date:20120525



On Thursday, May 24, 2012 4:21:54 PM UTC+2, Pete_Scholar wrote:

 This means my only working browser for dev is Chrome, as I'm running 
 Firefox 12.0.


FYI, Firefox 12 support has landed in the SVN 10 days ago: 
http://code.google.com/p/google-web-toolkit/source/browse/trunk/plugins/xpcom/prebuilt/gwt-dev-plugin.xpi
 

-- 
You received this message because you are subscribed to the Google Groups 
Google Web Toolkit group.
To view this discussion on the web visit 
https://groups.google.com/d/msg/google-web-toolkit/-/mNjN_vRyVzYJ.
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.com.
For more options, visit this group at 
http://groups.google.com/group/google-web-toolkit?hl=en.

Re: JSON Parsing in GWT Client

2012-05-25 Thread Raphael André Bauer

Hi,


the solution is really simple:
Use resty gwt: https://github.com/chirino/resty-gwt

A sample project including GWT configuration can be found here:
http://code.google.com/p/play-gae-gwt-dreamteam-showcase

restygwt drives one of the biggest websites in Germany. We never had
any performance issues.


Cheers,

Raphael

On Thu, May 24, 2012 at 10:52 AM, Thomas Broyer t.bro...@gmail.com wrote:


 On Thursday, May 24, 2012 8:56:53 AM UTC+2, dominikz wrote:

 I was just having the same problem. I evaluated the solutions yesterday
 and frankly speaking was surprised by the lack of straight answers from both
 GWT documentation and the internet.
 The number of projects there are is just amazing, but nothing seems to
 stand out.

 I don't like the idea of using Overlay Types since it's too much coding
 and the objects coded this way can be used only on browser-side
 I'm not sure what's the status of AutoBean
 (http://code.google.com/p/google-web-toolkit/wiki/AutoBean) since clicking
 on 'Project Home' here takes me to some I think very old start page for GWT.
 No official references from main GWT
 site https://developers.google.com/web-toolkit/

 AutoBean is the technology backing RequestFactory; it's stable and has very
 few known bugs in 2.4 (one or two, no more; all being already fixed in
 trunk; this is an all different story for RequestFactory).
 I just think nobody had/took the time to copy the wiki page to the dev guide
 at developers.google.com/web-toolkit.

 --
 You received this message because you are subscribed to the Google Groups
 Google Web Toolkit group.
 To view this discussion on the web visit
 https://groups.google.com/d/msg/google-web-toolkit/-/I6JfHi2Gw4QJ.

 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.com.
 For more options, visit this group at
 http://groups.google.com/group/google-web-toolkit?hl=en.



-- 
inc: http://ars-machina.raphaelbauer.com
tech: http://ars-codia.raphaelbauer.com
web: http://raphaelbauer.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.com.
For more options, visit this group at 
http://groups.google.com/group/google-web-toolkit?hl=en.

Re: Large GIN Modules and DevMode performance

Also thought about it how to make reloading a page faster in dev mode.

If I understand you correctly, one should create multiple EntryPoints + 
Host Pages for development where each EntryPoint only shows a subset of the 
complete application?

So it would look like:

- libfeature1.gwt.xml (no entry point and is included in 
ProductionApp.gwt.xml)
- devfeature1.gwt.xml (inherits libfeature1 and has an entry point that is 
able to present feature1)

and each feature/screen/whatever has its own GinModule, right?


@Joseph: How do you precompile modules to JavaScript and then reuse it? Or 
did I misunderstood you?


-- J.

-- 
You received this message because you are subscribed to the Google Groups 
Google Web Toolkit group.
To view this discussion on the web visit 
https://groups.google.com/d/msg/google-web-toolkit/-/k-W9MlGdq14J.
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.com.
For more options, visit this group at 
http://groups.google.com/group/google-web-toolkit?hl=en.

Re: Selection gets lost for multiple selections (when Scroll-Bar is used) in IE

2012-05-25 Thread sharath kumar

It is with the DualListField.Multiple selection gets lost in IE when
scrollbar is used.
Please reply back ASAP.

On Wed, Apr 4, 2012 at 4:33 PM, Philippe Lhoste phi...@gmx.net wrote:

 On 03/04/2012 12:26, sharath kumar wrote:

 What steps will reproduce the problem?
 1. Select first text field
 2. Scroll till the end to select all the items (Use scroll-bar to
 select the below item)
 3. Press shift and try to select last text field in order to select
 all the items in the window. The selection gets lost.


 In which component? A grid? Live or not?


 What is the expected output? What do you see instead?
 All the text fields should be selected. But the selection gets lost
 and one text field is selected randomly.

 What version of the product are you using? On what operating system?
 GXT 2.2.5   with Windows 7.


 Since this is a general GWT mailing list, I think this is better reported
 in the Sencha forums...


 Please provide any additional information below.

 Multiple selection happens if we don't select using scroll bar. If we
 use scroll bar, then only problem is coming.


 --
 Philippe Lhoste
 --  (near) Paris -- France
 --  http://Phi.Lho.free.fr http://phi.lho.free.fr/
 --  --  --  --  --  --  --  --  --  --  --  --  --  --


 --
 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.comgoogle-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=enhttp://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-toolkit@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.

zero width widgets

2012-05-25 Thread Matthew Pocock

Hi,

I seem to have recurrent problems with building up gwt layouts where things
don't appear, and when I dig into the elemnt tree in chrome, I find that
things are getting a zero width. Is there a standard cheat list I should be
going through to debug these kind of issues?

Thanks,

Matthew

-- 
Dr Matthew Pocock
Integrative Bioinformatics Group, School of Computing Science, Newcastle
University
mailto: turingatemyhams...@gmail.com
gchat: turingatemyhams...@gmail.com
msn: matthew_poc...@yahoo.co.uk
irc.freenode.net: drdozer
skype: matthew.pocock
tel: (0191) 2566550
mob: +447535664143

-- 
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.com.
For more options, visit this group at 
http://groups.google.com/group/google-web-toolkit?hl=en.

Re: Error with GWT 2.4



On Thursday, May 24, 2012 8:57:30 PM UTC+2, skippy wrote:

 IE8, Window 7, not chrome plug-in. 


Are you sure Chrome Frame is not installed, but disabled? DOMImplStandard 
should not be used in IE.
See http://code.google.com/p/google-web-toolkit/issues/detail?id=6665


-- 
You received this message because you are subscribed to the Google Groups 
Google Web Toolkit group.
To view this discussion on the web visit 
https://groups.google.com/d/msg/google-web-toolkit/-/xHiXcLtP6lAJ.
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.com.
For more options, visit this group at 
http://groups.google.com/group/google-web-toolkit?hl=en.

Re: Large GIN Modules and DevMode performance



On Friday, May 25, 2012 1:18:38 PM UTC+2, Jens wrote:

 Also thought about it how to make reloading a page faster in dev mode.

 If I understand you correctly, one should create multiple EntryPoints + 
 Host Pages for development where each EntryPoint only shows a subset of the 
 complete application?

 So it would look like:

 - libfeature1.gwt.xml (no entry point and is included in 
 ProductionApp.gwt.xml)
 - devfeature1.gwt.xml (inherits libfeature1 and has an entry point that is 
 able to present feature1)

 and each feature/screen/whatever has its own GinModule, right?


Basically, yes.
But you don't have to go as small as a screen per test harness.
 

 @Joseph: How do you precompile modules to JavaScript and then reuse it? Or 
 did I misunderstood you?


Maybe he's talking about com.google.gwt.dev.CompileModule, which is 
basically running the Precompile phase of a GWT compilation and serializing 
the resulting AST into a *.gwtar file. You'll find such *.gwtar files in 
gwt-user.jar starting with GWT 2.4 (maybe 2.3), but the *.gwtar files are 
very dependent upon the version of GWT, so it's not a portable thing. 
AFAIK, it was thought out to speed up gwt-user.jar, not for general 
consumption.

-- 
You received this message because you are subscribed to the Google Groups 
Google Web Toolkit group.
To view this discussion on the web visit 
https://groups.google.com/d/msg/google-web-toolkit/-/o2mkk6-WToUJ.
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.com.
For more options, visit this group at 
http://groups.google.com/group/google-web-toolkit?hl=en.

Re: Hibernate ClassCastException in GWT hosted mode only

2012-05-25 Thread norandom

Hi all,

here the procedure to apply the patch while  4.1.4 is not relased.
1. create a source folder (ie: src_patch) in your eclipse project
2. create the package org.hibernate.service.classloading.internal;
3. copy the patched source file ClassLoaderServiceImpl.java (attached) in 
the package (patch from v4.1.3)

https://lh6.googleusercontent.com/-XB7EAsl8xkU/T7-MgOAKV3I/E1M/Vdgxb3LRD5w/s1600/Capture.PNG

- eclipse will compile the class in web-inf/classes and it will override 
the old class

Vincent


Le jeudi 17 mai 2012 20:46:37 UTC+2, Christomania a écrit :

 Has somebody found a way to try the patch described here  
 https://hibernate.onjira.com/browse/HHH-7084 ?

 Le mardi 1 mai 2012 18:42:18 UTC+2, AK a écrit :

 Same question, as it has been almost 2 months.  Any word on ways to 
 get this to work without downgrading hibernate?



