Announcing GWT 1.6 Milestone 1

[Scott Blum]

Greetings GWT developers,

The GWT team is happy to announce the availability of 1.6 Milestone 1! Binary distributions are available for download directly from GWT's Google Code project.

http://code.google.com/p/google-web-toolkit/downloads/list?can=1&q=1.6.0

As always, milestone builds like this are use-at-your-own-risk. There are known bugs, and it definitely isn't ready for production use. Please expect some trial and error getting everything to work. The javadoc that comes bundled with the distribution should be up-to-date, but the online Developer Guide (http://code.google.com/docreader/#p=google-web-toolkit-doc-1-6) is still very much a work in progress. We will be updating it over the next several weeks. In lieu of an up-to-date Developer Guide and release notes, below are the major highlights relative to GWT 1.5.3.

*** New Project Structure in GWT 1.6 ***

One of the biggest changes to GWT 1.6 is a new project structure. The old output format has been replaced by the standard Java web app expanded "war" format, and the actual directory name does default to "/war". Note that the war directory is not only for compiler output; it is also intended to contain handwritten static resources that you want to be included in your webapp alongside GWT modules (that is, things you'd want to version control). Please also note that the "GWTShell" and "GWTCompiler" tools will maintain their legacy behavior, but they have been deprecated in favor of new "HostedMode" and "Compiler" tools which use the new war output. When 1.6 is officially released, we will be encouraging existing projects to update to the new directory format and to use the new tools to take advantage of new features and for compatibility with future GWT releases.

The sample projects provided in the GWT distribution provide an example of correct new project configurations. For more details on the specifics of the new project format, please see GWT 1.6 WAR design document (http://code.google.com/p/google-web-toolkit/wiki/WAR_Design_1_6).

A couple of important changes we should highlight here:

- Projects with server-side code (GWT RPC) must configure a "web.xml" file at "/war/WEB-INF/web.xml". This web.xml file must define and publish any servlets associated with the web application. See the included DynaTable sample. Additionally, server-side library dependencies must be copied into "/war/WEB-INF/lib". For example, any GWT RPC servlets must have a copy of gwt-servlet.jar in this folder.

- HTML host pages will no longer typically be located in a GWT module's public path. Instead, we'll be recommending that people take advantage of the natural web app behavior for serving static files by placing host pages anywhere in the war structure that makes sense. For exmaple, you might want to load a GWT module from a JSP page located in the root of your web app. To keep such handwritten static files separate from those produced by the GWT compiler, the latter will be placed into module-specific subdirectories. Any page that wishes to include a GWT module can do so via a script tag by referencing the GWT-produced "<module>.nocache.js script" within that module's subdirectory. As of 1.6, we'll be recommending that only module-specific resources used directly by GWT code, such as image files needed by widgets, should remain on the public path. See the included Showcase sample for some examples of this distinction.

- When you do need to load resources from a module's public path, always construct an absolute URL by prepending GWT.getModuleBaseURL(). For example, 'GWT.getModuleBaseURL() + "dir/file.ext"'. This advice has not changed, but in the past it was easy to be sloppy with this, because the host page and GWT module typically lived in the same directory, so using a relative URL would usually do the right thing. Now that GWT modules live in a subdirectory, you must reference public resources through GWT.getModuleBaseURL().

*** Hosted Mode Enhancements ***

Although the legacy GWTShell still uses an embedded Tomcat server, the new HostedMode runs Jetty instead. There is also a new "Restart Server" button on the main hosted mode window. Clicking this button restarts the internal Jetty server, which allows Java code changes to take effect on the server without having to completely exit and restart hosted mode. This is useful when making code changes to RPC servlets, or when serializable RPC types are modified and the server and client are out of sync.

*** New EventHandler System ***

Event handlers have been added to replace the old event listeners used by Widgets, History, and various other classes. The new system has a few differences from the old system:

- EventHandler methods always take a single parameter: the GwtEvent that the Widget fired. For example, ClickHandler has a single method onClick(ClickEvent).

