richfaces ajax-call and embedded gwt widget
Hello, This is actually a cross post from stackoverflow (http:// stackoverflow.com/questions/6574174/richfaces-ajax-call-and-embedded- gwt-widget). Wasn't sure where I could get a quicker suggestion, hope you don't mind :) I am using a4j:function call from richfaces library to rerender a segment on a website every 20 seconds. The problem is that there is a div element in this segment, that contains gwt widgets. So that everytime the richfaces ajax call is made and replaces the old html code I need to tell GWT that it has to reattach the widget (is a native call to GWT from within the rerendered area). This reattaching results in flickering behavior, you see the richfaces content first and in half a second GWT widgets appear. Does anyone know if there is a more elegant solution to make it user friendly (I cannot get rid of GWT/Richfaces mixture:)). I would like to have my GWT widgets be reattached first before the segment is displayed so that users dont see any blinking/resizing/flickering on the website. Guess, that there should be a hook in richfaces to call external code that would make changes to the new dom tree before it is displayed in the browser. Thank you. -- 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: com.google.gwt.user.client.rpc.IncompatibleRemoteServiceException
are you using external server and not embedded jetty. If so, you have
to deploy your module in exploded mode, that is not compressed. Then
point your devmode to the exploded folder on the server:
-war /home/${env_var:USER}/your-exploded-web.war
This allows using dev mode without having to recompile.
good luck
On 12 Feb., 20:06, Fernando Barbat wrote:
> Yes, it has a zero argument constructor. As I said, once I recompile,
> undeploy and deploy again, the problem goes away.
>
> I'm using Eclipse and deploy an EAR in Glasfish v3. This EAR contains
> an EJB jar, a GWT WAR and other jars. I noted that
> IncompatibleRemoteServiceException sometimes raises up after I change
> EJB code.
>
> It's not enough with undeploy and deploy. I also have to recompile to
> get everything working.
>
> Maybe it is a problem with packaging. Maybe the RPC generated security
> files are not getting into the EAR properly.
>
> Any idea about how to solve this? Any idea about how to disable that
> security feature?
>
> Thanks.
>
> On 11 feb, 16:58, nacho wrote:
>
>
>
>
>
>
>
> > I think that your trouble is that the classes compiled to JS are different
> > to the classes that you have in the server.
>
> > Using dev mode after change in a class that I use in a RPC I just reload the
> > page and the problem is gone.
--
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.
is using DivElement instead of Label recommended?
Hello,
I wonder if it is a usual practice to use Element objects instead of
Widgets in GWT to achieve better performacne. For instance, instead of
adding a list of labels to a HTMLPanel I could do the following:
HTMLPanel pnl = new HTMLPanel("myPanel");
DivElement div = DivElement.as(DOM.createDiv());
pnl.getElement().appendChild(div);
My motivation is that since these divs shouldn't act on any events
there is no reason to imposing more overhead and making them widgets.
That's what UiBinder at least does, as far as I understand.
I would have also used UiBinder instead, the problem is though that it
is a dynamic list of items changing at runtime and since can not be
created declaratively and statically?
Am I on the right path? What are possible repercussions, memory leaks
or maintainability? Thanks.
--
You received this message because you are subscribed to the Google Groups
"Google Web Toolkit" group.
To post to this group, send email to google-web-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.
how to add Element to HTMLPanel
His, I am trying to use as few widgets as possible with UiBinder. One template that is there are many instances of could be have DivElement as parent node, is there a way to add Elements to HTMLPanel after it is created. Am aware of the way to add Widget to HTMLPanel but not Element, maybe there is another panel that is as efficient as HTMLPanel - uses setInnerHTML() - and can dynamically add Element objects. Thanks. -- You received this message because you are subscribed to the Google Groups "Google Web Toolkit" group. To post to this group, send email to google-web-tool...@googlegroups.com. To unsubscribe from this group, send email to google-web-toolkit+unsubscr...@googlegroups.com. For more options, visit this group at http://groups.google.com/group/google-web-toolkit?hl=en.
When to use UiBinder and CellList instead of widgets and standard layout panels?
His, I am trying to optimize performance for a mobile gwt app. We have lots of HTMLPanels to layout widgets, and our speculation is that they may be causes of really bad sluggishness. Since there are regular updates to GUI we use add() method of HTMLPanel to render updated sections anew. Would switching to the new cell layout panels alleviate some of these performance pains. I am not quite getting if CellList is supposed to perform better in layouting components than standard panels. As we started finding ways to optimize for performance I have tried replacing some of the widgets with UiBinder templates + owning classes, but no real performance gains have been made. We are surely doing something wrong, could it be that you only gain performance improvements only if templates are as static as possible and UiFields are simplest elements as opposed to widgets. I am guessing that if UiFields in templates are layout panels or widgets (buttons in our case, mostly) than it cannot really use setInnerHtml to boost up performance and uses regular DOM manipulations instead. It would be great to hear some feedback on these two points. Is anyone aware of any opensource rich-data GWT project to be able to peek into its codebase? Thanks -- You received this message because you are subscribed to the Google Groups "Google Web Toolkit" group. To post to this group, send email to google-web-tool...@googlegroups.com. To unsubscribe from this group, send email to google-web-toolkit+unsubscr...@googlegroups.com. For more options, visit this group at http://groups.google.com/group/google-web-toolkit?hl=en.