-- 
You received this message because you are subscribed to the Google Groups 
Google Web Toolkit group.
To view this discussion on the web visit 
https://groups.google.com/d/msg/google-web-toolkit/-/21UXxnGc5D4J.
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.com.
For more options, visit this group at 
http://groups.google.com/group/google-web-toolkit?hl=en.

/*
 * Hibernate, Relational Persistence for Idiomatic Java
 *
 * Copyright (c) 2010, Red Hat Inc. or third-party contributors as
 * indicated by the @author tags or express copyright attribution
 * statements applied by the authors.  All third-party contributions are
 * distributed under license by Red Hat Inc.
 *
 * This copyrighted material is made available to anyone wishing to use, modify,
 * copy, or redistribute it subject to the terms and conditions of the GNU
 * Lesser General Public License, as published by the Free Software Foundation.
 *
 * This program is distributed in the hope that it will be useful,
 * but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY
 * or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU Lesser General Public License
 * for more details.
 *
 * You should have received a copy of the GNU Lesser General Public License
 * along with this distribution; if not, write to:
 * Free Software Foundation, Inc.
 * 51 Franklin Street, Fifth Floor
 * Boston, MA  02110-1301  USA
 */
package org.hibernate.service.classloading.internal;

import java.io.IOException;
import java.io.InputStream;
import java.net.URL;
import java.util.ArrayList;
import java.util.Enumeration;
import java.util.HashSet;
import java.util.Iterator;
import java.util.LinkedHashSet;
import java.util.List;
import java.util.Map;
import java.util.ServiceLoader;

import org.hibernate.cfg.AvailableSettings;
import org.hibernate.service.classloading.spi.ClassLoaderService;
import org.hibernate.service.classloading.spi.ClassLoadingException;

/**
 * Standard implementation of the service for interacting with class loaders
 *
 * @author Steve Ebersole
 */
public class ClassLoaderServiceImpl implements ClassLoaderService {
	private final ClassLoader classClassLoader;
	private final ClassLoader resourcesClassLoader;

	public ClassLoaderServiceImpl() {
		this( ClassLoaderServiceImpl.class.getClassLoader() );
	}

	public ClassLoaderServiceImpl(ClassLoader classLoader) {
		this( classLoader, classLoader, classLoader, classLoader );
	}

	public ClassLoaderServiceImpl(
			ClassLoader applicationClassLoader,
			ClassLoader resourcesClassLoader,
			ClassLoader hibernateClassLoader,
			ClassLoader environmentClassLoader) {
		// Normalize missing loaders
		if ( hibernateClassLoader == null ) {
			hibernateClassLoader = ClassLoaderServiceImpl.class.getClassLoader();
		}

		if ( environmentClassLoader == null || applicationClassLoader == null ) {
			ClassLoader sysClassLoader = locateSystemClassLoader();
			ClassLoader tccl = locateTCCL();
			if ( environmentClassLoader == null ) {
environmentClassLoader = sysClassLoader != null ? sysClassLoader : hibernateClassLoader;
			}
			if ( applicationClassLoader == null ) {
applicationClassLoader = tccl != null ? tccl : hibernateClassLoader;
			}
		}

		if ( resourcesClassLoader == null ) {
			resourcesClassLoader = applicationClassLoader;
		}

		final LinkedHashSetClassLoader classLoadingClassLoaders = new LinkedHashSetClassLoader();
		classLoadingClassLoaders.add( applicationClassLoader );
		classLoadingClassLoaders.add( hibernateClassLoader );
		classLoadingClassLoaders.add( environmentClassLoader );

		//this.classClassLoader = new ClassLoader() {
		System.out.println(PATCH: https://hibernate.onjira.com/browse/HHH-7084;);
		this.classClassLoader = new ClassLoader( applicationClassLoader ) {
			@Override
			protected Class? findClass(String name) throws ClassNotFoundException {
for ( ClassLoader loader : classLoadingClassLoaders ) {
	try {
		return loader.loadClass( name );
	}
	catch (Exception ignore) {
	}
}

Re: zero width widgets

2012-05-25 Thread Derek

I suspect this is from mixing layoutpanels with non-layoutpanels. By
and large, layoutpanels (anything with LayoutPanel as part of its
name) need to form an unbroken chain with a previous layoutpanel all
the way back to RootLayoutPanel. If you have an HTMLPanel or
simplepanel or other non-layoutpanel between layoutpanels, the inner
layoutpanel will be sized to zero unless you explicitly set its width
and height.

On May 25, 7:59 am, Matthew Pocock turingatemyhams...@gmail.com
wrote:
 Hi,

 I seem to have recurrent problems with building up gwt layouts where things
 don't appear, and when I dig into the elemnt tree in chrome, I find that
 things are getting a zero width. Is there a standard cheat list I should be
 going through to debug these kind of issues?

 Thanks,

 Matthew

 --
 Dr Matthew Pocock
 Integrative Bioinformatics Group, School of Computing Science, Newcastle
 University
 mailto: turingatemyhams...@gmail.com
 gchat: turingatemyhams...@gmail.com
 msn: matthew_poc...@yahoo.co.uk
 irc.freenode.net: drdozer
 skype: matthew.pocock
 tel: (0191) 2566550
 mob: +447535664143

-- 
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.com.
For more options, visit this group at 
http://groups.google.com/group/google-web-toolkit?hl=en.

Re: JSON Parsing in GWT Client

2012-05-25 Thread Alex opn

restygwt drives one of the biggest websites in Germany. We never had 
any performance issues.

You can't say which one, i suggest? :) Would be interesting to know.

Am Freitag, 25. Mai 2012 13:03:58 UTC+2 schrieb ra:

 Hi, 


 the solution is really simple: 
 Use resty gwt: https://github.com/chirino/resty-gwt 

 A sample project including GWT configuration can be found here: 
 http://code.google.com/p/play-gae-gwt-dreamteam-showcase 

 restygwt drives one of the biggest websites in Germany. We never had 
 any performance issues. 


 Cheers, 

 Raphael 

 On Thu, May 24, 2012 at 10:52 AM, Thomas Broyer t.bro...@gmail.com 
 wrote: 
  
  
  On Thursday, May 24, 2012 8:56:53 AM UTC+2, dominikz wrote: 
  
  I was just having the same problem. I evaluated the solutions yesterday 
  and frankly speaking was surprised by the lack of straight answers from 
 both 
  GWT documentation and the internet. 
  The number of projects there are is just amazing, but nothing seems to 
  stand out. 
  
  I don't like the idea of using Overlay Types since it's too much coding 
  and the objects coded this way can be used only on browser-side 
  I'm not sure what's the status of AutoBean 
  (http://code.google.com/p/google-web-toolkit/wiki/AutoBean) since 
 clicking 
  on 'Project Home' here takes me to some I think very old start page for 
 GWT. 
  No official references from main GWT 
  site https://developers.google.com/web-toolkit/ 
  
  AutoBean is the technology backing RequestFactory; it's stable and has 
 very 
  few known bugs in 2.4 (one or two, no more; all being already fixed in 
  trunk; this is an all different story for RequestFactory). 
  I just think nobody had/took the time to copy the wiki page to the dev 
 guide 
  at developers.google.com/web-toolkit. 
  
  -- 
  You received this message because you are subscribed to the Google 
 Groups 
  Google Web Toolkit group. 
  To view this discussion on the web visit 
  https://groups.google.com/d/msg/google-web-toolkit/-/I6JfHi2Gw4QJ. 
  
  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.com. 
  For more options, visit this group at 
  http://groups.google.com/group/google-web-toolkit?hl=en. 



 -- 
 inc: http://ars-machina.raphaelbauer.com 
 tech: http://ars-codia.raphaelbauer.com 
 web: http://raphaelbauer.com 


-- 
You received this message because you are subscribed to the Google Groups 
Google Web Toolkit group.
To view this discussion on the web visit 
https://groups.google.com/d/msg/google-web-toolkit/-/YEMebkQic2gJ.
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.com.
For more options, visit this group at 
http://groups.google.com/group/google-web-toolkit?hl=en.

Re: URL for GwtDevPluginSetup.exe is broken

2012-05-25 Thread Alex opn

Had the same yesterday, was able to download it from svn:

http://google-web-toolkit.googlecode.com/svn-history/trunk/plugins/ie/prebuilt/gwt-dev-plugin-x86.msi

I think with the above you get the newest available?

Also possible to add a specific revision (r10267 in this case):
http://google-web-toolkit.googlecode.com/svn-history/r10267/trunk/plugins/ie/prebuilt/gwt-dev-plugin-x86.msi

taken from:
http://stackoverflow.com/questions/6600226/trouble-installing-gwt-developer-plugin-for-ie-through-firewall

Regards

Am Mittwoch, 23. Mai 2012 18:16:18 UTC+2 schrieb gwtdev:

Since yesterday, I keep getting following error when load local GWT
application for the first time, please help where I can get
GwtDevPluginSetup.exe.