- Each GwtEvent contains accessors relevant to the event, such as the key that was pressed on KeyEvents. Native events provide access to the underlying native event object.

- Each EventHandler defines only one method, so you do not need to create empty methods just to satisfy the interface requirements.

For users who create their own Widgets, you no longer need to manage listeners manually. Every Widget has a HandlerManager that manages all of its handlers. For native events, such as ClickEvent, just call addDomHandler() from within your code to register a handler and sink the associated event on the Widget. When the native event is detected, the handler will automatically be called. For logical events, such as SelectionEvent, call addHandler() and fire the event manually using the fireEvent() method.

You can see examples of EventHandler usage in many of the updated GWT widgets and samples, or in new projects created with the new webAppCreator tool.

You can now trigger a native event on almost any Element. Create a new native event using the Document.create*Event() methods, then dispatch it on a specific Element by calling Element.dispatchEvent(). These methods allow you to expand your test coverage in ways that were previously impossible.

*** New Widgets ***

DatePicker

The new DatePicker and DateBox widgets allow your users to select a date from a calendar. The Showcase sample provides examples of both of these widgets.

LazyPanel

The new LazyPanel widget allows you to delay the creation of certain sections of your application until they are first accessed, improving startup performance. For example, if your application has a seldom used "Help" section, you can wrap it in a LazyPanel and create the user interface only if and when the user tries to access it. To use the LazyPanel, extend the class and override the abstract createWidget() method. The createWidget() method will be called the first time you call setVisible() on the LazyPanel.

*** Fixed Issues ***

Please see our bug tracker for a full list of fixed issues and enhancements (http://code.google.com/p/google-web-toolkit/issues/list?can=1&q=status:Fixed,FixedNotReleased%20milestone:1_6_RC&num=1000).

As always, please report bugs to our issue tracker (http://code.google.com/p/google-web-toolkit/issues/list) after doing a quick search to see if your issue has already been reported.  If you run into a serious issue, feel free to also reply back to this thread.

Happy coding,

Scott, on behalf of the GWT team

[Dop Sun]

Waiting for quite sometime, and I'm targeting 1.6 to build my GDE
project for quite some time, and finally, got a milestone release
ready. :) Great News!

> (http://code.google.com/p/google-web-toolkit/issues/list?can=1&q=statu...
> ).

[cyril.lakech]

Great news !!!! Félicitations ! We don't need to build from the source
now !

I'm goint to test this M1 next week !

Regards,

> (http://code.google.com/p/google-web-toolkit/issues/list?can=1&q=statu...
> ).

[Samyem Tuladhar]

Great! I was wondering if this version supports OOPHM?

Thanks,
Samyem

> (http://code.google.com/p/google-web-toolkit/issues/list?can=1&q=statu...
> ).

[Arthur Kalmenson]

It doesn't, OOPHM is suppose to be scheduled for 2.0. It is possible
to use OOPHM from the branch, although I haven't tried it myself.

--
Arthur Kalmenson

[Itamar Ravid]

Hi - I've perused the WAR Design document, but I've several questions that remained unanswered.

In my project, we use GWT-SL and Gilead - the former for exporting Spring beans as GWT-RPC services, and the latter for transferring Hibernate-enabled objects back and forth between the server and client.

A consequence of using GWT-SL is that I no longer map my servlets through web.xml - instead, I've a springDispatcher servlet which receives all requests and routes them to the appropriate service exports I've defined with GWT-SL, using the SimpleUrlHandlerMapping which Spring provides.

Now, since I'd like to route only GWT-RPC (and servlet) requests through the Dispatcher servlet, I've only routed /services/* to it.