Re: Recommendations for improving performance of a mobile application
Thanks. I am aware of the feature. But i think it's not an appropriate solution in the particular case. The thing is that all the client code is loaded on startup the only thing that happens when user changes pages is that new DTOs are sent over and the existing views (which are already present on client) are rendered anew to reflect new data. This rendering and attaching views to panels and to each other takes so much time. I have peeked into Google Wave html code. It's very clean and uses mostly divs, sparsely tables. I wonder what they are using to layout data: HorizontalPanel, HTMLPanel, Cell Layout, etc? As far as I know client code is of the moment not open sourced. On 12 Okt., 10:11, Jin wrote: > Denis, > > Not sure if you're already aware of this or whether it will help your > situation: There is a technique called code splitting where you split > up a GWT app into smaller components so that the visible components > load faster. > > Other components are loaded as needed or in the background after > visible components have been loaded. > > http://code.google.com/webtoolkit/doc/latest/DevGuideCodeSplitting.html > > Haven't tried it, but sounds promising. > > Cheers, > > Jin > > On Oct 12, 9:01 am, "Armishev, Sergey" wrote: > > > I am not specifically developing for mobile device (I am doing Network > > Management) and not consider myself guru in the mobile Web apps. But I did > > pretty good testing regarding performance of my GWT apps on mobile devices > > to allow our customers use their smart phones for network management. What > > I found that performance greatly depends on performance of Javascript > > engine running inside mobile browser. On my Android my alarm management app > > was running probably with the same performance as on my desktop. No visible > > performance degradation. On Windows Mobile 6.5 with default "Pocket > > Explorer" browser my app was just frozen! Only after installing Opera > > browser it started to run without visible performance problems. Just to let > > you know that I am using long polling to achieve real time visualization > > and operators can exchange instant messaging through the same GWT RPC calls. > > I can recommend to test your app on Android first to see the best > > performance and then go to other mobile devices > > > -Sergey > > > -Original > > Message-From:google-web-tool...@googlegroups.com[mailto:google-web-tool...@googlegroups.com]on > > Behalf Of denis56 > > Sent: Monday, October 11, 2010 4:44 PM > > To: Google Web Toolkit > > Subject: Re: Recommendations for improving performance of a mobile > > application > > > Hi, > > > We already do that. Only changed objects are sent over the wire. There > > is of course question of granularity, but i think we have struck a > > good balance between complexity and efficiency. Are you also > > developing for mobile? > > > On 11 Okt., 22:01, "Armishev, Sergey" wrote: > > > Regarding slow RPC calls. Just try to send incremental data changes, > > > only the data that been changed/added/deleted and cache main data on > > > client. You might have thousands of items but only few of them or > > > nothing been changed between RPC calls. I usually send full update when > > > number of changes exceeds certain critical level when cost of update is > > > higher then full update. Another hint is to change only data that > > > visible to user. Usually there are not much visible data particularly on > > > mobile device. > > > > -Sergey > > > > -Original Message-> From:google-web-toolkit@googlegroups.com > > > > [mailto:google-web-tool...@googlegroups.com]on Behalf Of denis56 > > > Sent: Monday, October 11, 2010 2:27 PM > > > To: Google Web Toolkit > > > Subject: Recommendations for improving performance of a mobile > > > application > > > > Hello to everyone, > > > > We decided to give it a try programming a mobile application with GWT > > > - frequent RPC updates, very component-oriented and frequently > > > changing data. At the moment we try to meet bearable performance > > > benchmarks since current performance is so dismal. To change to a > > > different page (fully dynamic) takes more than a minute at times ; > > > ( Clicking on clickable panels, which are present in large quantities > > > on every page, is also excruciatingly slow. > > > > I have watched videos from Google I/O Conference 2010 (Architecting > > > for performance with GWT, GWT's UI
Re: Recommendations for improving performance of a mobile application
Hi, We already do that. Only changed objects are sent over the wire. There is of course question of granularity, but i think we have struck a good balance between complexity and efficiency. Are you also developing for mobile? On 11 Okt., 22:01, "Armishev, Sergey" wrote: > Regarding slow RPC calls. Just try to send incremental data changes, > only the data that been changed/added/deleted and cache main data on > client. You might have thousands of items but only few of them or > nothing been changed between RPC calls. I usually send full update when > number of changes exceeds certain critical level when cost of update is > higher then full update. Another hint is to change only data that > visible to user. Usually there are not much visible data particularly on > mobile device. > > -Sergey > > > > -Original Message- > From: google-web-toolkit@googlegroups.com > > [mailto:google-web-tool...@googlegroups.com] On Behalf Of denis56 > Sent: Monday, October 11, 2010 2:27 PM > To: Google Web Toolkit > Subject: Recommendations for improving performance of a mobile > application > > Hello to everyone, > > We decided to give it a try programming a mobile application with GWT > - frequent RPC updates, very component-oriented and frequently > changing data. At the moment we try to meet bearable performance > benchmarks since current performance is so dismal. To change to a > different page (fully dynamic) takes more than a minute at times ; > ( Clicking on clickable panels, which are present in large quantities > on every page, is also excruciatingly slow. > > I have watched videos from Google I/O Conference 2010 (Architecting > for performance with GWT, GWT's UI overhaul, Faster apps faster) and > tried replacing some of the widgets with UiBinder. This had so far > brought only minimal performance improvements. There are some critical > observations I would like to share and maybe you would have valuable > suggestions > > * One of the things Joel Webber tells is that widgets are slow. > Unfortunately everything is a widget in our app: clickable panels for > selection and HTMLPanels/FlowPanels to layout clickables. There is > very little space to cut off on widget. What would