404. That’s an error.
The requested URL /tag/s/appguid%3D%7B9a5e649a-
ec63-4c7d-99bf-75adb345e7e5%7D%26lang%3Den%26appname%3DGWT
%2520Developer%2520Plugin%2520for%2520IE%2520%2528x86%2529%26needsadmin
%3Dfalse/gwt/plugins/ie/GwtDevPluginSetup.exe was not found on this
server. That’s all we know

--
You received this message because you are subscribed to the Google Groups
Google Web Toolkit group.
To view this discussion on the web visit
https://groups.google.com/d/msg/google-web-toolkit/-/vKxfIheh-VIJ.
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.com.
For more options, visit this group at
http://groups.google.com/group/google-web-toolkit?hl=en.

Re: Large GIN Modules and DevMode performance

2012-05-25 Thread Joseph Lust

Jens,

For clarification, I mean that in Maven we have a GWT project with all of 
our screens, a Server project with the serverside Java code, and a 
GWT-Common project with all of our common components.

So all of our common widgets are compiled in the GWT-Common project and 
wrapped up into a Jar as our common lib. Then our GWT 
project inherits these (via module.gwt.xml and a Maven dependency). AFAIK 
this reduces the size of the main code base we usually work in (just the 
GWT project) and thus Eclipse has less to work with/validate/etc and saves 
some overhead. We also have automation and GWTTestCases that run against 
the common lib as a separate project, so we save time during each build on 
our CI server (Team City).

I am sure Thomas knows more about the GWT compiler than I ever will, so I 
cannot speak to whether building in this way saves time building 
intermediate component code, or whether the code for the common components 
is still compiled with the main GWT project.



Sincerely,
Joseph

-- 
You received this message because you are subscribed to the Google Groups 
Google Web Toolkit group.
To view this discussion on the web visit 
https://groups.google.com/d/msg/google-web-toolkit/-/DR7Gx0pZ4-UJ.
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.com.
For more options, visit this group at 
http://groups.google.com/group/google-web-toolkit?hl=en.

Re: Export CellTable data to PDF and Excel Format using UIBinder

2012-05-25 Thread Joseph Lust

The gist is to:

   1. Get a list of all records you need to export from the celltable on 
   the client browser (and the sorting information).
   2. Return this to the server.
   3. Fetch all the information from (1).
   4. Build into an Excel or CSV output using POI or similar.

In reality it can be quite a pain. I implemented it in ExtJs grids using 
POI and had to have columns in any order user had rearranged them to, 
columns shown/hidden if user had changed them, numeric values as numeric 
type in exported Excel file... etc. You probably want to try to use a 
library to achieve this rather than write it yourself.

I've not used this library, but it looks like a good start: 
GWT-Table-To-Excel http://code.google.com/p/gwt-table-to-excel/
Note that the library uses HTML output, which is a hack (that even ASP.Net 
recommends... LOL). You'll get a warning if you open a HTML table in Excel 
that it is not the proper format, but then Excel will load it anyway. If 
you need rich, native Excel exports, you'll probably need to write your own 
POI framework.

This topic was covered earlier here: 
https://groups.google.com/forum/#!msg/google-web-toolkit/I9VEaFlkA1I/pDXGBBD2g_AJ


Sincerely,
Joseph

-- 
You received this message because you are subscribed to the Google Groups 
Google Web Toolkit group.
To view this discussion on the web visit 
https://groups.google.com/d/msg/google-web-toolkit/-/PfRTT2iTGgkJ.
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.com.
For more options, visit this group at 
http://groups.google.com/group/google-web-toolkit?hl=en.

Re: Large GIN Modules and DevMode performance



 For clarification, I mean that in Maven we have a GWT project with all of 
 our screens, a Server project with the serverside Java code, and a 
 GWT-Common project with all of our common components.

 So all of our common widgets are compiled in the GWT-Common project and 
 wrapped up into a Jar as our common lib. Then our GWT 
 project inherits these (via module.gwt.xml and a Maven dependency). AFAIK 
 this reduces the size of the main code base we usually work in (just the 
 GWT project) and thus Eclipse has less to work with/validate/etc and saves 
 some overhead. We also have automation and GWTTestCases that run against 
 the common lib as a separate project, so we save time during each build on 
 our CI server (Team City).


Ah ok makes more sense ;)
 

 I am sure Thomas knows more about the GWT compiler than I ever will, so I 
 cannot speak to whether building in this way saves time building 
 intermediate component code, or whether the code for the common components 
 is still compiled with the main GWT project.


 The GWT compiler just pulls in any of your source files you have made 
visible using source / super-source in your module.gwt.xml and compiles 
it. So you won't save time in dev mode or during compilation.

-- J.

-- 
You received this message because you are subscribed to the Google Groups 
Google Web Toolkit group.
To view this discussion on the web visit 
https://groups.google.com/d/msg/google-web-toolkit/-/u3bJT1bP6VcJ.
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.com.
For more options, visit this group at 
http://groups.google.com/group/google-web-toolkit?hl=en.

Re: JSON Parsing in GWT Client



 restygwt drives one of the biggest websites in Germany. We never had 
 any performance issues.

 You can't say which one, i suggest? :) Would be interesting to know.


Just a guess: probably studivz.net / meinvz.net. Its a german social 
network and has been rewritten in GWT some time ago. But you have to sign 
up and log in to switch to the new GWT version and see it in action. They 
use Rest like requests and they have forked restygwt on github 
(https://github.com/vznet).

-- J.

-- 
You received this message because you are subscribed to the Google Groups 
Google Web Toolkit group.
To view this discussion on the web visit 
https://groups.google.com/d/msg/google-web-toolkit/-/hLMv17IsStEJ.
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.com.
For more options, visit this group at 
http://groups.google.com/group/google-web-toolkit?hl=en.

Re: Large GIN Modules and DevMode performance



On Friday, May 25, 2012 4:23:02 PM UTC+2, Jens wrote:


  The GWT compiler just pulls in any of your source files you have made 
 visible using source / super-source in your module.gwt.xml and compiles 
 it. So you won't save time in dev mode or during compilation.


Actually even the contrary! The more JARs and directories you add to your 
classpath, the more time it needs to find a class/resource in it. (it's 
negligible here however, with only a single dependency).
I know someone made a modified gwt-maven-plugin that first unpacks every 
dependency in a single temporary directory and launching GWT with only that 
directory, gwt-user.jar and gwt-dev.jar in the classpath; he hasn't 
contributed it though (but said he likely would, IIRC).

-- 
You received this message because you are subscribed to the Google Groups 
Google Web Toolkit group.
To view this discussion on the web visit 
https://groups.google.com/d/msg/google-web-toolkit/-/aIMotzaF_8kJ.
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.com.
For more options, visit this group at 
http://groups.google.com/group/google-web-toolkit?hl=en.

Re: Large GIN Modules and DevMode performance

2012-05-25 Thread Nuno R

The startup time improved while breaking down into smaller parts,  but this 
explode the number of source to maintain only related with dev env.

After a couple of tests it seems the environment has big impact, I'm using 
a windows 7 machine and after a defrag startup time improved significantly 
and are now more closer to the results in a Linux environment that it is 
still faster then w7. 

On Thursday, May 24, 2012 8:58:06 AM UTC+1, Thomas Broyer wrote:



 On Thursday, May 24, 2012 4:25:34 AM UTC+2, Nuno R wrote:

 Hi,

 At the moment we've a large scale application that reached a bottleneck 
 at DevMode startup time.
 It takes ~120s to hit onModuleLoad() call and spends most of the time 
 generating and compiling GIN Injector ~85%.

 Our best result was to target gwt-UnitCache to a virtual disk in RAM, 
 this speeded things up but still not under 90s to reach onModuleLoad().
 Does exist any other way to improvide GIN processing without trying to 
 isolate groups of screens on their own GWT modules (by consequence smaller 
 GIN modules) and run only the one a developer is working on.
 I think we could not reach an acceptable startup time (20s) without it.

 This only brings some small concerns about mantaining dev mode code only 
 (besides gwt.xml files), at least ActivityMapper, HistoryMapper and GIN 
 module to switch.


 How do you tackle this issue on your large scale projects?


 FYI, the Wave team recommended (2 years ago) creating harnesses; i.e. 
 smaller standalone apps.
 In Wave, they made an EditorHarness for their operation-transform-aware 
 rich-text editor and an UndercurrentHarness for the wave panel (the one 
 that displays a wave, with threaded wavelets, and incremental data and 
 feature loading).
 In your case, create small and modular GinModules, and small EntryPoints 
 and Ginjectors specific to a particular set of screens.

 http://www.google.com/events/io/2010/sessions/gwt-continuous-build-testing.html
  (starts 
 around 42:30 in the video)




-- 
You received this message because you are subscribed to the Google Groups 
Google Web Toolkit group.
To view this discussion on the web visit 
https://groups.google.com/d/msg/google-web-toolkit/-/VbVHkpIYEpkJ.
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.com.
For more options, visit this group at 
http://groups.google.com/group/google-web-toolkit?hl=en.

Re: How to access a GWT Servlet from a C# application?

2012-05-25 Thread Clemens

Thank you, both of you, for your quick help. I was not aware that it is so 
easy.
@Joseph: This code snippet is working beautifully!

Am Montag, 21. Mai 2012 19:08:39 UTC+2 schrieb Clemens:

 Hy, 

 I am trying to access a GWT Servlet from a C# application. How do I do 
 that? 
 To explain it with the Sample Code: If I send Test to the server I 
 would like to get back Hello Test!. 