The difficulty arises when using Hosted Mode. First, I have to copy a modified web.xml to ${project}/tomcat/webapps/ROOT. This web.xml routes all requests to /services/* to my springDispatcher. This web.xml is different than the one I distribute with my WAR, since it also includes mapping for the GWTShellServlet. Also, in order for serialization to work, I have to compile my project once (using GWTCompiler) and copy all resulting *.gwt.rpc files to ${project}/tomcat/webapps/ROOT/${module}/. This occurs because the GWTShellServlet generates these dynamically as GWT-RPC requests arrive.

I've semi-automated the whole process using Ant tasks, and I'm aware of the gwt-maven plugin which provides some of this functionality, but I'd like to know whether the situation has been fully or partly addressed with the 1.6 release.

[Sebastien]

Hi,

I am interested to migrate from the 1.5.3 to the 1.6.0 M1. When I use
this M1 with -war option, the the compiler output is not like an
expanded war. I tried to specify all parameters (-war, -gen, -
extra ..) without success. The output of the compiler is always
similar to the output of the 1.5.3 compiler. The compiler indicates
the 1.6.0 version, so I am sure to use the 1.6. I am under linux
(ubuntu).

Do you have any idea why the compiler output is not like an expanded
war ?

Regards,
Seb

On 6 fév, 16:26, Scott Blum <sco...@google.com> wrote:

> (http://code.google.com/p/google-web-toolkit/issues/list?can=1&q=statu...
> ).

[Dop Sun]

I guess in the sample projects (included in the M1 build), there is
build file for each and every of them, and within that, there build
tasks for the project. Maybe, you can have a try what there?

[Duong BaTien]

> The GWT team is happy to announce the availability of 1.6 Milestone 1!
> Binary distributions are available for download directly from GWT's Google
> Code project.

> Happy coding,

> Scott, on behalf of the GWT team

Greetings:

Thanks for the tremendous efforts from GWT team and supporting
communities. Here are some preliminary exploration and wish list of the
coming doc:

1) Using the webAppCreator (ant 1.7.0, jdk 1.6) i create an application,
run the hosted mode, then the build to compile the app. In the app war
directory, i see the hosted.html and app.nocache.js, but Not the .css
and .html file as listed in the compiler output of the StockWatcher 1.6
example.

2) The current gwt-incubator_1-5_Dec_28.jar is not compatible with
GWT-1.6.0. Hope the new compatible version will be out soon. I am still
working through the examples to understand the new war structure,
especially the project that uses different modular GWT modules and the
location of public files for different modules. Some documentation and
suggested best practices may be required.

3) With new GWT Event Handler, the stockwatcher example shows a
container class to use a single handler method for event (ClickHandler)
coming from different event publishers (but1 and but2). GWT team may
share its experience and approach for all GWT event types in 1 container
class, especially the way to integrate GWT event handler with OpenAjax
Hub, leveraging GWT JSO and coming UIBinder ;-)

4) In the doc Articles and Tutorials, i see GWT for JSON Mashups and
Application of Facebook. It may be a good idea to centralize in 1 place
possible GWT applications and how-to to leverage Google API and
facilities: Gears, AppEngine, google-ajax-examples, etc.

Thanks

Duong BaTien
DBGROUPS and BudhNet

[Sebastien]

Hi,

I take a look on the build.xml of samples. I didn't see any special
difference.
Here is a part of my build.xml :
<java maxmemory="1200m" classname="com.google.gwt.dev.Compiler"
fork="true" failonerror="true" taskName="gwtc">
<arg value="com.raisepartner.prism.PRISM" />
<classpath>
<pathelement location="src/main"/>
<fileset dir="${prj.lib}"><include name="**/*.jar"/></fileset>
</classpath>
</java>
=====>
[gwtc] Compiling module com.raisepartner.prism.PRISM
[gwtc] Compiling 10
permutations
[gwtc] Permutation compile
succeeded
[gwtc] Linking into
war
[gwtc] Link
succeeded
[gwtc] Compilation succeeded --
483,002s

The output is 'war' directory at the root of my project. This
directory contains only the module directory. There is no WEB-INF
directory created. Is it the expected behavior ?
By looking the samples I understand that my main html page (the one
calling the nocache.js) must me copied manually at the root of the war
directory. The content of the <module>.public is copied into war/
<module>. I expected the content would be copied at the root of the
'war' directory.