be the most > lightweight component to implement onMouseDown-Event (a Button or > clickable FlowPanel)? Is it worth the effort to replace all HTMLPanel/ > FlowPanel possible with new Cell Widgets (I am thinking about CellList > in particular). Are they plain faster for layouting components? > > * Running Firebug Profiler and Speedtracer shows that PRC requests > take the most time. On page change, for example, an PRC call would > return a rather big list of DTOs - they have hierarchical logical > structure - and each is then injected into corresponding view that > attaches itself to parent view for display. So profiler says that > add()-calls take the most time. But I can hardly imagine a way to get > rid of these add calls because the views have to be attached to the > owning views. Is there a better strategy to display lists of lists of > data? > > * We use tables for layouting and in the process of converting that > into divs, as well as converting syles to eliminate style chaining. > Could that improve performance of mobile browsers significantly? One > of the insights Speedracer gave is that parsing HTML and recalculating > CSS styles takes a whole lot of time. > > Would really appreciate every suggestion. Thanks, > Denis > > -- > You received this message because you are subscribed to the Google > Groups "Google Web Toolkit" group. > To post to this group, send email to > google-web-tool...@googlegroups.com. > To unsubscribe from this group, send email to > google-web-toolkit+unsubscr...@googlegroups.com. > For more options, visit this group > athttp://groups.google.com/group/google-web-toolkit?hl=en. > > style='font-size:8.0pt;font-family:"Arial","sans-serif";color:#003366'> > _ > This electronic message and any files transmitted with it contains > information from iDirect, which may be privileged, proprietary > and/or confidential. It is intended solely for the use of the individual > or entity to whom they are addressed. If you are not the original > recipient or the person responsible for delivering the email to the > intended recipient, be advised that you have received this email > in error, and that any use, dissemination, forwarding, printing, or > copying of this email is strictly prohibited. If you received this email > in error, please delete it and immediately notify the sender. > _
Recommendations for improving performance of a mobile application
Hello to everyone, We decided to give it a try programming a mobile application with GWT - frequent RPC updates, very component-oriented and frequently changing data. At the moment we try to meet bearable performance benchmarks since current performance is so dismal. To change to a different page (fully dynamic) takes more than a minute at times ; ( Clicking on clickable panels, which are present in large quantities on every page, is also excruciatingly slow. I have watched videos from Google I/O Conference 2010 (Architecting for performance with GWT, GWT's UI overhaul, Faster apps faster) and tried replacing some of the widgets with UiBinder. This had so far brought only minimal performance improvements. There are some critical observations I would like to share and maybe you would have valuable suggestions * One of the things Joel Webber tells is that widgets are slow. Unfortunately everything is a widget in our app: clickable panels for selection and HTMLPanels/FlowPanels to layout clickables. There is very little space to cut off on widget. What would be the most lightweight component to implement onMouseDown-Event (a Button or clickable FlowPanel)? Is it worth the effort to replace all HTMLPanel/ FlowPanel possible with new Cell Widgets (I am thinking about CellList in particular). Are they plain faster for layouting components? * Running Firebug Profiler and Speedtracer shows that PRC requests take the most time. On page change, for example, an PRC call would return a rather big list of DTOs - they have hierarchical logical structure - and each is then injected into corresponding view that attaches itself to parent view for display. So profiler says that add()-calls take the most time. But I can hardly imagine a way to get rid of these add calls because the views have to be attached to the owning views. Is there a better strategy to display lists of lists of data? * We use tables for layouting and in the process of converting that into divs, as well as converting syles to eliminate style chaining. Could that improve performance of mobile browsers significantly? One of the insights Speedracer gave is that parsing HTML and recalculating CSS styles takes a whole lot of time. Would really appreciate every suggestion. Thanks, Denis -- You received this message because you are subscribed to the Google Groups "Google Web Toolkit" group. To post to this group, send email to google-web-tool...@googlegroups.com. To unsubscribe from this group, send email to google-web-toolkit+unsubscr...@googlegroups.com. For more options, visit this group at http://groups.google.com/group/google-web-toolkit?hl=en.
Could it be that GWT internally modifies cookie?
His to everyone, We are experiencing a wicked behavior on the web page that uses one GWT component running regular updates. Maybe someone could provide an insight. The thing is that after a couple of PRC requests you can see in Firebug that the sessionid sent to client on previous response does not correspond to the sessionid that is sent on subsequent PRC request. As a side effect users get wrong session cookie and resulting in unintended log outs. Really strange and happens only irregularly. On of the suspicions is that GWT might internally manipulate sessions (we are using GWT 2.0) because no cookies are set by us explicitly? Thanks. -- You received this message because you are subscribed to the Google Groups "Google Web Toolkit" group. To post to this group, send email to google-web-tool...@googlegroups.com. To unsubscribe from this group, send email to google-web-toolkit+unsubscr...@googlegroups.com. For more options, visit this group at http://groups.google.com/group/google-web-toolkit?hl=en.
is it possible & how to debug productive code?
His,
I wonder if it is possible to debug a gwt site running on a productive
webserver. Say there is a website
"example.com" with some issues that you want to inspect real-time and
in-production through a DevMode.
I created following eclipse launch directive
-noserver -whitelist .* -war /home/${env_var:USER}/jboss/server/
default/deploy/example-war -bindAddress 0.0.0.0 -codeServerPort 9993 -
startupUrl
http://example.com:8080/example.html
xyz.Example
So I basically using local codebase and want to debug against a
compiled and deployed instance running on a remote (not in the LAN)
webserver.
