/dev/js/JsInlinerTest.java


Index: dev/core/src/com/google/gwt/dev/js/JsInliner.java
===
--- dev/core/src/com/google/gwt/dev/js/JsInliner.java   (revision 7879)
+++ dev/core/src/com/google/gwt/dev/js/JsInliner.java   (working copy)
@@ -673,6 +673,12 @@
 maintainsOrder = false;
   }
 }
+
+public void endVisit(JsNew x, JsContextJsExpression ctx) {
+  if (unevaluated.size()  0) {
+maintainsOrder = false;
+  }
+}

 @Override
 public void endVisit(JsNameRef x, JsContextJsExpression ctx) {
Index: dev/core/test/com/google/gwt/dev/js/JsInlinerTest.java
===
--- dev/core/test/com/google/gwt/dev/js/JsInlinerTest.java  (revision 7879)
+++ dev/core/test/com/google/gwt/dev/js/JsInlinerTest.java  (working copy)
@@ -80,6 +80,29 @@
 compare(expected, input);
   }

+  /**
+   * Test that a new expression breaks argument ordering.
+   */
+  public void testOrderingNew() throws Exception {
+StringBuffer code = new StringBuffer();
+// A static variable x
+code.append(var x;);
+
+// foo() uses x
+code.append(function foo() { alert('x = ' + x); });
+
+// callee does new foo before evaluating its argument
+code.append(function callee(arg) { new foo(); return arg; });
+
+// caller invokes callee with a multi that initializes x
+code.append(function caller() { callee((x=1,2)); });
+
+// bootstrap the program
+code.append(caller(););
+
+compare(code.toString(), code.toString());
+  }
+
   public void testSelfRecursion() throws Exception {
 String input = function a1() { return blah  b1() }
 + function b1() { return bar  a1()} + function c() { a1() }  
c();



--
http://groups.google.com/group/Google-Web-Toolkit-Contributors

To unsubscribe, reply using remove me as the subject.

[gwt-contrib] Re: Have JsInliner refuse to interpose a new expression (issue344801)


Can anyone familiar with GWT's optimization review this?

http://gwt-code-reviews.appspot.com/344801/show

--
http://groups.google.com/group/Google-Web-Toolkit-Contributors

To unsubscribe, reply using remove me as the subject.

[gwt-contrib] Re: Have JsInliner refuse to interpose a new expression (issue344801)

2010-04-13 Thread scottb


LGTM, nits.


http://gwt-code-reviews.appspot.com/344801/diff/1/2
File dev/core/src/com/google/gwt/dev/js/JsInliner.java (right):

http://gwt-code-reviews.appspot.com/344801/diff/1/2#newcode677
dev/core/src/com/google/gwt/dev/js/JsInliner.java:677: public void
endVisit(JsNew x, JsContextJsExpression ctx) {
I assume the logic is essentially the same as for
endVisit(JsInvocation)?  Tiny bit of javadoc noting that would be good.
Also, I think this wants to sort down below endVisit(JsNameRef).

http://gwt-code-reviews.appspot.com/344801/diff/1/3
File dev/core/test/com/google/gwt/dev/js/JsInlinerTest.java (right):

http://gwt-code-reviews.appspot.com/344801/diff/1/3#newcode120
dev/core/test/com/google/gwt/dev/js/JsInlinerTest.java:120:
System.err.println(Input vs );
Unrelated to your patch, but maybe yank this line while you're here?

http://gwt-code-reviews.appspot.com/344801/show

--
http://groups.google.com/group/Google-Web-Toolkit-Contributors

To unsubscribe, reply using remove me as the subject.