 Regards, 
 Clemens


Am Montag, 21. Mai 2012 19:08:39 UTC+2 schrieb Clemens:

 Hy, 

 I am trying to access a GWT Servlet from a C# application. How do I do 
 that? 
 To explain it with the Sample Code: If I send Test to the server I 
 would like to get back Hello Test!. 

 Regards, 
 Clemens

-- 
You received this message because you are subscribed to the Google Groups 
Google Web Toolkit group.
To view this discussion on the web visit 
https://groups.google.com/d/msg/google-web-toolkit/-/sqxnVypZ2iQJ.
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.com.
For more options, visit this group at 
http://groups.google.com/group/google-web-toolkit?hl=en.

Re: Large GIN Modules and DevMode performance



 After a couple of tests it seems the environment has big impact, I'm using 
 a windows 7 machine and after a defrag startup time improved significantly 
 and are now more closer to the results in a Linux environment that it is 
 still faster then w7. 


On Win7 you should also clean your temp files regularly to make Eclipse 
happy, 
see: http://code.google.com/p/google-web-toolkit/issues/detail?id=5261 . 
Not sure if all these temp files also slow down dev mode but maybe it helps.

-- J.

-- 
You received this message because you are subscribed to the Google Groups 
Google Web Toolkit group.
To view this discussion on the web visit 
https://groups.google.com/d/msg/google-web-toolkit/-/shXHkSXOI5EJ.
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.com.
For more options, visit this group at 
http://groups.google.com/group/google-web-toolkit?hl=en.

Re: zero width widgets

2012-05-25 Thread Matthew Pocock

Ah, thanks. I had a DockLayoutPanel and was using a SimplePane as a
place-holder in east, then later filling this SimplePane with a chat
control. Perhaps there is a better way to dynamically replace
DockLayoutPanel children. Anyway, I now tell the SimplePane to have width
and height 100% (and a few other components, just to be sure), and it's
laying out as I asked now.

Cheers.

Matthew

On 25 May 2012 14:49, Derek derekad...@gmail.com wrote:

 I suspect this is from mixing layoutpanels with non-layoutpanels. By
 and large, layoutpanels (anything with LayoutPanel as part of its
 name) need to form an unbroken chain with a previous layoutpanel all
 the way back to RootLayoutPanel. If you have an HTMLPanel or
 simplepanel or other non-layoutpanel between layoutpanels, the inner
 layoutpanel will be sized to zero unless you explicitly set its width
 and height.

 On May 25, 7:59 am, Matthew Pocock turingatemyhams...@gmail.com
 wrote:
  Hi,
 
  I seem to have recurrent problems with building up gwt layouts where
 things
  don't appear, and when I dig into the elemnt tree in chrome, I find that
  things are getting a zero width. Is there a standard cheat list I should
 be
  going through to debug these kind of issues?
 
  Thanks,
 
  Matthew
 
  --
  Dr Matthew Pocock
  Integrative Bioinformatics Group, School of Computing Science, Newcastle
  University
  mailto: turingatemyhams...@gmail.com
  gchat: turingatemyhams...@gmail.com
  msn: matthew_poc...@yahoo.co.uk
  irc.freenode.net: drdozer
  skype: matthew.pocock
  tel: (0191) 2566550
  mob: +447535664143

 --
 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.com.
 For more options, visit this group at
 http://groups.google.com/group/google-web-toolkit?hl=en.




-- 
Dr Matthew Pocock
Integrative Bioinformatics Group, School of Computing Science, Newcastle
University
mailto: turingatemyhams...@gmail.com
gchat: turingatemyhams...@gmail.com
msn: matthew_poc...@yahoo.co.uk
irc.freenode.net: drdozer
skype: matthew.pocock
tel: (0191) 2566550
mob: +447535664143

-- 
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.com.
For more options, visit this group at 
http://groups.google.com/group/google-web-toolkit?hl=en.

Re: zero width widgets

Instead of SimplePanel you could use SimpleLayoutPanel.

http://google-web-toolkit.googlecode.com/svn/javadoc/2.4/com/google/gwt/user/client/ui/SimpleLayoutPanel.html

-- J.

-- 
You received this message because you are subscribed to the Google Groups 
Google Web Toolkit group.
To view this discussion on the web visit 
https://groups.google.com/d/msg/google-web-toolkit/-/zWrc3c6gFO4J.
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.com.
For more options, visit this group at 
http://groups.google.com/group/google-web-toolkit?hl=en.

Push a message from Websphere 7 to GWT client

2012-05-25 Thread James

Is there an native library to push a message from Websphere 7 to GWT client 
in IE7?

James

-- 
You received this message because you are subscribed to the Google Groups 
Google Web Toolkit group.
To view this discussion on the web visit 
https://groups.google.com/d/msg/google-web-toolkit/-/YgIXHyNTe_cJ.
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.com.
For more options, visit this group at 
http://groups.google.com/group/google-web-toolkit?hl=en.

Howto? include an outside JAR (separate project) with business objects that can be used on the Server and in the Browser Client

2012-05-25 Thread Hanasaki Jiji

I have a separate project that contains all the business domain object 
classes and their respective JUnit test cases.  There is no GWT in this 
project and a desire to keep this as purely a business object class.  How 
can the GWT complier be told generate the needed client side code so the 
same domain objects can be used in GWT based client code called by the GUI 
and RPC's?  I should mention that the BO project JAR built by Maven and 
stored in a corporate repository and the GWT project is built with the 
CodeHaus.org maven plugin (http://mojo.codehaus.org/gwt-maven-plugin/)  I 
suppose there could be two answers to this: 1. Standard GWT using ANT and 
2. a Maven plugin specific way (including the BO classes in the ...client 
package is not an option since the resulting .class files are used in other 
projects).   Thus, this may end up cross posted in the Maven plugin list 
and GWT list.  Please pardon the cross posting should it be necessary.

Thank you.

-- 
You received this message because you are subscribed to the Google Groups 
Google Web Toolkit group.
To view this discussion on the web visit 
https://groups.google.com/d/msg/google-web-toolkit/-/lX3FvND7VjQJ.
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.com.
For more options, visit this group at 
http://groups.google.com/group/google-web-toolkit?hl=en.

Re: SplitLayoutPanel center widget min size

2012-05-25 Thread manstis

Yay, I like :)

If you have a similar patch for when a SplitLayoutPanel contains another 
SplitLayoutPanel where all Widgets have a minimum size and a panel in the 
outermost SplitLayoutPanel is shrunk but can't be collapsed beyond the sum 
of it's child SplitLayoutPanel minimum sizes I'd love to hear ;)

Otherwise I'll have to take a look...

Great patch though.

On Friday, 20 January 2012 14:22:12 UTC, Patrick Tucker wrote:

 7135 was merged into 4731: 
 http://code.google.com/p/google-web-toolkit/issues/detail?id=4731 

 On Jan 20, 8:37 am, Patrick Tucker tucker...@gmail.com wrote: 
  My solution if anyone is interested:
 http://code.google.com/p/google-web-toolkit/issues/detail?id=7135 
  
  On Jan 18, 12:07 pm, Patrick Tucker tucker...@gmail.com wrote: 
  
  
  
   Has anyone played with this? 
  
   I know the current implementation does not allow a minimumsizeto be 
   set on the center widget, what I am wondering is if anyone has 
   modified the current implementation or an older implementation to 
   enforce the minimum. 
  
   My layout needs to enforce that the center widget does not get fully 
   hidden. 
  
   Thanks, 
   Pat- Hide quoted text - 
  
  - Show quoted text -

-- 
You received this message because you are subscribed to the Google Groups 
Google Web Toolkit group.
To view this discussion on the web visit 
https://groups.google.com/d/msg/google-web-toolkit/-/p5vFEtKiK-kJ.
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.com.
For more options, visit this group at 
http://groups.google.com/group/google-web-toolkit?hl=en.

adding eclipse/GWT/app engine project to version control (git)

2012-05-25 Thread Michael

After looking around on here and the web I found conflicting
information about what should be version controlled in a eclipse/app
engine/gwt project.  So I played around a bit and wanted to share what
I found to work for us. We are using git, but this information is very
quickly transferred to any version control.  I stated by using
eclipses create web application wizard.  I want to share this project
and it's settings among a few developers so I source control
the .settings folder which has eclipse project settings.

My .gitignor file is (files and directories that are not version
controlled):
/gwt-unitCache
/test-classes
/war/projectname
/war/WEB-INF/deploy
/war/WEB-INF/classes

I needed the /test/projectname directory for eclipse not to give me an
error (I don't know why though) so to source control that I just put
an empty .gitignor file there.  Git won't source control an empty
directory.

That's it.  This has been working for us for a couple weeks now.  We
don't seem to be including any files that a GWT compile and app engine
deploy create.  We are also just able to clone the repo and import the
project right in to eclipse and we can compile.

I am open to other suggestions, or anything we may be missing.

Michael

-- 
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.com.
For more options, visit this group at 
http://groups.google.com/group/google-web-toolkit?hl=en.

Re: adding eclipse/GWT/app engine project to version control (git)

2012-05-25 Thread Nuno R

Take a look at a maven setup.