The browser starts with the following url
http://example.com:8080/example.html?gwt.codesvr=192.168.0.2:9993 but
I don't see connection tab being opened in the DevMode console.
Could it be that I have messed it up with ports and ip addresses?
thanks
--
You received this message because you are subscribed to the Google Groups
"Google Web Toolkit" group.
To post to this group, send email to google-web-tool...@googlegroups.com.
To unsubscribe from this group, send email to
google-web-toolkit+unsubscr...@googlegroups.com.
For more options, visit this group at
http://groups.google.com/group/google-web-toolkit?hl=en.
Re: IncompatibleRemoteServiceException with -noserver argument in gwt 2.0
Thanks a lot for your help. I figured out that you need to deploy in exploded war when using an external server, since on every start and page refresh DevMode would generate a new rpc policy file in my project which was actually missing in the war on the server and since the problem. On 28 Jan., 17:02, Rajeev Dayal wrote: > Hi, > > If you want to pick up changes to your server code while the development > server is running, you need to deploy your application in exploded-war > format to JBoss. > > When you run GWT in -noserver mode, you should point your -war directory > over to your JBoss server's deployment directory - that way, any files > generated by GWT will be dumped into your running server's directory. > > Unfortunately, there is a bug in GWT's RPC system that will prevent the > server from picking up changes to the RPC serialization policy file, even if > you copy the modified serialization file into the running server's > deployment directory: > > http://code.google.com/p/google-web-toolkit/issues/detail?id=4548 > > I'd recommend that you star this issue if you want it to be prioritized. > > Rajeev > > On Thu, Jan 14, 2010 at 12:38 PM, denis56 wrote: > > > His, > > > I have been struggling with > > com.google.gwt.user.client.rpc.IncompatibleRemoteServiceException > > (which finds several entries in web) > > for a while and was able to find a clumsy workaround. Could anyone > > please suggest a solution in the following case as I believe it may > > be either my configuration error or gwt bug. > > > Here is a sample project structure to refer to: > > - java > > - modules > > | -- gwtclient > > > There is an ant task that compiles gwt code and puts it generated code > > under modules/gwtclient, this folder is than zipped as war file and > > deployed to jboss. It all runs then smoothly in a browser. > > > However should I start DevMode with "-noserver" option and the "-war" > > option pointing at NOT PACKAGED directory modules/gwtclient in the > > project structure, the following exception > > "com.google.gwt.user.client.rpc.IncompatibleRemoteServiceException: > > Type 'XYZ' was not assignable to > > 'com.google.gwt.user.client.rpc.IsSerializable' and did not have a > > custom field serializer. For security purposes, this type will not be > > deserialized." > > will be thrown. > > I noticed that on every start DevMode creates new .gwt.rpc file under > > modules/gwtclient (and not to deployed war) and that the exception > > disappears if the modules/gwtclient is packaged and redeployed as > > DevMode runs. > > > From > >http://groups.google.com/group/google-web-toolkit/browse_thread/threa... > > I understand that the server (jboss' tomcat?) cannot find > > suitable .gwt.rpc file and falls back to LegacySerializationPolicy > > which demands IsSerializable interface. > > > Shouldn't DevMode running with -noserver argument be able create > > new .gwt.prc files direct in the war file deployed to jboss or not > > create it at all? > > Is this a bug or my DevMode configuration mistake? > > > Thanks > > > -- > > You received this message because you are subscribed to the Google Groups > > "Google Web Toolkit" group. > > To post to this group, send email to google-web-tool...@googlegroups.com. > > To unsubscribe from this group, send email to > > google-web-toolkit+unsubscr...@googlegroups.com > > . > > For more options, visit this group at > >http://groups.google.com/group/google-web-toolkit?hl=en. -- You received this message because you are subscribed to the Google Groups "Google Web Toolkit" group. To post to this group, send email to google-web-tool...@googlegroups.com. To unsubscribe from this group, send email to google-web-toolkit+unsubscr...@googlegroups.com. For more options, visit this group at http://groups.google.com/group/google-web-toolkit?hl=en.