[gwt-contrib] Re: Have JsInliner refuse to interpose a new expression (issue344801)

2010-04-13 Thread scottb


Oh, I had one final thought on this, that might be worth a note for the
future, if my logic is sound:

You might imagine having special-cased JsInvocations to an empty target
method to not break evaluation order (since they do nothing).  As a
practical matter, however, we didn't bother doing so because such
constructs would have already been inlined by the time we get here.
This is NOT true for a JsNew op, which can never be inlined.  We might
want to add a TODO that a JsNew op with an empty target method could be
special-cased to not break EvalutationOrder.


http://gwt-code-reviews.appspot.com/344801/show

--
http://groups.google.com/group/Google-Web-Toolkit-Contributors

To unsubscribe, reply using remove me as the subject.

[gwt-contrib] Re: Have JsInliner refuse to interpose a new expression (issue344801)


Thanks, Scott.

I added the TODO at the place indicated.  It applies more broadly,
though: any place the inliner looks at a function call, it would be nice
to know what if any side effects that function call has and would be
affected by.


http://gwt-code-reviews.appspot.com/344801/diff/1/2
File dev/core/src/com/google/gwt/dev/js/JsInliner.java (right):

http://gwt-code-reviews.appspot.com/344801/diff/1/2#newcode677
dev/core/src/com/google/gwt/dev/js/JsInliner.java:677: public void
endVisit(JsNew x, JsContextJsExpression ctx) {
Done and done.

http://gwt-code-reviews.appspot.com/344801/diff/1/3
File dev/core/test/com/google/gwt/dev/js/JsInlinerTest.java (right):

http://gwt-code-reviews.appspot.com/344801/diff/1/3#newcode120
dev/core/test/com/google/gwt/dev/js/JsInlinerTest.java:120:
System.err.println(Input vs );
On 2010/04/13 15:07:30, scottb wrote:

Unrelated to your patch, but maybe yank this line while you're here?


Done.

http://gwt-code-reviews.appspot.com/344801/show

--
http://groups.google.com/group/Google-Web-Toolkit-Contributors

[gwt-contrib] Re: Fixes ICE in JsInliner (issue284802)


LGTM

http://gwt-code-reviews.appspot.com/284802/show

--
http://groups.google.com/group/Google-Web-Toolkit-Contributors

To unsubscribe, reply using remove me as the subject.

[gwt-contrib] [google-web-toolkit] r7912 committed - Added method to change the size of a MutableArray. Added factory metho...


Revision: 7912
Author: rchan...@google.com
Date: Tue Apr 13 05:32:59 2010
Log: Added method to change the size of a MutableArray. Added factory  
method to
construct MeutableArray of specific size. These provide a O(n) way to  
initialize
a MutableArray as shown by  
com.google.gwt.collections.MutableArrayBenchmarkTest


Review at http://gwt-code-reviews.appspot.com/319801

http://code.google.com/p/google-web-toolkit/source/detail?r=7912

Added:
 /trunk/bikeshed/test/com/google/gwt/collections/ClientMutableArrayTest.java
 /trunk/bikeshed/test/com/google/gwt/collections/GwtImmutableArrayTest.java
  
/trunk/bikeshed/test/com/google/gwt/collections/MutableArrayBenchmarkTest.java

Modified:
 /trunk/bikeshed/src/com/google/gwt/collections/CollectionFactory.java
 /trunk/bikeshed/src/com/google/gwt/collections/MutableArray.java
 /trunk/bikeshed/test/com/google/gwt/collections/ObjectArrayTest.java

===
--- /dev/null
+++  
/trunk/bikeshed/test/com/google/gwt/collections/ClientMutableArrayTest.java	 
Tue Apr 13 05:32:59 2010

@@ -0,0 +1,27 @@
+/*
+ * Copyright 2009 Google Inc.
+ *
+ * Licensed under the Apache License, Version 2.0 (the License); you may  
not
+ * use this file except in compliance with the License. You may obtain a  
copy of

+ * the License at
+ *
+ * http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an AS IS BASIS,  
WITHOUT

+ * WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the
+ * License for the specific language governing permissions and limitations  
under