So my conclusion is that there is no change compared to the
1.5.3 ... :-(

Regards,
Seb

[logicpeters]

"Note that the
war directory is not only for compiler output; it is also intended to
contain handwritten static resources that you want to be included in
your
webapp alongside GWT modules (that is, things you'd want to version
control)."

I'm not following... Do you mean for us to include these static
resources in the compiler output directory and maintain this directory
under source control (with generated code alongside static source,
such as index.html and css files)? That seems like a bad idea to me.
Why not enforce the standard J2EE project path (see Sun's blueprints),
and maintain these files under /src/conf.

[Scott Blum]

On Sun, Feb 8, 2009 at 12:11 PM, Itamar Ravid <itamar...@gmail.com> wrote:

The difficulty arises when using Hosted Mode. First, I have to copy a modified web.xml to ${project}/tomcat/webapps/ROOT. This web.xml routes all requests to /services/* to my springDispatcher. This web.xml is different than the one I distribute with my WAR, since it also includes mapping for the GWTShellServlet. Also, in order for serialization to work, I have to compile my project once (using GWTCompiler) and copy all resulting *.gwt.rpc files to ${project}/tomcat/webapps/ROOT/${module}/. This occurs because the GWTShellServlet generates these dynamically as GWT-RPC requests arrive.

I've semi-automated the whole process using Ant tasks, and I'm aware of the gwt-maven plugin which provides some of this functionality, but I'd like to know whether the situation has been fully or partly addressed with the 1.6 release.

Itamar,

I think you should find 1.6 much easier once you've switched over from GWTShell to HostedMode.  Just use your real web.xml in hosted mode to map in your springDispatcher; you no longer need to use GWTShellServlet at all.

Scott

[Scott Blum]

Hi Sebastien,

It's actually fairly compatible, the difference is more in attitude and expectation.  You would provide the WEB-INF directory yourself, GWT no longer tries to control everything.  The model is more that GWT can plug in easily to existing web apps.  For example, you would also want to setup build rules to compile your server-side code into WEB-INF/classes.

Scott

[Scott Blum]

On Mon, Feb 9, 2009 at 9:47 AM, logicpeters <logic...@gmail.com> wrote:

"Note that the
war directory is not only for compiler output; it is also intended to
contain handwritten static resources that you want to be included in
your
webapp alongside GWT modules (that is, things you'd want to version
control)."

I'm not following...  Do you mean for us to include these static
resources in the compiler output directory and maintain this directory
under source control (with generated code alongside static source,
such as index.html and css files)?  That seems like a bad idea to me.

 

Not exactly... the compiler output directory is considered to be "/war/<moduleName>".... NOT "/war" itself really.  In other words, you can put whatever resources you want into the top-level war directory, the compiler generated files will always be put into the moduleName subdirectory under this, so as not to conflict.

Why not enforce the standard J2EE project path (see Sun's blueprints),
and maintain these files under /src/conf.

We might have to look into this.

Scott

 

[Itamar Ravid]

Awesome! Thanks, Scott.

[maku]

Hi,

from my point of view it is nice that GWT 1.6 brings a more
transparent project structure.

BUT, what we really need is a concept which fits into enterprise
development strategies.

e.g. I'm not able to see how using maven and continous integraton and
developing can be achievend in a painless way.
- running application in hosted mode maybe with an own server (-
noserver mode))
- working with more than one project
- clean seperation of generated and not generated code
- and so on...

Yes, I know it is possible to handle this stuff and we did it ... But
we spent already so much time on this infrastructural things. And I
guess we are not alone....
And to get a simple to use infrastructure is almost impossible.