Re: problems running application in development mode using -noserver switch and cache.html/nocache.js files not served from the webapp root
do you get a pop-up message that your browser cannot connect to hosted mode? if yes, it might help setting "-bindAddress 0.0.0.0" to your host (source: http://groups.google.com/group/google-web-toolkit/browse_thread/thread/e292666b63d182b3/aa53947cd7525cda?lnk=gst&q=2.0.1#aa53947cd7525cda) did you download gwt-plugin for firefox? can you try it out in another browser with gwt plug-in? On 7 Feb., 20:52, mooreds wrote: > Hi folks, > > I'm trying to get GWT2.0.1development mode working with the - > noserver option. The main issue is that while running in development > mode, any changes I make in the GWT java classes are not reflected > when I refresh the browser. > > This is a bit of a nonstandard setup, so let me outline it a bit. > > We have a number of apps that depend on some other cvs modules: > > * gwtapp1 > * gwtapp2 > * gwtlib1 > * gwtlib2 > > All of these depend on code running on the server (json files, RPC) > and I couldn't figure out how to get them working with the built in > Jetty server. (For one thing, I couldn't figure out how to have the > different projects all compile into one WEB-INF/classes directory.) > > gwtapp1 compiles the GWT into the root directory, and works just fine. > > gwtapp2 compiles the GWT into /static/gwt/ One other difference that > may be relevant is that gwtapp2 uses the cross site linker. However, > removing that line from the .gwt.xml file didn't seem to make a > difference in the behavior. Another difference is the directory that > gwtapp2 compiles into actually has 3 different modules in it (but I'm > not trying to touch any of the other modules). > > It works fine in 'production' mode, when built via ant. But it > doesn't work in development mode. > > Here's the arguments I'm using for the eclipse launcher: > > -noserver -startupUrlhttp://localhost:8080/HomePage.do > com.foo.gwtapp2 (I've tried different URLs, including a static html > page) > > The GWT development mode window pops up just fine, but I never see the > 'FF' tab pop up when I click 'launch default browser'. I do see this > url in the browser window: > http://localhost:8080/HomePage.do?gwt.codesvr=192.168.3.103:9997 > > I see no messages in the development mode window. However, when I > turn the logLevel up to DEBUG (-logLevel DEBUG) I see the 'Loading > Modules' line. When I am working with gwtapp1 (which works) I also > see a 'Connection received from xxx.xxx.xxx.xxx' message. > > Before the 'Connection recieved', I see the following TRACE messages > when running either app: > > 00:00:02.078 [TRACE] Invoking Linker RPC policy file manifest > 00:00:02.078 [TRACE] Invoking Linker Standard > 00:00:02.094 [DEBUG] Attempting to optimize JS > 00:00:02.141 [TRACE] Invoking Linker Export CompilationResult > symbol maps > 00:00:02.141 [TRACE] Invoking Linker Emit compile report > artifacts > 00:00:02.141 [TRACE] Linking compilation into C:\eclipse-workspace > \account\war\com.foo.gwtapp[12] > > I imagine that the Connection is the issue, but am not quite sure how > to debug it. > > I've installed the Google FF plugin, and am using FF 3.6. This is an > old project so it was not created with the google eclipse plugin. > Running on eclipse 3.4.2 on Windows XP, if that matters. > > I've reviewed this > FAQ:http://code.google.com/webtoolkit/doc/latest/FAQ_DebuggingAndCompilin... > > and made sure that I have all the .rpc files in the /static/gwt > directory. This directory is browser accessible (ie, when I put a > file here:http://localhost:8080/static/gwt/a.txt, I can read it with > my browser). > > I also played around with the -codeServerPort argument, but that > didn't seem to make any difference. In the thought that it was > perhaps a urlrewrite issue (gwtapp2 does some url rewriting) I put a > static html file in the /static/gwt directory, but that still didn't > lead to a Connection being made. > > I have done some searching on the web and in the GWT google group, but > haven't found much else. I also didn't find anything when searching > through the bug list. > > What am I missing? Can anyone give me further places I should look or > ideas? > > Thanks, > Dan -- You received this message because you are subscribed to the Google Groups "Google Web Toolkit" group. To post to this group, send email to google-web-tool...@googlegroups.com. To unsubscribe from this group, send email to google-web-toolkit+unsubscr...@googlegroups.com. For more options, visit this group at http://groups.google.com/group/google-web-toolkit?hl=en.
Re: Problem running application GWT 2.0 with external server and DevMode, unable to find module file
see another post http://groups.google.com/group/google-web-toolkit/browse_thread/thread/1ac1304e312e7290 make sure that src folder is in your classpath in eclipse "In Classpath tab: under User Entries add your source folder. " On 15 Jan., 13:06, mlopez wrote: > Hello Chris, > > Firstly I start the Tomcat web server at localhost:8080, > my launch configuration is all that I write in the first post: Run -> > Run configurations -> > > Java Application: > in the Main tab as Main class I'm using: com.google.gwt.dev.DevMode, > and as Project Iam > > using my GWT project: myapp. > in the Arguments tab as Program Arguments I'm using the previous ones: > > -noserver > -war "C:\workspace\.metadata\.plugins\org.eclipse.wst.server.core > \tmp0\wtpwebapps\myapp\gwt" > -startupUrlhttp://localhost:8080/MyAppWeb/gwt/com.xxx.web.MyApp/MyApp.html > com.xxx.web.client.MyApp > > and as VM arguments I'm using: -Xmx256M > > That's my entire launch configuration. > > And as I said in the first post after the GWT Development Mode window > starts I get the next error: > > Unable to find com.xxx.web.client.MyApp.gwt.xml on your classpath; > could be a typo, or maybe > you forgot to include a classpath entry for source? > > I'm not sure if what you want is my module gwt.xml file, is this one: > > > "http://google-web-toolkit.googlecode.com/svn/trunk/distro-source/core/ > src/gwt-module.dtd"> > > name='com.xxx.architecture.commons.gwt.web.ItxCommonsGwt' /> > name='com.xxx.compras.comunes.gwt.precompras.web.CVPComunesPreCompras' / > > > > > > > > > > > > > > > > > > > > > The problem is only launching in DevMode, being as if I go to the > startupUrl from a web navigator I can access to the web application > without any problem, but I can't debug my web application. > > I also have tried launch the web application using GWT Web App launch > config but I have another problem when the application calls to any > RPC service, I have created other topic for this > problem:http://groups.google.com/group/google-web-toolkit/browse_thread/threa... > > Thank you for your interest and request me all the information that > you could need. > > Marcos Lopez. -- You received this message because you are subscribed to the Google Groups "Google Web Toolkit" group. To post to this group, send email to google-web-tool...@googlegroups.com. To unsubscribe from this group, send email to google-web-toolkit+unsubscr...@googlegroups.com. For more options, visit this group at http://groups.google.com/group/google-web-toolkit?hl=en.