+ * the License.
+ */
+package com.google.gwt.collections;
+
+/**
+ * Re-run {...@link ObjectArrayTest} tests under GWT.
+ */
+public class ClientMutableArrayTest extends ObjectArrayTest {
+  @Override
+  public String getModuleName() {
+return com.google.gwt.collections.Collections;
+  }
+
+}
===
--- /dev/null
+++  
/trunk/bikeshed/test/com/google/gwt/collections/GwtImmutableArrayTest.java	 
Tue Apr 13 05:32:59 2010

@@ -0,0 +1,27 @@
+/*
+ * Copyright 2009 Google Inc.
+ *
+ * Licensed under the Apache License, Version 2.0 (the License); you may  
not
+ * use this file except in compliance with the License. You may obtain a  
copy of

+ * the License at
+ *
+ * http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an AS IS BASIS,  
WITHOUT

+ * WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the
+ * License for the specific language governing permissions and limitations  
under

+ * the License.
+ */
+package com.google.gwt.collections;
+
+/**
+ * Re-run {...@link ImmutableArrayTest} tests under GWT.
+ */
+public class GwtImmutableArrayTest extends ImmutableArrayTest {
+  @Override
+  public String getModuleName() {
+return com.google.gwt.collections.Collections;
+  }
+
+}
===
--- /dev/null
+++  
/trunk/bikeshed/test/com/google/gwt/collections/MutableArrayBenchmarkTest.java	 
Tue Apr 13 05:32:59 2010

@@ -0,0 +1,58 @@
+/*
+ * Copyright 2009 Google Inc.
+ *
+ * Licensed under the Apache License, Version 2.0 (the License); you may  
not
+ * use this file except in compliance with the License. You may obtain a  
copy of

+ * the License at
+ *
+ * http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an AS IS BASIS,  
WITHOUT

+ * WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the
+ * License for the specific language governing permissions and limitations  
under

+ * the License.
+ */
+package com.google.gwt.collections;
+
+import com.google.gwt.benchmarks.client.Benchmark;
+import com.google.gwt.benchmarks.client.IntRange;
+import com.google.gwt.benchmarks.client.Operator;
+import com.google.gwt.benchmarks.client.RangeField;
+
+/**
+ * Benchmarks the performance of various MutableArray methods.
+ */
+public class MutableArrayBenchmarkTest extends Benchmark {
+
+  final IntRange elemRange = new IntRange(0, 5000, Operator.ADD, 100);
+
+  @Override
+  public String getModuleName() {
+return com.google.gwt.collections.Collections;
+  }
+
+  public void testAddGrowth() {
+  }
+
+  public void testAddGrowth(@RangeField(elemRange) Integer numElements) {
+MutableArrayInteger ma = CollectionFactory.createMutableArray();
+
+for (int i = 0; i  numElements; i++) {
+  ma.add(i);
+}
+  }
+
+  public void testSetSizeGrowth() {
+  }
+
+  public void testSetSizeGrowth(@RangeField(elemRange) Integer  
numElements) {

+MutableArrayInteger ma = CollectionFactory.createMutableArray();
+
+ma.setSize(numElements, null);
+for (int i = 0; i  numElements; i++) {
+  ma.set(i, i);
+}
+  }
+
+}

[gwt-contrib] [google-web-toolkit] r7913 committed - Have JsInliner refuse to interpose a new expression...


Revision: 7913
Author: sp...@google.com
Date: Tue Apr 13 06:08:53 2010
Log: Have JsInliner refuse to interpose a new expression
before a substituted argument.

Review at http://gwt-code-reviews.appspot.com/344801

Review by: sco...@google.com
http://code.google.com/p/google-web-toolkit/source/detail?r=7913

Modified:
 /trunk/dev/core/src/com/google/gwt/dev/js/JsInliner.java
 /trunk/dev/core/test/com/google/gwt/dev/js/JsInlinerTest.java

===
--- /trunk/dev/core/src/com/google/gwt/dev/js/JsInliner.java	Tue Mar 23  
14:56:21 2010
+++ /trunk/dev/core/src/com/google/gwt/dev/js/JsInliner.java	Tue Apr 13  
06:08:53 2010