Don't need to keep in the repository any files related with IDE 
configuration or project dependencies.
http://maven.apache.org/plugins/maven-eclipse-plugin/ 
http://mojo.codehaus.org/gwt-maven-plugin/ 
After cloning a repo running *maven eclipse:eclipse* generates all the 
configuration files needed by eclipse for a successful import.

Example:
https://github.com/dankurka/mgwt/blob/master/pom.xml 

On Friday, May 25, 2012 10:54:51 PM UTC+1, Michael wrote:

 After looking around on here and the web I found conflicting 
 information about what should be version controlled in a eclipse/app 
 engine/gwt project.  So I played around a bit and wanted to share what 
 I found to work for us. We are using git, but this information is very 
 quickly transferred to any version control.  I stated by using 
 eclipses create web application wizard.  I want to share this project 
 and it's settings among a few developers so I source control 
 the .settings folder which has eclipse project settings. 

 My .gitignor file is (files and directories that are not version 
 controlled): 
 /gwt-unitCache 
 /test-classes 
 /war/projectname 
 /war/WEB-INF/deploy 
 /war/WEB-INF/classes 

 I needed the /test/projectname directory for eclipse not to give me an 
 error (I don't know why though) so to source control that I just put 
 an empty .gitignor file there.  Git won't source control an empty 
 directory. 

 That's it.  This has been working for us for a couple weeks now.  We 
 don't seem to be including any files that a GWT compile and app engine 
 deploy create.  We are also just able to clone the repo and import the 
 project right in to eclipse and we can compile. 

 I am open to other suggestions, or anything we may be missing. 

 Michael 



-- 
You received this message because you are subscribed to the Google Groups 
Google Web Toolkit group.
To view this discussion on the web visit 
https://groups.google.com/d/msg/google-web-toolkit/-/2b7kh38jmLEJ.
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.com.
For more options, visit this group at 
http://groups.google.com/group/google-web-toolkit?hl=en.

UIB click handlers, GWTQuery, and TD's ;)

2012-05-25 Thread Roger Studner

I'm building an app that has a really good use case for me to have my own/raw 
HTML Tables   (I thought of flextable.. but the api pains me).

Anyway.. i'm trying to figure out something and am a bit stuck, thought the 
crowd might have some ideas.

piece of code:

GQuery td = $(td colspan='x'/);
ThingyBarUIB thingyBar = new ThingyBarUIB();
thingyBar.loadData(someObject);
td.append(thingyBar.getElement());   
addClickHandlerToBar(someObject, thingyBar);

Now.. I easily add click handlers via the method/gwtquery all is well.

I'd really like, just to be.. I dunno.. closer to GWT.. to be able to use 
@UiHandler on the UIBinder backed via class ThingyBarUIB.

I realize this 'fails' the second I do thingyBar.getElement().

Any known techniques to get around this i.e. use UiHandler?  reAttach()? 

Thanks!
Roger

-- 
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.com.
For more options, visit this group at 
http://groups.google.com/group/google-web-toolkit?hl=en.

Re: UIB click handlers, GWTQuery, and TD's ;)

If you want to use @UiHandler and thus GWT's event handling I think you 
just have to call onAttach()/onDetach().

So your code would be:

GQuery td = $(td colspan='x'/);
ThingyBarUIB thingyBar = new ThingyBarUIB(); //UiBinder with @UiHandler
thingyBar.loadData(someObject);
thingyBar.attachTo(td); // implemented as: void attachTo(Element e) { 
e.append(getElement()); onAttach(); }

 and later on when you dont need thingybar anymore:

thingyBar.detach(); //implemented as: void detach() { 
getElement().removeFromParent(); onDetach(); }

If you, for some reason, cant say when to detach() you could add your 
thingyBar to RootPanel.detachOnWindowClose(Widget w); in your attachTo() 
method. This would be similar to what Button.wrap(Element e) does. Take a 
look at the source code. But as you add a widget to a td cell you should 
know when to remove it, so calling a detach() method should be preferred.

-- J.

-- 
You received this message because you are subscribed to the Google Groups 
Google Web Toolkit group.
To view this discussion on the web visit 
https://groups.google.com/d/msg/google-web-toolkit/-/Zzt1RztjTRcJ.
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.com.
For more options, visit this group at 
http://groups.google.com/group/google-web-toolkit?hl=en.

[gwt-contrib] Re: Issue 7038: CompositeEditor and ListEditor optimizations (issue1664803)

On Fri, May 25, 2012 at 12:48 AM, Brian Slesinsky skybr...@google.com wrote:
 On Sat, May 19, 2012 at 9:16 AM, Thomas Broyer t.bro...@gmail.com wrote:

 So I agree that is
 seems logical that a live view would only last until the next
 flush() or at most until the next setValue(). But this isn't
 well-documented in this class or anywhere else that I've found so far.

 To
 me, https://developers.google.com/web-toolkit/doc/latest/DevGuideUiEditors is
 clear enough. YMMV though.

 It does explain a lot of things but I'm missing the part where it
 explains the lifetime of the view returned by getEditors(). I'd expect
 that to be explained in the javadoc for getEditors itself.

Right. We should add that.

 So maybe we want to do that now? I also wonder if, for consistency, we
 want to make sure the live view goes stale on *every* call to
 setValue(), which would make our optimization a bit trickier; we
 should probably create a new ListValueProvider and copy all the
 subeditors into it.


 ListEditorWrapper could track this in detach() and from then on act as an
 immutable list (throwing exceptions if you try to change it).

 Sure, that seems like a separate issue though.

 Here's my concern: if you call setValue() twice with the same list,
 this patch will make the live view returned by getEditors() last
 longer than before. So I think maybe we shouldn't reuse the
 ListEditorWrapper instance, just the Editors themselves.
 ListEditorWrapper looks cheap to construct, so we might as well be
 consistent and avoid changing behavior.

 On the other hand, maybe you're expecting that the caller *knows* that
 they're calling setValue() twice with the same List, so they expect
 the view to last longer as a result, sort of like if we were returning
 the same list again without wrapping it. If so, maybe this is a change
 in behavior to be closer to what the user expects?

I think the expected use of getEditors() is that users don't keep an
handle on it.
Even if they, if they call getEditors() after each setValue(), then it
doesn't really matter whether the returned value will be the same
instance or a different one.
How about documenting that? (users shouldn't hold a reference on the
returned list, or if they do, they should refresh it after any call
to setValue(); the returned list only guaranteed to be the same
between calls to setValue, and shouldn't be used after a call to
setValue).

 If you ask me, for consistency, I'd
 remove the null-check in ListEditor and let it fail for 'null' values, just
 like it did originally. It might have been added by mistake, I don't know,
 it was added as part of a bigger change and wasn't tracked/documented at
 all, so who knows?

 I'm not sure what you mean. Do you mean that setValue() should just
 throw an exception for null lists? Or does it blow up later? Or do we
 want to avoid blowing up?

Throw on 'null' (it will actually blow up with an NPE in
ListEditorWrapper's constructor, when retrieving its size).
I would just add an 'assert' in setValue as a hint for developers, and
let it blow in ListEditorWrapper in prod mode.

 That's the kind of behavior I'd rather fight against: getValue just after
 setValue should try to preserve the value (null vs. empty string)

 Yes, good point. Except that in this case, there is no getValue()
 method, since this isn't a LeafValueEditor and doesn't implement
 HasValue.

Yes, that's what I meant by Only if it was then flushed into the
edited property (otherwise your changes to the empty list would be
lost) and We could basically merge OptionalFieldEditor functionality
into ListEditor: ListEditor would have to be made a LeafValueEditor
if we wanted to turn nulls into empty lists.

 It looks like all changes get sent to the backing list via flush()?

Yes

This is the general contract of the Editor framework: do not change
the edited object in real-time, instead queue changes and only apply
them on flush(); similarly, LeafValueEditors are not supposed to
modify the object in-place, but return the new value from
getValue(), otherwise they should be ValueAwareEditors with no
sub-editor.
Maybe it should be made clearer in the doc?

 If we call setValue() with a null list, I don't think flush() can do
 anything, because we don't have any reference to a backing model to
 update! Therefore, any changes made in the ListEditor cannot be saved?
 You'd have to use an OptionalFieldEditor if you want to save it.

 So it seems like at most we can display a read-only, empty list.
 There's no point in letting the user edit something they can't save.

Exactly. Unless we change ListEditor to be a LeafValueEditor, which
would be equivalent to merging OptionalFieldEditor functionality into
ListEditor.


To sum up: ListEditor is currently able to display 'null' lists as if
they were empty lists, but then you cannot edit the list's content (as
there's actually no list), and more importantly getEditors() will
throw a misleading exception, and flush() will throw (so

[gwt-contrib] Contributor License Agreements question

2012-05-25 Thread Alexandre Ardhuin

Hi all,

A few days ago I submitted a patch (
http://gwt-code-reviews.appspot.com/1712803/ ) after having followed the
steps described at
https://developers.google.com/web-toolkit/makinggwtbetter#submittingpatches.
My CTO had signed the Corporate Contributor License Agreement (
http://code.google.com/legal/corporate-cla-v1.0.html ) . He had provided my
name and my corporate email ( alexandre.ardhuin at deveryware com ) in
Schedule A. But with that email, I could not sign in at
http://gwt-code-reviews.appspot.com , a Google Account was required. So I
signed with a google account ( alexandre.ardhuin.dw at gmail com ) to
submit my patch.