IncompatibleRemoteServiceException with -noserver argument in gwt 2.0
His, I have been struggling with com.google.gwt.user.client.rpc.IncompatibleRemoteServiceException (which finds several entries in web) for a while and was able to find a clumsy workaround. Could anyone please suggest a solution in the following case as I believe it may be either my configuration error or gwt bug. Here is a sample project structure to refer to: - java - modules | -- gwtclient There is an ant task that compiles gwt code and puts it generated code under modules/gwtclient, this folder is than zipped as war file and deployed to jboss. It all runs then smoothly in a browser. However should I start DevMode with "-noserver" option and the "-war" option pointing at NOT PACKAGED directory modules/gwtclient in the project structure, the following exception "com.google.gwt.user.client.rpc.IncompatibleRemoteServiceException: Type 'XYZ' was not assignable to 'com.google.gwt.user.client.rpc.IsSerializable' and did not have a custom field serializer. For security purposes, this type will not be deserialized." will be thrown. I noticed that on every start DevMode creates new .gwt.rpc file under modules/gwtclient (and not to deployed war) and that the exception disappears if the modules/gwtclient is packaged and redeployed as DevMode runs. >From >http://groups.google.com/group/google-web-toolkit/browse_thread/thread/d026ef687c504eef# I understand that the server (jboss' tomcat?) cannot find suitable .gwt.rpc file and falls back to LegacySerializationPolicy which demands IsSerializable interface. Shouldn't DevMode running with -noserver argument be able create new .gwt.prc files direct in the war file deployed to jboss or not create it at all? Is this a bug or my DevMode configuration mistake? Thanks -- You received this message because you are subscribed to the Google Groups "Google Web Toolkit" group. To post to this group, send email to google-web-tool...@googlegroups.com. To unsubscribe from this group, send email to google-web-toolkit+unsubscr...@googlegroups.com. For more options, visit this group at http://groups.google.com/group/google-web-toolkit?hl=en.
Re: dynamic Messages/ImageBundle possible?
Thanks a lot,
that's indeed what i needed. It took me sometime to realize what the
requirement to put Messages.properties files in "classpath root"
really means. Is there a possibility to have Messages.properties
reside within the same package where the code using it (maybe though
an interface), since it makes it somewhat uncomfortable the way it
works now, more even so when you have several gwt modules in a single
project that should use difference bundles.
thanks
On Nov 25, 10:12 pm, kgeri wrote:
> [Not sure if my previous post went through, so...
> ]
>
> For dynamic (string-accessible) message support take a look at
> this:http://code.google.com/p/gwt-dmesg
>
> I wrote it because we've faced the same problem a million times now in
> our GWT projects and it was annoying :)
>
> Hope it helps, comments and ideas are welcome!
>
> On nov. 24, 23:36, denis56 wrote:
>
> > Oh, thanks for the hint.
>
> > Are you aware of the similar class for Messages?
>
> > On 23 Nov., 16:44, "Dean S. Jones" wrote:
>
> > > this might help a little
>
> > >http://code.google.com/p/asmodaiosgwt/
>
> > > On Nov 23, 10:40 am, denis56 wrote:
>
> > > > Hi there,
>
> > > > I wonder if it is possible to use Internationalization and
> > > > ImageBundles in GWT in a more dynamic way. For instance, if there is a
> > > > client method
>
> > > > void createSportTitleAndImage(String sportId) {
>
> > > > }
>
> > > > is it somehow possible to have the "sportId" argument being
> > > > dynamically plugged into the relevant ImageBundle/Messages method to
> > > > return appropriate code without the if/else-jungle.
>
> > > > So instead of:
>
> > > > void createSportTitleAndImage(String sportId) {
> > > > if (sport.equals("basketball"))
> > > > ImageBundle.getBasketballImage();
> > > > Messages.getBasketballText();
> > > > else if (sport.equals("handball"))
> > > > ImageBundle.getHandballImage();
> > > > Messages.getHandballText();
> > > > else ...
>
> > > > }
>
> > > > to have something like :
>
> > > > void createSportTitleAndImage(String sportId) {
> > > > ImageBundle.getSportImageById(sportId);
> > > > Messages.getSportTextById(sportId);
>
> > > > }
>
> > > > where getSportByName(sportId) would fetch resource for the key "base +
> > > > sportId" ?
>
> > > > That would really simplify things:)
> > > > Thanks
--
You received this message because you are subscribed to the Google Groups
"Google Web Toolkit" group.
To post to this group, send email to google-web-tool...@googlegroups.com.
To unsubscribe from this group, send email to
google-web-toolkit+unsubscr...@googlegroups.com.
For more options, visit this group at
http://groups.google.com/group/google-web-toolkit?hl=en.
Re: dynamic Messages/ImageBundle possible?
Oh, thanks for the hint.
Are you aware of the similar class for Messages?
On 23 Nov., 16:44, "Dean S. Jones" wrote:
> this might help a little
>
> http://code.google.com/p/asmodaiosgwt/
>
> On Nov 23, 10:40 am, denis56 wrote:
>
>
>
> > Hi there,
>
> > I wonder if it is possible to use Internationalization and
> > ImageBundles in GWT in a more dynamic way. For instance, if there is a
> > client method
>
> > void createSportTitleAndImage(String sportId) {
>
> > }
>
> > is it somehow possible to have the "sportId" argument being
> > dynamically plugged into the relevant ImageBundle/Messages method to
> > return appropriate code without the if/else-jungle.