@@ -673,11 +673,28 @@
 maintainsOrder = false;
   }
 }
-
+
 @Override
 public void endVisit(JsNameRef x, JsContextJsExpression ctx) {
   checkName(x.getName());
 }
+
+@Override
+public void endVisit(JsNew x, JsContextJsExpression ctx) {
+  /*
+   * Unless all arguments have already been evaluated, assume
+   * that invoking the new expression might interfere with
+   * the evaluation of the argument.
+   *
+   * It would be possible to allow this if the invoked function
+   * either does nothing or does nothing that affects the
+   * remaining arguments.  However, currently there is no
+   * analysis of the invoked function.
+   */
+  if (unevaluated.size()  0) {
+maintainsOrder = false;
+  }
+}

 @Override
 public void endVisit(JsThisRef x, JsContextJsExpression ctx) {
===
--- /trunk/dev/core/test/com/google/gwt/dev/js/JsInlinerTest.java	Wed Oct  
28 09:10:53 2009
+++ /trunk/dev/core/test/com/google/gwt/dev/js/JsInlinerTest.java	Tue Apr  
13 06:08:53 2010

@@ -79,6 +79,29 @@
 + function c1() { return ex2 ? a1() :c1(); } c1();
 compare(expected, input);
   }
+
+  /**
+   * Test that a new expression breaks argument ordering.
+   */
+  public void testOrderingNew() throws Exception {
+StringBuffer code = new StringBuffer();
+// A static variable x
+code.append(var x;);
+
+// foo() uses x
+code.append(function foo() { alert('x = ' + x); });
+
+// callee does new foo before evaluating its argument
+code.append(function callee(arg) { new foo(); return arg; });
+
+// caller invokes callee with a multi that initializes x
+code.append(function caller() { callee((x=1,2)); });
+
+// bootstrap the program
+code.append(caller(););
+
+compare(code.toString(), code.toString());
+  }

   public void testSelfRecursion() throws Exception {
 String input = function a1() { return blah  b1() }
@@ -94,7 +117,6 @@
 input = optimize(input, JsSymbolResolver.class,  
FixStaticRefsVisitor.class,

 JsInliner.class, JsUnusedFunctionRemover.class);
 expected = optimize(expected);
-System.err.println(Input vs );
 assertEquals(expected, input);
   }
 }

--
http://groups.google.com/group/Google-Web-Toolkit-Contributors

To unsubscribe, reply using remove me as the subject.

[gwt-contrib] [google-web-toolkit] r7914 committed - Fixes ICE in JsInliner....


Revision: 7914
Author: sco...@google.com
Date: Tue Apr 13 07:20:13 2010
Log: Fixes ICE in JsInliner.

We are getting a compiler crash if you try to inline a static JSNI method  
that references 'this' (weird case).  This change disables inlining such  
methods.


http://gwt-code-reviews.appspot.com/284802/show
Found by: spoon
Review by: spoon

http://code.google.com/p/google-web-toolkit/source/detail?r=7914

Modified:
 /trunk/dev/core/src/com/google/gwt/dev/js/JsInliner.java

===
--- /trunk/dev/core/src/com/google/gwt/dev/js/JsInliner.java	Tue Apr 13  
06:08:53 2010
+++ /trunk/dev/core/src/com/google/gwt/dev/js/JsInliner.java	Tue Apr 13  
07:20:13 2010

@@ -1368,25 +1368,36 @@

   /**
* Detects uses of parameters that would produce incorrect results if  
inlined.