Regards,
Martin

[logicpeters]

I haven't yet played with the new project structure, but speaking with
1.4 and 1.5 experience, I agree with Martin - that it is somewhat
difficult to build GWT projects with "enterprise continuous
integration systems", such as Maven (which, in particular, depends on
a certain project structure). A little more configurability in where
source can be located may be the best solution to this issue.

[Mark Renouf]

On Feb 6, 10:26 am, Scott Blum <sco...@google.com> wrote:

> DatePicker
>
> The new DatePicker and DateBox widgets allow your users to select a date
> from a calendar. The Showcase sample provides examples of both of these
> widgets.

Ooh. Does the inclusion of DatePicker imply that
ImmutableResourceBundle is also ready? I'm trying to get my head
around how styling works in DatePicker. It looks rather complex...

> LazyPanel
>
> The new LazyPanel widget allows you to delay the creation of certain
> sections of your application until they are first accessed, improving
> startup performance.

We've rolled our own using this exact technique, to make a
LazyTabPanel, etc but this will be easy to retrofit. Great addition,
thanks!

[Alex dP (Violet UML Editor, WebVNC...)]

Congratulation for this release.

Unfortunately, there's still nothing about Maven support. No official
repository, no official plugin.... I hope this will arrive this
Milestone 2.

Regards,
Alexandre de Pellegrin

> (http://code.google.com/p/google-web-toolkit/issues/list?can=1&q=statu...
> ).

[Mark Renouf]

On Feb 10, 11:45 pm, Mark Renouf <mark.ren...@gmail.com> wrote:
> On Feb 6, 10:26 am, Scott Blum <sco...@google.com> wrote:
>
> > DatePicker
>
> > The new DatePicker and DateBox widgets allow your users to select a date
> > from a calendar. The Showcase sample provides examples of both of these
> > widgets.
>
> Ooh. Does the inclusion of DatePicker imply that
> ImmutableResourceBundle is also ready? I'm trying to get my head
> around how styling works in DatePicker. It looks rather complex...

To answer my own question, I just looked at the release branch and the
ImmutableResourceBundle/StyleInjector support was removed. It's got
the standard list of CSS class selectors in the Javadoc, and a set of
default styles has been added to the GWT theme css.

[Scott Blum]

On Wed, Feb 11, 2009 at 4:29 AM, Alex dP (Violet UML Editor, WebVNC...) <alexandre.d...@gmail.com> wrote:

Congratulation for this release.

Unfortunately, there's still nothing about Maven support. No official
repository, no official plugin.... I hope this will arrive this
Milestone 2.

Regards,
Alexandre de Pellegrin

This is not likely for 1.6... we simply don't have any Maven users (much less experts) among the contributors.  If this is something a lot of people would find valuable, it would be great if someone in the community could step up.

I will note that we made some changes in 1.6 regarding ClassLoaders that should remove barriers to running GWT hosted mode (and unit tests) from within Maven.

Scott

[flyin...@gmail.com]

Well, I just finished changing every single listener i have into
handlers and have to rewrite widgets a little that fires listeners. I
have no idea what to do with the build.xml thing to make it using new
compiler.

On Feb 12, 9:26 am, Scott Blum <sco...@google.com> wrote:
> On Wed, Feb 11, 2009 at 4:29 AM, Alex dP (Violet UML Editor, WebVNC...) <
>

[Arthur Kalmenson]

> Congratulation for this release.
>
> Unfortunately, there's still nothing about Maven support. No official
> repository, no official plugin.... I hope this will arrive this
> Milestone 2.
>
> Regards,
> Alexandre de Pellegrin

I think you'll note that GWT is in the central Maven repository
(http://repo1.maven.org/maven2/com/google/gwt/) and there are
currently at least two Maven plugins (being merged), the gwt-maven
plugin: http://code.google.com/p/gwt-maven/ and the Codehaus GWT
plugin: http://mojo.codehaus.org/gwt-maven-plugin/. I and a number of
other people use Maven every day, and it works OK. I'm sure it could
be better, but it does work.