>
> > So instead of:
>
> > void createSportTitleAndImage(String sportId) {
> > if (sport.equals("basketball"))
> > ImageBundle.getBasketballImage();
> > Messages.getBasketballText();
> > else if (sport.equals("handball"))
> > ImageBundle.getHandballImage();
> > Messages.getHandballText();
> > else ...
>
> > }
>
> > to have something like :
>
> > void createSportTitleAndImage(String sportId) {
> > ImageBundle.getSportImageById(sportId);
> > Messages.getSportTextById(sportId);
>
> > }
>
> > where getSportByName(sportId) would fetch resource for the key "base +
> > sportId" ?
>
> > That would really simplify things:)
> > Thanks
--
You received this message because you are subscribed to the Google Groups
"Google Web Toolkit" group.
To post to this group, send email to google-web-tool...@googlegroups.com.
To unsubscribe from this group, send email to
google-web-toolkit+unsubscr...@googlegroups.com.
For more options, visit this group at
http://groups.google.com/group/google-web-toolkit?hl=en.
is the next version going to be 2.0?
His, One more question. I am in the midst of one project development running on gwt1.7. Is the next version going to be a jump to 2.0. Would you recommend switching to 2.0 code base right away ? thanks -- You received this message because you are subscribed to the Google Groups "Google Web Toolkit" group. To post to this group, send email to google-web-tool...@googlegroups.com. To unsubscribe from this group, send email to google-web-toolkit+unsubscr...@googlegroups.com. For more options, visit this group at http://groups.google.com/group/google-web-toolkit?hl=.
dynamic Messages/ImageBundle possible?
Hi there,
I wonder if it is possible to use Internationalization and
ImageBundles in GWT in a more dynamic way. For instance, if there is a
client method
void createSportTitleAndImage(String sportId) {
}
is it somehow possible to have the "sportId" argument being
dynamically plugged into the relevant ImageBundle/Messages method to
return appropriate code without the if/else-jungle.
So instead of:
void createSportTitleAndImage(String sportId) {
if (sport.equals("basketball"))
ImageBundle.getBasketballImage();
Messages.getBasketballText();
else if (sport.equals("handball"))
ImageBundle.getHandballImage();
Messages.getHandballText();
else ...
}
to have something like :
void createSportTitleAndImage(String sportId) {
ImageBundle.getSportImageById(sportId);
Messages.getSportTextById(sportId);
}
where getSportByName(sportId) would fetch resource for the key "base +
sportId" ?
That would really simplify things:)
Thanks
--
You received this message because you are subscribed to the Google Groups
"Google Web Toolkit" group.
To post to this group, send email to google-web-tool...@googlegroups.com.
To unsubscribe from this group, send email to
google-web-toolkit+unsubscr...@googlegroups.com.
For more options, visit this group at
http://groups.google.com/group/google-web-toolkit?hl=.
chrome/safari error
His I am getting the following error in Chrome (and Safari also doesn't work; Firefox and IE are ok) "TypeError: Object # has no method 'write'" Same problem was reported in http://groups.google.com/group/google-web-toolkit/browse_frm/thread/e84e07a4c36f91db/593492d3689bf731?lnk=gst&q=gwt++Object+%23%3Ca+Document%3E+has+no+method+%27write%27#593492d3689bf731 I am also using JSF with RichFaces and the code works without RichFaces code. Has anyone idea/experience of solving the issue? Many thanks, Denis --~--~-~--~~~---~--~~ 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: rebuilding gwt 1.7
His, thanks for your replies. I think I am missing it. Have got Compiling module myModule [java]Refreshing module from source [java] Validating newly compiled units [java] Removing units with errors [java] [ERROR] Errors in 'jar:file:/home/share/gwt- linux-1.7.0/gwt-user.jar!/com/google/gwt/user/client/impl/ HTTPRequestImplIE6.java' [java][ERROR] Line 18: The import user cannot be resolved [java][ERROR] Line 19: Syntax error on token "super", Identifier expected [java][ERROR] Line 27: JavaScriptObject cannot be resolved to a type [java][ERROR] Errors in 'jar:file:/home/share/gwt-linux-1.7.0/ gwt-user.jar!/com/google/gwt/user/client/impl/HTTPRequestImplIE6.java' [java] [ERROR] Line 18: The import user cannot be resolved [java] [ERROR] Line 19: Syntax error on token "super", Identifier expected [java] [ERROR] Line 27: JavaScriptObject cannot be resolved to a type [java][ERROR] Cannot proceed due to previous errors after I replaced HTTPRequestImplIE6.java in gwt1.7 jar with the one I' ve changed Tried also recompiling the whole gwt 1.7, but seeing another message: compile: [mkdir] Created dir: /home/gwt_1.7_svn/build/out/user/bin [gwt.javac] Compiling 802 source files to /home/gwt_1.7_svn/build/out/ user/bin [gwt.javac] /home/gwt_1.7_svn/user/src/com/google/gwt/user/client/impl/ HTTPRequestImplIE6.java:19: expected [gwt.javac] import user.super.com.google.gwt.emul.java.lang.Override; [gwt.javac] ^ [gwt.javac] /home/gwt_1.7_svn/user/src/com/google/gwt/user/client/impl/ HTTPRequestImplIE6.java:19: class, interface, or enum expected [gwt.javac] import user.super.com.google.gwt.emul.java.lang.Override; [gwt.javac] ^ [gwt.javac] /home//gwt_1.7_svn/user/src/com/google/gwt/user/client/ impl/HTTPRequestImplIE6.java:19: class, interface, or enum expected [gwt.javac] import user.super.com.google.gwt.emul.java.lang.Override; [gwt.javac] ^ [gwt.javac] 3 errors what am I doing wrong? thanks On 17 Aug., 12:29, Thomas Broyer wrote: > On 17 août, 10:53, denis56 wrote: > > > has anybody tried recompiling gwt 1.7? > [...] > > I would like to recompile it to > > resolvehttp://code.google.com/p/google-web-toolkit/issues/detail?id=3608for > > IE 6, Chrome and Safari. > > You do not have to recompile GWT if you only touch "client > code" (given that this is code that'll be compiled to JavaScript from > the Java *source*). > > You can either patch your gwt-user.jar (extract and modify the > appropriate file, and repackage it within the JAR) or provide your own > copied/modified version as an "overlay" (just put it earlier in the > classpath). --~--~-~--~~~---~--~~ 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 -~--~~~~--~~--~--~---
rebuilding gwt 1.7
has anybody tried recompiling gwt 1.7? under what location current gwt 1.7 distribution is stored, http://code.google.com/p/google-web-toolkit/source/browse/#svn/releases/1.7 ? I would like to recompile it to resolve http://code.google.com/p/google-web-toolkit/issues/detail?id=3608 for IE 6, Chrome and Safari. thanks --~--~-~--~~~---~--~~ You received this message because you are subscribed to the Google Groups "Google Web Toolkit" group. To post to this group, send email to Google-Web-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: location for images
Thanks,
That's just what I wanted to hear.