-   * Generally speaking, we disallow the use of parameters as lvalues.
+   * Generally speaking, we disallow the use of parameters as lvalues. Also
+   * detects trying to inline a method which references 'this' where the  
call

+   * site has no qualifier.
*/
   private static class ParameterUsageVisitor extends JsVisitor {
-private boolean lvalue = false;
+private final boolean hasThisExpr;
 private final SetJsName parameterNames;
-
-public ParameterUsageVisitor(SetJsName parameterNames) {
+private boolean violation = false;
+
+public ParameterUsageVisitor(boolean hasThisExpr, SetJsName  
parameterNames) {

+  this.hasThisExpr = hasThisExpr;
   this.parameterNames = parameterNames;
 }

 @Override
 public void endVisit(JsNameRef x, JsContextJsExpression ctx) {
   if (ctx.isLvalue()  isParameter(x)) {
-lvalue = true;
+violation = true;
   }
 }

-public boolean parameterAsLValue() {
-  return lvalue;
+@Override
+public void endVisit(JsThisRef x, JsContextJsExpression ctx) {
+  if (!hasThisExpr) {
+violation = true;
+  }
+}
+
+public boolean hasViolation() {
+  return violation;
 }

 /**
@@ -1911,9 +1922,10 @@
 }

 // Check that parameters aren't used in such a way as to prohibit  
inlining

-ParameterUsageVisitor v = new ParameterUsageVisitor(parameterNames);
+ParameterUsageVisitor v = new ParameterUsageVisitor(thisExpr != null,
+parameterNames);
 v.accept(toInline);
-if (v.parameterAsLValue()) {
+if (v.hasViolation()) {
   return false;
 }

--
http://groups.google.com/group/Google-Web-Toolkit-Contributors

[gwt-contrib] [google-web-toolkit] r7915 committed - Rolling back in r7776, which improves JsInliner, along with r7913 and ...


Revision: 7915
Author: sp...@google.com
Date: Tue Apr 13 10:38:07 2010
Log: Rolling back in r7776, which improves JsInliner, along with r7913 and  
r7914, which fix

bugs in it.  The merge commands were:

svn merge --ignore-ancestry -r 7775:7776  
https://google-web-toolkit.googlecode.com/svn/trunk
svn merge --ignore-ancestry -r 7912:7914  
https://google-web-toolkit.googlecode.com/svn/trunk



http://code.google.com/p/google-web-toolkit/source/detail?r=7915

Modified:
  
/branches/snapshot-2010.03.29-r7809/dev/core/src/com/google/gwt/dev/js/JsHoister.java
  
/branches/snapshot-2010.03.29-r7809/dev/core/src/com/google/gwt/dev/js/JsInliner.java
  
/branches/snapshot-2010.03.29-r7809/dev/core/test/com/google/gwt/dev/js/JsInlinerTest.java


===
---  
/branches/snapshot-2010.03.29-r7809/dev/core/src/com/google/gwt/dev/js/JsHoister.java	 
Mon Apr  5 14:44:40 2010
+++  
/branches/snapshot-2010.03.29-r7809/dev/core/src/com/google/gwt/dev/js/JsHoister.java	 
Tue Apr 13 10:38:07 2010

@@ -221,16 +221,9 @@
   stack.push(x);
 }

-/**
- * A this reference can only effectively be hoisted if the call site  
is

- * qualified by a JsNameRef, so we'll ignore this for now.
- */
 @Override
 public void endVisit(JsThisRef x, JsContextJsExpression ctx) {
-  // Set a flag to indicate that we cannot continue, and push a null so
-  // we don't run out of elements on the stack.
-  successful = false;
-  stack.push(null);
+  stack.push(new JsThisRef(x.getSourceInfo()));
 }

 public JsExpression getExpression() {
===
---  
/branches/snapshot-2010.03.29-r7809/dev/core/src/com/google/gwt/dev/js/JsInliner.java	 
Mon Apr  5 14:44:40 2010
+++  
/branches/snapshot-2010.03.29-r7809/dev/core/src/com/google/gwt/dev/js/JsInliner.java	 
Tue Apr 13 10:38:07 2010