How can I check to be identified as a cla-signer ?
Will my patch be taken in account ? I don't see the review made by Thomas
in this contributors group and I wonder why.

Thanks in advance,
Alexandre

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

[gwt-contrib] Re: Contributor License Agreements question



On Friday, May 25, 2012 2:55:21 PM UTC+2, Alexandre Ardhuin wrote:

 I don't see the review made by Thomas in this contributors group and I 
 wonder why.


Ah, you didn't CC google-web-toolkit-contributors@googlegroups.com.

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

Re: [gwt-contrib] Re: Contributor License Agreements question

2012-05-25 Thread Alexandre Ardhuin

2012/5/25 Thomas Broyer t.bro...@gmail.com

 On Friday, May 25, 2012 2:55:21 PM UTC+2, Alexandre Ardhuin wrote:

 I don't see the review made by Thomas in this contributors group and I
 wonder why.


 Ah, you didn't CC 
 google-web-toolkit-contributors@googlegroups.com.http://groups.google.com/group/Google-Web-Toolkit-Contributors


Oops, I though it was done automatically. I have updated the issue.

Thanks for the information.

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

Re: [gwt-contrib] Contributor License Agreements question

2012-05-25 Thread Rajeev Dayal

Hey Alexandre,

You're not able to verify that you signed the CLA in the corporate case. I
looked at our corporate CLA list, and I don't see your name (or that of
your company) listed there.

I'll do some more digging to see what happened. Did you receive any sort of
confirmation e-mail? If so, could you forward it to me directly?


Thanks,
Rajeev


On Fri May 25 08:55:21 GMT-400 2012, Alexandre Ardhuin 
alexandre.ardh...@gmail.com wrote:

 Hi all,

 A few days ago I submitted a patch (
 http://gwt-code-reviews.appspot.com/1712803/ ) after having followed the
 steps described at
 https://developers.google.com/web-toolkit/makinggwtbetter#submittingpatches.
 My CTO had signed the Corporate Contributor License Agreement (
 http://code.google.com/legal/corporate-cla-v1.0.html ) . He had provided
 my name and my corporate email ( alexandre.ardhuin at deveryware com ) in
 Schedule A. But with that email, I could not sign in at
 http://gwt-code-reviews.appspot.com , a Google Account was required. So
 I signed with a google account ( alexandre.ardhuin.dw at gmail com ) to
 submit my patch.

 How can I check to be identified as a cla-signer ?
 Will my patch be taken in account ? I don't see the review made by Thomas
 in this contributors group and I wonder why.

 Thanks in advance,
 Alexandre

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

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

Re: [gwt-contrib] Contributor License Agreements question

2012-05-25 Thread Alexandre Ardhuin

Confirmation email forwarded.

Thanks,
Alexandre

2012/5/25 Rajeev Dayal rda...@google.com

 Hey Alexandre,

 You're not able to verify that you signed the CLA in the corporate case. I
 looked at our corporate CLA list, and I don't see your name (or that of
 your company) listed there.

 I'll do some more digging to see what happened. Did you receive any sort
 of confirmation e-mail? If so, could you forward it to me directly?


 Thanks,
 Rajeev


 On Fri May 25 08:55:21 GMT-400 2012, Alexandre Ardhuin 
 alexandre.ardh...@gmail.com wrote:

 Hi all,

 A few days ago I submitted a patch (
 http://gwt-code-reviews.appspot.com/1712803/ ) after having followed the
 steps described at
 https://developers.google.com/web-toolkit/makinggwtbetter#submittingpatches.
 My CTO had signed the Corporate Contributor License Agreement (
 http://code.google.com/legal/corporate-cla-v1.0.html ) . He had provided
 my name and my corporate email ( alexandre.ardhuin at deveryware com ) in
 Schedule A. But with that email, I could not sign in at
 http://gwt-code-reviews.appspot.com , a Google Account was required.
 So I signed with a google account ( alexandre.ardhuin.dw at gmail com ) to
 submit my patch.

 How can I check to be identified as a cla-signer ?
 Will my patch be taken in account ? I don't see the review made by Thomas
 in this contributors group and I wonder why.

 Thanks in advance,
 Alexandre

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

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

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

[gwt-contrib] Re: Resubmission of Issue 1686803 (issue1717803)


http://gwt-code-reviews.appspot.com/1717803/

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

Re: [gwt-contrib] HELP: Generated Javascript contains wrong prototype assignments

2012-05-25 Thread Stephen Haberman


 Using trunk with your patch does not solve the issue.

Well, shoot. I've tried to recreate your issue and am not having any
luck.

I modified the Hello sample to have a few run asyncs, and put different
constructor calls to the same class in each one, but it's working
annoyingly well.

I can get it to output the secondary split point (your activities
splitpoint) with a defineSeed without the cast map...which should
reproduce your error, but GWT only does this if it's statically removed
the cast (as in my trivial sample it can tell it's not needed), so it
doesn't end up calling dynamicCast anyway.

If I turn on draftCompile to avoid the inlining (and subsequent cast
removal), the cast remains, but then GWT is smart and *does* output the
cast map in the 2nd split point's defineSeed for the split
point-specific constructor, so then dynamicCast works just fine.

It seems like you must be getting to a point where GWT *thinks* there
are no casts, and so does not output the cast map for that constructor,
but then there ends up being a cast.

Unfortunately, I'm not familiar with where/how the cast map derivation
is done, and so am generally not having a lot of progress hunting this
down.

- Stephen

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

Re: [gwt-contrib] HELP: Generated Javascript contains wrong prototype assignments

Hehe yeah its pretty hard to reproduce it in a new project. Even in our app 
branch for that issue I have to delete things out very carefully. As soon 
as I delete too much or something wrong, the ClassCastException disappears 
because split points are generated slightly different or GWT optimizes code 
in a way that it does not call dynamicCast() anymore.
Another example: In our app we also use GIN and after upgrading GIN from 
r236 to latest trunk (r251) the issue disappears because now GIN generates 
its Ginjector slightly different which results in different split points 
assignments (I think the previously non working constructor is now in the 
left over fragment and works again.).

You can see how fragile the situation must be and thats probably why its 
hard to reproduce. Seems like our app hits that single if statement in the 
compiler/optimizer/codesplitter that no one else hits ;-)


-- J.

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

Re: [gwt-contrib] HELP: Generated Javascript contains wrong prototype assignments

2012-05-25 Thread Ray Cromwell

Boy will this be a hard one to fix without a stable repro case. I hope
you keep around a repeatable broken build. The code responsible for
generating prototype assignments is here, you can see:

http://code.google.com/p/google-web-toolkit/source/browse/trunk/dev/core/src/com/google/gwt/dev/jjs/impl/FragmentExtractor.java

In extractStatements() and maybeRemoveCtorsFromDefineSeedStatement()
you can see the logic it uses to compute which ctors to setup. What's
supposed to happen is that in the exclusive fragments,you get another
defineSeed() statement, e.g.

defineSeed(id, superId, castMap, CustomPlace_1);

this will essentially do
CustomPlace_1.prototype = AbstractPlace prototype

There's only a few cases I can see where it can fail to emit this:
1) extractStatements() believes CustomPlace itself is not live in this
fragment (seems unlikely, does the exclusive fragment contain any
other code related to CustomPlace? e.g. a call to its methods?)

2) maybeRemoveCtorsFromDefineSeedStatement() believes that
CustomPlace_1 either a) isn't live (not called in the fragment) or b)
is already loaded from a prior fragment

2b) would represent a serious bug in the compiler's control flow
analyzer/CodeSplitter logic, because the compiler would have to arrive
at the following:


I) when doing the initial fragment, it believes CustomPlace 0 and 2
are live but not 1, so it omits 1
   isLive(CustomPlace_1) returns false
   but somehow, alreadyLoaded(CustomPlace_1) gets set to true
II) when going the exclusive fragment, it believes
   isLive(CustomPlace_1) returns true
   but somehow isAlreadyLoaded(CustomPlace_1) is true


Needless to say it would be very hard to debug this without a small repro case.

As far as the Cast maps are concerned, they are computed in
CastNormalizer, but they are computed prior to code splitting.

-Ray


On Wed, May 23, 2012 at 6:12 AM, Jens jens.nehlme...@gmail.com wrote:
 Hi,

 I hope you can give me some advice on the following problem we currently
 have as I have no idea where to start.

 In our app we use activities and have used an ActivityProxy (pretty similar
 to the solution in
 http://code.google.com/p/google-web-toolkit/issues/detail?id=5129) to code
 split all of our activities and everything works great. Now we have
 refactored our code so that our ActivityProxy acts like an ActivityMapper so
 we can code split groups of activities to reduce the number of split points
 and make split points a bit bigger. This works in general, but now we got
 unexpected ClassCastException's for some of our classes which clearly should
 not happen: We cast specific implementations in an interface they implement
 (e.g. MyPlace implements IPlace, MyCommand implements ActionCommand).