--
Arthur Kalmenson

On Wed, Feb 11, 2009 at 4:29 AM, Alex dP (Violet UML Editor,

[Scott Blum]

Would people be happier if we called this the "official community-supported GWT/Maven integration"? :)

[Allahbaksh]

GWT 1.6 M1 log.tld file not found error
Hi,
I was using GWT 1.5. Today I moved my application to GWT1.6M1 so as to
try the rich features. When I move the project I am getting below
error when I start my application.

Please note if I create some example application it works fine. Can
some one throw some light on the same.

Regards,
Allahbaksh

java.io.FileNotFoundException: C:\workspace\ProjectName\war\WEB-INF
\log.tld (The system cannot find the file specified)
at java.io.FileInputStream.open(Native Method)
at java.io.FileInputStream.<init>(Unknown Source)
at java.io.FileInputStream.<init>(Unknown Source)
at sun.net.www.protocol.file.FileURLConnection.connect(Unknown
Source)
at sun.net.www.protocol.file.FileURLConnection.getInputStream
(Unknown
Source)
at
com.sun.org.apache.xerces.internal.impl.XMLEntityManager.setupCurrentEntity
(Unknown Source)
at
com.sun.org.apache.xerces.internal.impl.XMLVersionDetector.determineDocVersion
(Unknown Source)
at
com.sun.org.apache.xerces.internal.parsers.XML11Configuration.parse
(Unknown Source)
at
com.sun.org.apache.xerces.internal.parsers.XML11Configuration.parse
(Unknown Source)
at com.sun.org.apache.xerces.internal.parsers.XMLParser.parse
(Unknown
Source)
at
com.sun.org.apache.xerces.internal.parsers.AbstractSAXParser.parse
(Unknown Source)
at com.sun.org.apache.xerces.internal.jaxp.SAXParserImpl
$JAXPSAXParser.parse(Unknown Source)
at javax.xml.parsers.SAXParser.parse(Unknown Source)
at org.mortbay.xml.XmlParser.parse(XmlParser.java:188)
at org.mortbay.xml.XmlParser.parse(XmlParser.java:204)
at
org.mortbay.jetty.webapp.TagLibConfiguration.configureWebApp
(TagLibConfiguration.java:227)
at org.mortbay.jetty.webapp.WebAppContext.startContext
(WebAppContext.java:1217)
at org.mortbay.jetty.handler.ContextHandler.doStart
(ContextHandler.java:513)
at org.mortbay.jetty.webapp.WebAppContext.doStart
(WebAppContext.java:
448)
at com.google.gwt.dev.shell.jetty.JettyLauncher
$WebAppContextWithReload.doStart(JettyLauncher.java:236)
at org.mortbay.component.AbstractLifeCycle.start
(AbstractLifeCycle.java:39)
at org.mortbay.jetty.handler.HandlerWrapper.doStart
(HandlerWrapper.java:130)
at org.mortbay.jetty.Server.doStart(Server.java:222)
at org.mortbay.component.AbstractLifeCycle.start
(AbstractLifeCycle.java:39)
at com.google.gwt.dev.shell.jetty.JettyLauncher.start
(JettyLauncher.java:283)
at com.google.gwt.dev.HostedMode.doStartUpServer
(HostedMode.java:368)
at com.google.gwt.dev.HostedModeBase.startUp
(HostedModeBase.java:587)
at com.google.gwt.dev.HostedModeBase.run(HostedModeBase.java:
394)
at com.google.gwt.dev.HostedMode.main(HostedMode.java:231)

On Feb 13, 8:26 pm, Scott Blum <sco...@google.com> wrote:
> Would people be happier if we called this the "official community-supported
> GWT/Maven integration"? :)
>

> On Fri, Feb 13, 2009 at 10:08 AM, Arthur Kalmenson <arthur.k...@gmail.com>wrote:
>
> > I think you'll note that GWT is in the central Maven repository
> > (http://repo1.maven.org/maven2/com/google/gwt/) and there are
> > currently at least two Maven plugins (being merged), the gwt-maven