@@ -136,7 +136,7 @@
*/
   affectedBySideEffects = true;
 }
-
+
 @Override
 public void endVisit(JsObjectLiteral x, JsContextJsExpression ctx) {
   affectedBySideEffects = true;
@@ -600,6 +600,9 @@
* invocations occur.
*/
   private static class EvaluationOrderVisitor extends JsVisitor {
+public static final JsName THIS_NAME = (new JsScope(fake scope) {
+}).declareName(this);
+
 private boolean maintainsOrder = true;
 private final ListJsName toEvaluate;
 private final ListJsName unevaluated;
@@ -670,11 +673,39 @@
 maintainsOrder = false;
   }
 }
-
+
 @Override
 public void endVisit(JsNameRef x, JsContextJsExpression ctx) {
-  JsName name = x.getName();
-
+  checkName(x.getName());
+}
+
+@Override
+public void endVisit(JsNew x, JsContextJsExpression ctx) {
+  /*
+   * Unless all arguments have already been evaluated, assume
+   * that invoking the new expression might interfere with
+   * the evaluation of the argument.
+   *
+   * It would be possible to allow this if the invoked function
+   * either does nothing or does nothing that affects the
+   * remaining arguments.  However, currently there is no
+   * analysis of the invoked function.
+   */
+  if (unevaluated.size()  0) {
+maintainsOrder = false;
+  }
+}
+
+@Override
+public void endVisit(JsThisRef x, JsContextJsExpression ctx) {
+  checkName(THIS_NAME);
+}
+
+public boolean maintainsOrder() {
+  return maintainsOrder  unevaluated.size() == 0;
+}
+
+private void checkName(JsName name) {
   if (!toEvaluate.contains(name)) {
 return;
   }
@@ -683,10 +714,6 @@
 maintainsOrder = false;
   }
 }
-
-public boolean maintainsOrder() {
-  return maintainsOrder  unevaluated.size() == 0;
-}

 /**
  * Determine if an expression contains a reference to a strict  
parameter.

@@ -913,6 +940,7 @@
   }

   inlining.push(invokedFunction);
+  x = tryToUnravelExplicitCall(x);
   JsExpression op = process(x, callerFunction, invokedFunction);

   if (x != op) {
@@ -1030,6 +1058,7 @@

   ListJsExpression hoisted = new ArrayListJsExpression(
   statements.size());
+  JsExpression thisExpr = ((JsNameRef)  
x.getQualifier()).getQualifier();

   ListJsName localVariableNames = new ArrayListJsName();
   boolean sawReturnStatement = false;

@@ -1102,13 +1131,14 @@
   }

   // Confirm that the expression conforms to the desired heuristics
-  if (!isInlinable(program, callerFunction, invokedFunction,
+  if (!isInlinable(program, callerFunction, invokedFunction, thisExpr,
   x.getArguments(), op)) {
 return x;
   }

   // Perform the name replacement
-  NameRefReplacerVisitor v = new NameRefReplacerVisitor(x,  
invokedFunction);

+  NameRefReplacerVisitor v = new NameRefReplacerVisitor(thisExpr,
+  x.getArguments(), invokedFunction.getParameters());
   for (ListIteratorJsName

[gwt-contrib] [google-web-toolkit] r7916 committed - Implement selection on trees...


Revision: 7916
Author: r...@google.com
Date: Tue Apr 13 08:09:30 2010
Log: Implement selection on trees
Rename 'onDataChange' - 'setData', etc.
Minor stylistic fixes

Review at http://gwt-code-reviews.appspot.com/333802

Review by: j...@google.com
http://code.google.com/p/google-web-toolkit/source/detail?r=7916

Added:
 /trunk/bikeshed/src/com/google/gwt/bikeshed/list/shared/DataChanged.java
Deleted:
 /trunk/bikeshed/src/com/google/gwt/bikeshed/list/shared/ListEvent.java
  
/trunk/bikeshed/src/com/google/gwt/bikeshed/list/shared/SizeChangeEvent.java

Modified:
 /trunk/bikeshed/src/com/google/gwt/bikeshed/list/client/HasCell.java
 /trunk/bikeshed/src/com/google/gwt/bikeshed/list/client/ListView.java
  