 Currently we have found 3 classes where this happens and they have in common
 that the compile report says that each of this classes belongs to two
 different split points (no code in left over and no code in initial). The
 first split point is our web app split point (the whole app is behind a
 split point so we dont have to load the whole app to decide if someone is
 logged in or not) and the second split point is the mentioned group of
 activities (so the second split point could be treated as a child to the web
 app split point). Whenever this happens, the GWT compiler forgets to insert
 a prototype assignment which results in ClassCastExceptions.

 Some details of whats going on using on of our Place classes that cant be
 cast into its interface anymore:

 Our place class definitions:

 public interface IPlace {
   String getPlaceNameAndToken();
   String getToken();
 }

 public abstract class AbstractPlace extends Place implements IPlace {

   public AbstractPlace(String placeName, IPlace parent) {
     //assign to fields
   }

   //implement getPlaceNameAndToken() using placeName and getToken()

 }

 public class CustomPlace extends AbstractPlace {

   public CustomPlace() {
     this(null, null) //note: 1-arg constructor is skipped
   }

   public CustomPlace(Long dataBaseId1) {
     this(dataBaseId1, null);
   }

   public CustomPlace(Long dataBaseId1, Long dataBaseId2) {
     super(CustomPlace, new SomeParentPlace());
     //assign parameters to fields
   }

   //implement getToken(); from IPlace

 }

 Now in our web app split point we have code that looks like new
 CustomPlace().getPlaceNameAndToken() to fill a Hyperlink. In our second
 split point (that contains a group of Activities), the CustomActivity uses
 the 1-arg constructor and the 2-arg constructor of CustomPlace to create
 places for placeController.goTo() calls.

 After compiling things in PRETTY mode the following has happend:

 The first split point (web app split point) contains the following code:

 function CustomPlace_0(){
   $clinit_Place();
   CustomPlace_2.call(this, null, null);
 }

 function CustomPlace_2(dataBaseId1, dataBaseId2){
   $clinit_Place();
   AbstractPlace_0.call(this, 'custom', new SomeParentPlace_0);
   this.dataBaseId1 =

[gwt-contrib] Re: Add Map support to RequestFactory (issue 6132056)

2012-05-25 Thread James Horsley

I added tests for complex keys and complex values which're passing so it
should do yes.

On 25 May 2012 17:01, alexisnou...@gmail.com wrote:

 Hello
 With your patch, does the function .with() works for a map for example
 in an entity:

 Map(String,ServiceParam) params;

 requestContext.find(id).with(**params.ServiceParam.Reader);

 ServiceParam and Reader are also Entity.
 Will it work?

 thanks,




 http://codereview.appspot.com/**6132056/http://codereview.appspot.com/6132056/


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

[gwt-contrib] Re: Issue 7038: CompositeEditor and ListEditor optimizations (issue1664803)

2012-05-25 Thread Brian Slesinsky

On Fri, May 25, 2012 at 1:43 AM, Thomas Broyer t.bro...@gmail.com wrote:

 I think the expected use of getEditors() is that users don't keep an
 handle on it.
 Even if they, if they call getEditors() after each setValue(), then it
 doesn't really matter whether the returned value will be the same
 instance or a different one.
 How about documenting that? (users shouldn't hold a reference on the
 returned list, or if they do, they should refresh it after any call
 to setValue(); the returned list only guaranteed to be the same
 between calls to setValue, and shouldn't be used after a call to
 setValue).

Yes, that seems fine as a specification. In getEditors: The returned
list will be live until the next call to setValue() and shouldn't be
used after that.


 If you ask me, for consistency, I'd
 remove the null-check in ListEditor and let it fail for 'null' values, just
 like it did originally. It might have been added by mistake, I don't know,
 it was added as part of a bigger change and wasn't tracked/documented at
 all, so who knows?

 I'm not sure what you mean. Do you mean that setValue() should just
 throw an exception for null lists? Or does it blow up later? Or do we
 want to avoid blowing up?

 Throw on 'null' (it will actually blow up with an NPE in
 ListEditorWrapper's constructor, when retrieving its size).
 I would just add an 'assert' in setValue as a hint for developers, and
 let it blow in ListEditorWrapper in prod mode.

 That's the kind of behavior I'd rather fight against: getValue just after
 setValue should try to preserve the value (null vs. empty string)

 Yes, good point. Except that in this case, there is no getValue()
 method, since this isn't a LeafValueEditor and doesn't implement
 HasValue.

 Yes, that's what I meant by Only if it was then flushed into the
 edited property (otherwise your changes to the empty list would be
 lost) and We could basically merge OptionalFieldEditor functionality
 into ListEditor: ListEditor would have to be made a LeafValueEditor
 if we wanted to turn nulls into empty lists.

 It looks like all changes get sent to the backing list via flush()?

 Yes

 This is the general contract of the Editor framework: do not change
 the edited object in real-time, instead queue changes and only apply
 them on flush(); similarly, LeafValueEditors are not supposed to
 modify the object in-place, but return the new value from
 getValue(), otherwise they should be ValueAwareEditors with no
 sub-editor.
 Maybe it should be made clearer in the doc?

 If we call setValue() with a null list, I don't think flush() can do
 anything, because we don't have any reference to a backing model to
 update! Therefore, any changes made in the ListEditor cannot be saved?
 You'd have to use an OptionalFieldEditor if you want to save it.

 So it seems like at most we can display a read-only, empty list.
 There's no point in letting the user edit something they can't save.

 Exactly. Unless we change ListEditor to be a LeafValueEditor, which
 would be equivalent to merging OptionalFieldEditor functionality into
 ListEditor.


 To sum up: ListEditor is currently able to display 'null' lists as if
 they were empty lists, but then you cannot edit the list's content (as
 there's actually no list), and more importantly getEditors() will
 throw a misleading exception, and flush() will throw (so ListEditor
 can really only be used with 'null' lists in read-only use cases,
 where you call edit() on the EditorDriver without ever calling
 flush()).
 This is something we should fix IMO. There are three solutions:
 1. make getEditors() return an empty list instead of throwing and
 avoid the NPE in flush(). Breaking change: getEditors() no longer
 throw if you call it before calling setValue() for the first time.
 2. make setValue throw on a 'null' value, mandating the use of an
 OptionalFieldEditor if the list can be 'null'. This is consistent with
 all other editors (most importantly simple Editor-s with sub-editors,
 e.g. all the examples from the doc except LabelDecorator), with the
 exception of LeafValueEditors which are expected to accept nulls (at
 the editor's discretion actually, but all the LeafValueEditors
 provided in gwt-user.jar accept nulls). Breaking change: well, anyone
 using a ListEditor in a read-only scenario where 'null' values are
 expected would have to update his code to add an OptionalFieldEditor
 in from of the ListEditor.
 3. merging OptionalFieldEditor functionality into ListEditor,
 turning a 'null' value into an empty list so it can be edited. But
 then one could no longer tell the difference between 'null' and an
 empty list, and we'd have to add code so that getValue returns 'null'
 if 'null' was passed to setValue and getList() is empty (but never
 return 'null' if a non-null value was passed to setValue!).

 Patch Set #3 implements #1 above, as it's less likely to break
 anyone's code (quite the contrary), at the expense of being
 inconsistent with most other

[gwt-contrib] Re: Add Map support to RequestFactory (issue 6132056)

2012-05-25 Thread James Horsley

I read Alexis's email too hastily and hadn't noted he was asking about
nested complex types. I'll add some tests to see how it behaves. I'll be
interested to dig into what AutoBean's behaviour is here as that's what I
based on, mimicked, and leveraged for this patch.

On 25 May 2012 19:38, t.bro...@gmail.com wrote:

 On 2012/05/25 17:09:32, james.horsley wrote:

 I added tests for complex keys and complex values which're passing so

 it

 should do yes.


 I don't see any change to addPathsToResolution...
 Also, how would it work? Say I have a MapA,B, A has a reference c to
 a proxy C, and B has a reference d to a proxy D. What should I pass to
 Request#with() to get the Ds of B without the Cs of A? with(theMap.d)
 ? Now what if A and B both have a property e referencing a proxy E, I
 want B's Es but not A's Es, with(theMap.e) wouldn't work here.
 I think maybe there should be special subproperties keys and values,
 so one can say: with(myMap.keys.c, myMap.values.d, myMap.values.e)
 to get A's Cs, and B's Ds and Es (without A's Es!)



 http://codereview.appspot.com/**6132056/http://codereview.appspot.com/6132056/


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

[gwt-contrib] Re: Generated EditorContext class names take actual parameterization into account. (issue1352806)

2012-05-25 Thread skybrian


LGTM. (Some optional nitpicks.)