> > plugin:http://code.google.com/p/gwt-maven/and the Codehaus GWT

[Jim Tomlinson]

Is com.google.gwt.dev.Compiler with this release supposed to run on
the Mac? Looks like library(s) are being shipped for it, but with OS X
10.5.6 and java 6 ("Java(TM) SE Runtime Environment (build 1.6.0_07-
b06-153)") I'm getting:

...
[INFO] Unpacking /Users/jimt/.m2/repository/com/google/gwt/gwt-dev/
1.6.0/gwt-dev-1.6.0-mac-libs.zipto
/Users/jimt/.m2/repository/com/google/gwt/gwt-dev/1.6.0
with Includes null and excludes:null
[INFO] [gwt:compile {execution: default}]
[INFO] establishing classpath list (buildClaspathList - scope =
COMPILE)
[INFO] google.webtoolkit.home (gwtHome) *not* set, using project POM
for GWT dependencies
Compiling module org.kuali.student.core.organization.ui.OrgEntry
Invalid memory access of location 00000000 rip=01160767

/Users/jimt/src/kuali/student/ks-core/branches/ks-core-dev/core-ui/
target/compile.sh: line 5: 6139 Segmentation fault "/System/
Library/Frameworks/JavaVM.framework/Versions/1.6.0/Home/bin/java" -
XstartOnFirstThread -Xmx512m -cp $CLASSPATH
com.google.gwt.dev.Compiler -gen /Users/jimt/src/kuali/student/ks-core/
branches/ks-core-dev/core-ui/target/.generated -logLevel INFO -style
OBF -war /Users/jimt/src/kuali/student/ks-core/branches/ks-core-dev/
core-ui/target/ks-core-ui-0.0.1-SNAPSHOT
org.kuali.student.core.organization.ui.OrgEntry
[INFO]
------------------------------------------------------------------------
[ERROR] BUILD ERROR
[INFO]
------------------------------------------------------------------------
[INFO] Exception attempting to run script - compile.sh

I saw this same issue on Ubuntu with 1.6.0_07. Upgrading to _12 fixed
the problem there, but there's no such upgrade avenue (yet) on OS X.
Known issue? Thanks.

- Jim Tomlinson

On Feb 6, 7:26 am, Scott Blum <sco...@google.com> wrote:
> Greetings GWT developers,
>
> The GWT team is happy to announce the availability of 1.6 Milestone 1!
> Binary distributions are available for download directly from GWT's Google
> Code project.
>
> http://code.google.com/p/google-web-toolkit/downloads/list?can=1&q=1.6.0

...

[Scott Blum]

What happens if you use a Java 1.5 VM?

[Jim Tomlinson]

Well, our project has lots of dependencies on java6, so there are
beaucoup failures, but it doesn't get a segment violation as it does
with 1.6.0_07. Guess I'll muddle along with my Ubuntu VM until the
next milestone. Here's hoping the Mac (and its java6 64-bit libraries)
get full support soon. Thanks.

- Jim Tomlinson

On Feb 18, 1:01 pm, Scott Blum <sco...@google.com> wrote:
> What happens if you use a Java 1.5 VM?
>

[Netsurfer]

Hi,
I was using GWT 1.5 since some weeks ago, now I moved my application
to GWT1.6M1 in order to use the new war structure with a Spring server
application context.
Everything works fine with HostedMode: this version simplifies a lot
web frameworks integration.

Now the problem:
I want to test my application with GwtTestCase but it uses JUnitShell
that extends GwtShell and there isn't any implementation like a
JUnitHostedMode that use the new war structure or, more in general,
the possibility to use HostedMode instead of GwtShell in a
GwtTestCase.
The consequence is that I'm not able to test my GWT Spring application
because every file inside the new war structure isn't reachable on a
GwtShell and the Spring context isn't loaded.

Is there any solution?