/trunk/bikeshed/src/com/google/gwt/bikeshed/list/client/PagingTableListView.java

 /trunk/bikeshed/src/com/google/gwt/bikeshed/list/client/SimpleCellList.java
  
/trunk/bikeshed/src/com/google/gwt/bikeshed/list/shared/AbstractListViewAdapter.java
  
/trunk/bikeshed/src/com/google/gwt/bikeshed/list/shared/DefaultSelectionModel.java

 /trunk/bikeshed/src/com/google/gwt/bikeshed/list/shared/SelectionModel.java
  
/trunk/bikeshed/src/com/google/gwt/bikeshed/tree/client/SideBySideTreeNodeView.java
  
/trunk/bikeshed/src/com/google/gwt/bikeshed/tree/client/SideBySideTreeView.java
  
/trunk/bikeshed/src/com/google/gwt/bikeshed/tree/client/StandardTreeNodeView.java

 /trunk/bikeshed/src/com/google/gwt/bikeshed/tree/client/TreeNodeView.java
 /trunk/bikeshed/src/com/google/gwt/bikeshed/tree/client/TreeView.java
 /trunk/bikeshed/src/com/google/gwt/bikeshed/tree/client/TreeViewModel.java
  
/trunk/bikeshed/src/com/google/gwt/sample/bikeshed/cookbook/client/BasicTreeRecipe.java
  
/trunk/bikeshed/src/com/google/gwt/sample/bikeshed/cookbook/client/Cookbook.java
  
/trunk/bikeshed/src/com/google/gwt/sample/bikeshed/cookbook/client/MailRecipe.java
  
/trunk/bikeshed/src/com/google/gwt/sample/bikeshed/cookbook/client/MultiSelectionModel.java
  
/trunk/bikeshed/src/com/google/gwt/sample/bikeshed/cookbook/client/MyTreeViewModel.java
  
/trunk/bikeshed/src/com/google/gwt/sample/bikeshed/stocks/client/StocksDesktop.java
  
/trunk/bikeshed/src/com/google/gwt/sample/bikeshed/stocks/client/StocksMobile.java
  
/trunk/bikeshed/src/com/google/gwt/sample/bikeshed/stocks/client/TransactionTreeViewModel.java
  
/trunk/bikeshed/src/com/google/gwt/sample/bikeshed/stocks/shared/StockQuote.java


===
--- /dev/null
+++  
/trunk/bikeshed/src/com/google/gwt/bikeshed/list/shared/DataChanged.java	 
Tue Apr 13 08:09:30 2010

@@ -0,0 +1,70 @@
+/*
+ * Copyright 2010 Google Inc.
+ *
+ * Licensed under the Apache License, Version 2.0 (the License); you may  
not
+ * use this file except in compliance with the License. You may obtain a  
copy of

+ * the License at
+ *
+ * http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an AS IS BASIS,  
WITHOUT

+ * WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the
+ * License for the specific language governing permissions and limitations  
under

+ * the License.
+ */
+package com.google.gwt.bikeshed.list.shared;
+
+import java.util.List;
+
+/**
+ * Represents a list event.
+ *
+ * @param T the type of data in the list
+ */
+public class DataChangedT {
+
+  private final int length;
+  private final int start;
+  private final ListT values;
+
+  /**
+   * Creates a {...@link DataChanged} instance.
+   *
+   * @param start the start index of the data
+   * @param length the length of the data
+   * @param values the new values
+   */
+  public DataChanged(int start, int length, ListT values) {
+this.start = start;
+this.length = length;
+this.values = values;
+  }
+
+  /**
+   * Get the length of the changed data set.
+   *
+   * @return the length of the data set
+   */
+  public int getLength() {
+return length;
+  }
+
+  /**
+   * Get the start index of the changed data.
+   *
+   * @return the start index
+   */
+  public int getStart() {
+return start;
+  }
+
+  /**
+   * Gets the value.
+   *
+   * @return the value
+   */
+  public ListT getValues() {
+return values;
+  }
+}
===
--- /trunk/bikeshed/src/com/google/gwt/bikeshed/list/shared/ListEvent.java	 
Fri Apr  9 11:12:18 2010