It seems a little odd that you'd want to do this. I'd expect that the
subtype (Manager or Intern) would have additional fields to edit, so you
wouldn't be able to just reuse the editor for the parent type. But it
should still be allowed.



http://gwt-code-reviews.appspot.com/1352806/diff/2001/user/test/com/google/gwt/editor/client/SimpleBeanEditorTest.java
File user/test/com/google/gwt/editor/client/SimpleBeanEditorTest.java
(right):

http://gwt-code-reviews.appspot.com/1352806/diff/2001/user/test/com/google/gwt/editor/client/SimpleBeanEditorTest.java#newcode67
user/test/com/google/gwt/editor/client/SimpleBeanEditorTest.java:67:
interface Driver
Having two levels of class nesting is rather confusing. Could you move
the Driver and other 4 inner classes up a level?

http://gwt-code-reviews.appspot.com/1352806/diff/2001/user/test/com/google/gwt/editor/client/SimpleBeanEditorTest.java#newcode73
user/test/com/google/gwt/editor/client/SimpleBeanEditorTest.java:73:
static class SubPerson1 extends Person {
I think this might be easier to follow if we made the example more
concrete. Maybe rename these?
HasPersons - Department
SubPerson1 - Manager
SubPerson2 - Intern

http://gwt-code-reviews.appspot.com/1352806/

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

[gwt-contrib] Re: Resubmitting code review 1660804 (issue1716803)


On 2012/05/24 18:06:50, rchandia wrote:

Submitted as r10994

http://gwt-code-reviews.appspot.com/1716803/

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

[gwt-contrib] Re: Issue 7230: allow absolute paths in ui:style's src= (issue1660804)


On 2012/05/24 18:07:45, rchandia wrote:

Reposting for internal resubmission at:
http://gwt-code-reviews.appspot.com/1716803


Submitted as r10994

http://gwt-code-reviews.appspot.com/1660804/

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

[gwt-contrib] Re: Resubmission of Issue 1686803 (issue1717803)


Submitter as r10995

http://gwt-code-reviews.appspot.com/1717803/

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

[gwt-contrib] Re: Add Map support to RequestFactory (issue 6132056)

AutoBean has no equivalent to these partial loading; just like a POJO has
no notion of lazy loading, compared to JPA entities.
Le 25 mai 2012 21:06, James Horsley james.hors...@gmail.com a écrit :

 I read Alexis's email too hastily and hadn't noted he was asking about
 nested complex types. I'll add some tests to see how it behaves. I'll be
 interested to dig into what AutoBean's behaviour is here as that's what I
 based on, mimicked, and leveraged for this patch.

 On 25 May 2012 19:38, t.bro...@gmail.com wrote:

 On 2012/05/25 17:09:32, james.horsley wrote:

 I added tests for complex keys and complex values which're passing so

 it

 should do yes.


 I don't see any change to addPathsToResolution...
 Also, how would it work? Say I have a MapA,B, A has a reference c to
 a proxy C, and B has a reference d to a proxy D. What should I pass to
 Request#with() to get the Ds of B without the Cs of A? with(theMap.d)
 ? Now what if A and B both have a property e referencing a proxy E, I
 want B's Es but not A's Es, with(theMap.e) wouldn't work here.
 I think maybe there should be special subproperties keys and values,
 so one can say: with(myMap.keys.c, myMap.values.d, myMap.values.e)
 to get A's Cs, and B's Ds and Es (without A's Es!)



 http://codereview.appspot.com/**6132056/http://codereview.appspot.com/6132056/




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

[gwt-contrib] Re: HorizontalSplitPanel should use ScheduledCommand instead of Command (issue1718803)


http://gwt-code-reviews.appspot.com/1718803/

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

[gwt-contrib] Re: HorizontalSplitPanel should use ScheduledCommand instead of Command (issue1694803)


On 2012/05/24 20:25:26, rchandia wrote:

LGTM



Reposting internally at:
http://gwt-code-reviews.appspot.com/1718803


Submitted as r10996

http://gwt-code-reviews.appspot.com/1694803/

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

[gwt-contrib] Re: HorizontalSplitPanel should use ScheduledCommand instead of Command (issue1718803)


On 2012/05/25 21:16:45, rchandia wrote:

Submitted as r10996

http://gwt-code-reviews.appspot.com/1718803/

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

[gwt-contrib] Re: Fix for issue 5952: RequestContext#isChanged. (issue1601806)


On 2012/05/25 21:43:28, sjostrand.jonas wrote:

When a proxy is edit()ed not only sub ENTITY proxies but also sub

VALUE proxies

are automatically edit()ed and considered as changed by the auto bean

diff. Can

this be fixed by the same technique? I tried the patch and it didn't

solve the

problem already.


This is the TODO at
http://gwt-code-reviews.appspot.com/1601806/diff/45014/user/src/com/google/web/bindery/requestfactory/shared/impl/AbstractRequestContext.java#newcode1214

The idea is that the ValueProxy wouldn't be sent to the server despite
being edit()ed, because it wouldn't be referenced by the entity proxy
(in the sense that the diff for the entity proxy wouldn't include the
value proxy, as the value proxy before and after would compare
equal).
The only other solution would be to remove the auto-edit feature, or
change how it's done.

http://gwt-code-reviews.appspot.com/1601806/

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

[gwt-contrib] ExternalTextResourceGenerator is locale/machine-dependent. (issue1716804)


Reviewers: ,

Message:
Looking for a reviewer on this.



Note there are other uses of getBytes() without a specified encoding:
 - in XsrfProtectedServiceServlet and XsrfTokenServiceServlet (though
it's not that important, as it's about creating and validating a token;
it could be an issue only if you load-balance between instances that
don't run with the same file.encoding),
 - in RunAsyncfailureServlet in unit tests,
 - in StringTest in unit-tests; that one is expected.

Note also that all uses of new String() and getBytes() when a String
charsetName is passed could be changed to using Charset charset,
resolving that one only once. According to
http://code.google.com/p/google-web-toolkit/source/detail?r=10933 that
could possibly speed things up (including maybe the compiler).
I'll leave that improvement to another CL.

Description:
ExternalTextResourceGenerator is locale/machine-dependent.

Fixes issue 5200


Please review this at https://gwt-code-reviews.appspot.com/1716804/

Affected files:
  M user/src/com/google/gwt/resources/rg/ExternalTextResourceGenerator.java


Index:  
user/src/com/google/gwt/resources/rg/ExternalTextResourceGenerator.java
diff --git  
a/user/src/com/google/gwt/resources/rg/ExternalTextResourceGenerator.java  
b/user/src/com/google/gwt/resources/rg/ExternalTextResourceGenerator.java
index  
7bb445e67456669508b2bfd13c42843638652861..043075997136ed733dfa7f1c81aeb8dd652e6405  
100644
---  
a/user/src/com/google/gwt/resources/rg/ExternalTextResourceGenerator.java
+++  
b/user/src/com/google/gwt/resources/rg/ExternalTextResourceGenerator.java

@@ -105,7 +105,7 @@ public final class ExternalTextResourceGenerator extends

 urlExpression = context.deploy(
  
context.getClientBundleType().getQualifiedSourceName().replace('.', '_')
-+ _jsonbundle.txt, text/plain,  
wrappedData.toString().getBytes(), true);
++ _jsonbundle.txt, text/plain,  
Util.getBytes(wrappedData.toString()), true);


 TypeOracle typeOracle = context.getGeneratorContext().getTypeOracle();
 JClassType stringType = typeOracle.findType(String.class.getName());
@@ -171,7 +171,7 @@ public final class ExternalTextResourceGenerator extends
   }

   private String getMd5HashOfData() {
-return Util.computeStrongName(data.toString().getBytes());
+return Util.computeStrongName(Util.getBytes(data.toString()));
   }

   private boolean shouldUseJsonp(ResourceContext context, TreeLogger  
logger) {



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

[gwt-contrib] Re: Issue 7038: CompositeEditor and ListEditor optimizations (issue1664803)


On 2012/05/25 17:58:21, skybrian wrote:


In getEditors: The returned
list will be live until the next call to setValue() and shouldn't be
used after that.


[...]


Yeah, #3 is out. Option #2 would make sense if we expected that
complaining early means that most people would fix the problem before
it reaches production. I'm not convinced of that; I think people won't
test the corner cases that result in null list fields in their
AutoBeans (particularly if those corner cases come from the server).


AFAICT, that's already an issue with other editors though: in the
DevGuide for editors [1], if the address or manager is null, then
you'll have an NPE at runtime.

(note: not necessarily AutoBeans, any POJO can be edited with the Editor
framework ;-) )

[1]
https://developers.google.com/web-toolkit/doc/latest/DevGuideUiEditors#Quickstart


So I think #1 might actually be right; accept that the list might be
null and don't break, but don't let the user edit the field either.
Also, I see that having a null is actually the initial state of the
ListEditor so it seems okay to go back to that state.



It seems okay for getEditors to return an empty list when we have no
backing list (yes, we have no editors).



If that seems okay, here's the javadoc to fix so we don't have to have
this discussion again:



- ListEditor constructor: The ListEditor will have no backing list
until setValue() is called with a non-null value.



- setValue(): If a null is passed in, the ListEditor will have no
backing list and edits cannot be made.



- getEditors(): If there is no backing list, an empty list will be

returned.


- getList(): Returns null if there is no backing list and edits
cannot be made.


Done.

https://gwt-code-reviews.appspot.com/1664803/

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

[gwt-contrib] Re: ExternalTextResourceGenerator is locale/machine-dependent. (issue1716804)

2012-05-25 Thread jat


LGTM

https://gwt-code-reviews.appspot.com/1716804/

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

[gwt-contrib] Re: Fix for issue 5952: RequestContext#isChanged. (issue1601806)