On 11 Aug., 17:32, Sean wrote:
> I had problems with this too.
>
> Regular images, that are NOT in an Image Bundle can go in your war
> file. I typically keep a WAR/images folder and drop everything in
> there. Then a reference to them would be Image test = new Image(images/
> test.png); if the image file is WAR/images/test.png
>
> ImageBundle pictures have to be kept in the code files. It's kind of
> confusing, but essentially these pictures get compiled using the code
> and put into one single image. They will be kept somewhere in (typical
> folder structure) com/client/[AppName]/...
>
> An example from the API:
>
> �...@resource("com/client/myapp/icons/btn_cancel_icon.png")
> public AbstractImagePrototype cancelButtonIcon();
>
> Hope this helps.
>
> On Aug 10, 5:55 am, denis56 wrote:
>
>
>
> > His,
>
> > In GWT 1.5 one needed to put images under public directory, with gwt
> > 1.6/1.7 the recommended location is somewhere under your war
> > directory. I use normal images as well as ImageBundle in my project.
> > Could someone please clarify where image bundles are supposed to go
> > now - stay under public directory or go together with other non-
> > imagebundle files out to war?
>
> > Thannks
--~--~-~--~~~---~--~~
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
-~--~~~~--~~--~--~---
location for images
His, In GWT 1.5 one needed to put images under public directory, with gwt 1.6/1.7 the recommended location is somewhere under your war directory. I use normal images as well as ImageBundle in my project. Could someone please clarify where image bundles are supposed to go now - stay under public directory or go together with other non- imagebundle files out to war? Thannks --~--~-~--~~~---~--~~ 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 -~--~~~~--~~--~--~---
how to optimize compile speed, reuse of compiled module code possible
Wanted to recheck my understanding of speeding up compile time. As I read in Java, Ubuntu, ... blog at http://blog.bazoud.com/post/2008/07/31/Can-I-speed-up-the-GWT-compiler-Part-II, - setting "user.agent" property value to your development browser's enging - using only default locale for compilation - and ant's uptodate task to compile modules only when their code has changed speeds the things up. My project consists of loosely coupled entrypoint-less gwt-modules being included by main module (with entrypoint defined). I am wondering if it is possible somehow to extend usefullness of ant's uptodate task functionality to enable finer granularity of compile. So if I changed module A code (with entrypoint defined, which includes module B) and start a compiler, than the compiler compiles only the module A's code and reuses module B's code which presumably was already once compiled before? Another question of interest, do my settings of "user.agent" and locales in the top level module overwrite these of the included ones. In end effect, if I want maximul speed-up, do I need to have all included modules each specifying user.agent and default locale? Cheers, denis --~--~-~--~~~---~--~~ 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: gwt 1.7 app - blank page in ie7, safari
I had similar issues as my app code embedded in new environment stopped working. Some general recommendations I could make are: * try using oophm (http://code.google.com/p/google-web-toolkit/wiki/ UsingOOPHM), it is much easier to track down problems in oophm, albeit getting oophm to work may turn out treaky * try switching other non-gwt javascript libraries/functions, which may interfere with gwt functioning to see if it would work otherwhise * try running on another hosted mode browser (easier with oophm, requieres platform change with in-process hosted mode). On Aug 4, 2:12 pm, sam wrote: > I am having similar kinda problem. I am creating a Google Map > Application that has a draggable marker with search option. I am using > GWT and eclipse. My search button works fine in the hosted mode, but > when I compile and run it off the browser, my search button does not > work. My button seems to be firing other commands like pop ups. > > thanks, > sam > > On Aug 3, 3:52 pm, otismo wrote: > > > It wasn't a syntax issue. It was a coding problem. I'm using various > > client-side datastores (i.e. localstorage, userdata, whatwg, etc.). > > The datastores behave differently on store and load and were > > triggering different code paths. I've got it fixed. > > > Thanks for your help! > > > On Aug 3, 11:11 am, Jason Parekh wrote: > > > > Hey Peter, > > > > Could you try changing your tag to include the attribute: > > > type="text/javascript": > > > > > > > > > > > > >