+++ /dev/null
@@ -1,70 +0,0 @@
-/*
- * Copyright 2010 Google Inc.
- *
- * Licensed under the Apache License, Version 2.0 (the License); you may  
not
- * use this file except in compliance with the License. You may obtain a  
copy of

- * the License at
- *
- * http://www.apache.org/licenses/LICENSE-2.0
- *
- * Unless required by applicable law or agreed to in writing, software
- * distributed under the License is distributed on an AS IS BASIS,  
WITHOUT

- * WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the
- * License for the specific language governing permissions and

[gwt-contrib] Re: Added method to change the size of a MutableArray. Added factory method to construct a nonempty ... (issue319801)

2010-04-13 Thread rchandia


http://gwt-code-reviews.appspot.com/319801/show

--
http://groups.google.com/group/Google-Web-Toolkit-Contributors

To unsubscribe, reply using remove me as the subject.

[gwt-contrib] Re: Added method to change the size of a MutableArray. Added factory method to construct a nonempty ... (issue319801)

2010-04-13 Thread rchandia


Dan, Freeland:

Please disregard this last patch. I will not submit it. Actually I do
not know why it got appended to this issue instead of becoming a new
issue.

Do I need to do something else to clean up?

On 2010/04/13 18:36:44, rchandia wrote:




http://gwt-code-reviews.appspot.com/319801/show

--
http://groups.google.com/group/Google-Web-Toolkit-Contributors

Re: [gwt-contrib] Have JsInliner refuse to interpose a new expression (issue344801)

2010-04-13 Thread Scott Blum

I *think* it's the latter.

On Tue, Apr 13, 2010 at 3:51 PM, Ray Cromwell cromwell...@gmail.com wrote:

 What's the motivation for this change? Did some code actually trip
 this up, or was this more a result of the new way constructor
 functions do work, whereas in the past, the majority of JsNew
 operations didn't do anything interesting?

 -Ray



-- 
http://groups.google.com/group/Google-Web-Toolkit-Contributors

[gwt-contrib] [google-web-toolkit] r7917 committed - Fix for delegate-of-the-delegate NPE after ListView.Delegate refactor


Revision: 7917
Author: rj...@google.com
Date: Tue Apr 13 12:46:18 2010
Log: Fix for delegate-of-the-delegate NPE after ListView.Delegate refactor

http://code.google.com/p/google-web-toolkit/source/detail?r=7917

Modified:
  
/trunk/bikeshed/src/com/google/gwt/valuestore/client/ValuesListViewTable.java


===
---  
/trunk/bikeshed/src/com/google/gwt/valuestore/client/ValuesListViewTable.java	 
Fri Apr  9 11:12:18 2010
+++  
/trunk/bikeshed/src/com/google/gwt/valuestore/client/ValuesListViewTable.java	 
Tue Apr 13 12:46:18 2010

@@ -92,6 +92,9 @@
*/
   public void setDelegate(ValuesListView.Delegate delegate) {
 this.delegate = delegate;
+if (delegate != null) {
+  delegate.onRangeChanged(0, table.getNumDisplayedItems());
+}
   }

   public void setValueList(ListValuesK newValues) {
@@ -103,7 +106,10 @@
   @Override
   protected void onRangeChanged(ListViewValuesK view) {
 Range range = view.getRange();
-getDelegate().onRangeChanged(range.getStart(), range.getLength());
+Delegate delegate = getDelegate();
+if (delegate != null) {
+  delegate.onRangeChanged(range.getStart(), range.getLength());
+}
   }
 };
   }

--
http://groups.google.com/group/Google-Web-Toolkit-Contributors

To unsubscribe, reply using remove me as the subject.

[gwt-contrib] [google-web-toolkit] r7918 committed - Reverts r7916 (Implement selection on trees), which causes...