Thanks in advanced,
Marco

On 6 Feb, 16:26, Scott Blum <sco...@google.com> wrote:
> Greetings GWT developers,
>
> The GWT team is happy to announce the availability of 1.6 Milestone 1!
> Binary distributions are available for download directly from GWT's Google
> Code project.
>
> http://code.google.com/p/google-web-toolkit/downloads/list?can=1&q=1.6.0
>

[Scott Blum]

To be honest, we kind of punted on a full JUnitShell port for a couple of reasons:

1) Backwards compatibility.  If we updated JUnitShell to use HostedMode, it seems like old test cases would probably break.  The idea of a JUnitHostedMode is a good one... but how would we be able to tell which one the user meant, since the JUnit framework is in control?

2) It should work okay for pure client-side testing.  I take it from your message that you're doing something more sophisticated, that involves server-side code also?

[Arthur Kalmenson]

> 1) Backwards compatibility. If we updated JUnitShell to use HostedMode, it
> seems like old test cases would probably break. The idea of a
> JUnitHostedMode is a good one... but how would we be able to tell which one
> the user meant, since the JUnit framework is in control?

So does this mean that GWTTestCases still spawns off a headless hosted
mode that's using an embedded Tomcat and not an embedded Jetty?

--
Arthur Kalmenson

[Scott Blum]

On Fri, Feb 20, 2009 at 8:29 AM, Arthur Kalmenson <arthu...@gmail.com> wrote:

> 1) Backwards compatibility.  If we updated JUnitShell to use HostedMode, it
> seems like old test cases would probably break.  The idea of a
> JUnitHostedMode is a good one... but how would we be able to tell which one
> the user meant, since the JUnit framework is in control?

So does this mean that GWTTestCases still spawns off a headless hosted
mode that's using an embedded Tomcat and not an embedded Jetty?

That is correct.  It sounds like what I'm hearing though, is that there would be a lot of interest in us providing an explicit upgrade path, so that tests could be transitioned.  If we did this, it might even make sense to deprecate the old-style but keep it in for now for backwards compatibility.

[Arthur Kalmenson]

--
Arthur Kalmenson

>> So does this mean that GWTTestCases still spawns off a headless hosted
>> mode that's using an embedded Tomcat and not an embedded Jetty?
>
> That is correct. It sounds like what I'm hearing though, is that there
> would be a lot of interest in us providing an explicit upgrade path, so that
> tests could be transitioned. If we did this, it might even make sense to
> deprecate the old-style but keep it in for now for backwards compatibility.

OK, because my understanding was that the embedded Jetty will allow
hosted mode to start faster, which would be a great improvement for
GWTTestSuites/GWTTestCases. Thanks!

[Brian]

Migrated my app to 1.6, and am really liking the war structure and the
speed increase in booting. By migrating to the new war structure, I
just got rid of my local jetty server.

Anyway, I'm plowing through 200 warnings or so on deprecated usage,
and am just punting on one type. Basically I'm hoping someone tells
me what to do by the time I clear up everything else...

It's the conversion from ClickHandler to ClickEvent. How do I convert
the click handler on a button from a ClickListener to a ClickHandler,
and still compare on the widget?

convert:
public void onClick(Widget widget) {
if (widget == someWidget) {
// do something...
}
}

to:
public void onClick(ClickEvent event) {
// tell me what to write here, please :-)
}

[Isaac Truett]

// From memory, without running it past a compiler, I think it is:
Widget widget = (Widget) event.getSource();

[Brian]

Thanks. I should have read the webpage on doing the conversion before
posting.

Well that was really painful to go through and change the event
handlers to the new way. Wow there's a lot of @Deprecated stuff now.
It's nice that the javadocs include some info on the new way.

Cool though to finally delete all the stub methods from the Listeners
interfaces that weren't in use (onKeyDown, etc).

On Feb 24, 2:40 pm, Isaac Truett <itru...@gmail.com> wrote:
> // From memory, without running it past a compiler, I think it is:
> Widget widget = (Widget) event.getSource();
>