Re: [gwt-contrib] Re: chrome broken

2010-08-20 Thread John LaBanca
Do you see this error every time, or just intermittently? Thanks, John LaBanca jlaba...@google.com On Thu, Aug 19, 2010 at 11:58 PM, Pascal Patry i...@invalidip.com wrote: On Thursday, August 19, 2010 23:50:44 Ray Cromwell wrote: That's interesting, because Longs are supposed to be

Re: [gwt-contrib] Re: chrome broken

2010-08-20 Thread Pascal Patry
On Friday, August 20, 2010 14:24:47 John LaBanca wrote: Do you see this error every time, or just intermittently? Every time with a RequestObjectLong. I suspect any other kind of request using RequestObject would also trigger it. As I previously mentioned, it's a chrome only issue. --

Re: [gwt-contrib] Re: chrome broken

2010-08-20 Thread John LaBanca
We tracked it down to a Chrome dev mode oddity. Chrome dev mode adds a __gwt_ObjectId attribute to all objects passed between Java and JSNI. I'm testing a simple fix now. Thanks, John LaBanca jlaba...@google.com On Fri, Aug 20, 2010 at 3:08 PM, Pascal Patry i...@invalidip.com wrote: On

Re: [gwt-contrib] Re: chrome broken

2010-08-20 Thread Pascal Patry
On Friday, August 20, 2010 15:31:55 John LaBanca wrote: We tracked it down to a Chrome dev mode oddity. Chrome dev mode adds a __gwt_ObjectId attribute to all objects passed between Java and JSNI. I'm testing a simple fix now. Great, thank you --

Re: [gwt-contrib] Re: chrome broken

2010-08-20 Thread Amit Manjhi
Can you try John's fix at http://gwt-code-reviews.appspot.com/760803/show ? Thanks. On Fri, Aug 20, 2010 at 12:33 PM, Pascal Patry i...@invalidip.com wrote: On Friday, August 20, 2010 15:31:55 John LaBanca wrote: We tracked it down to a Chrome dev mode oddity. Chrome dev mode adds a

Re: [gwt-contrib] Re: chrome broken

2010-08-20 Thread Pascal Patry
On Friday, August 20, 2010 15:51:10 Amit Manjhi wrote: Can you try John's fix at http://gwt-code-reviews.appspot.com/760803/show ? Thanks. Yes, it won't be long.. -- http://groups.google.com/group/Google-Web-Toolkit-Contributors

Re: [gwt-contrib] Re: chrome broken

2010-08-20 Thread Patrick Julien
I actually reported this issue already a while back: http://code.google.com/p/google-web-toolkit/issues/detail?id=5110 On Fri, Aug 20, 2010 at 3:31 PM, John LaBanca jlaba...@google.com wrote: We tracked it down to a Chrome dev mode oddity.  Chrome dev mode adds a __gwt_ObjectId attribute to

Re: [gwt-contrib] Re: chrome broken

2010-08-20 Thread Pascal Patry
On Friday, August 20, 2010 16:05:32 Patrick Julien wrote: I actually reported this issue already a while back: http://code.google.com/p/google-web-toolkit/issues/detail?id=5110 On Fri, Aug 20, 2010 at 3:31 PM, John LaBanca jlaba...@google.com wrote: We tracked it down to a Chrome dev mode

[gwt-contrib] Re: chrome broken

2010-08-19 Thread rjrjr
We believe this was fixed at r8572. Can anyone confirm? On Aug 18, 3:45 pm, Ray Cromwell cromwell...@gmail.com wrote: Thanks, there will be a fix for this by COB today. -Ray On Wed, Aug 18, 2010 at 2:38 PM, Pascal Patry i...@invalidip.com wrote: On Wednesday, August 18, 2010 16:41:39 you

Re: [gwt-contrib] Re: chrome broken

2010-08-19 Thread Pascal Patry
On Thursday, August 19, 2010 13:48:53 rjrjr wrote: We believe this was fixed at r8572. Can anyone confirm? sure, won't be long. -- http://groups.google.com/group/Google-Web-Toolkit-Contributors

Re: [gwt-contrib] Re: chrome broken

2010-08-19 Thread Pascal Patry
On Thursday, August 19, 2010 13:48:53 rjrjr wrote: We believe this was fixed at r8572. Can anyone confirm? It fixed the latest issue in Firefox, however, Chrome is still broken. Stack Trace: java.lang.IllegalArgumentException: invoke arguments: JS value of type int, expected

Re: [gwt-contrib] Re: chrome broken

2010-08-19 Thread Pascal Patry
On Thursday, August 19, 2010 14:15:56 Ray Ryan wrote: Your gwt-servlet jar might be stale, try re-genning it. (ant dist-dev) I don't think, since I'm usually building as: ant clean; ant buildonly doc; cd distro-source; ant It did make a difference by fixing Firefox completely. The problem is

Re: [gwt-contrib] Re: chrome broken

2010-08-19 Thread Pascal Patry
On Thursday, August 19, 2010 16:36:36 Ray Cromwell wrote: Could you use the Chrome Inspector to post the JSON payload that came back? Sure, the response that fails on the client side is: {result:3,related:{}} The request here was a RequestObjectLong and I do expect 3 to be the response of this

Re: [gwt-contrib] Re: chrome broken

2010-08-19 Thread Ray Cromwell
That's interesting, because Longs are supposed to be serialized as strings not JSON numbers, I'll have to check the server code. Would you mind posting the Request interface method definition? Does the class referenced by @Service actually return a Long/long? -Ray On Thu, Aug 19, 2010 at 2:00