Ryan, I agree that the security token in the gadget holder should be done on the navigate. It appears this was the intention, but I think there are a few things about security tokens that haven¹t been completed in the common container (I wish these holes were documented). My first attempt WAS to patch the code, however we¹ve been trying to use the maven artifacts and not override any of the base stuff but instead plug in where we think appropriate. I guess as far as the community is concerned I should be patching the code, however I think the google guys have a better idea of what the intentions were here. As far as intercepting the RPC call. I hate this. I really wish I could figure out how to bubble down the gadget security token to the jsonrpcrequest. Thanks for the confirmations.
doug On 7/20/11 7:42 PM, "Ryan J Baxter" <[email protected]> wrote: > Doug I am glad you found a solution. I feel like the first part where you > are setting the security token in the gadget holder, should be done by the > common container code automatically when you call > container.navigateGadget. Maybe you can look into doing this can creating > a patch? > > The second part where you are intercepting the RCP call and getting the > security token out of the gadget holder, still seems a little odd....but I > can't think of a better solution right now. > > -Ryan > > Email: [email protected] > Phone: 978-899-3041 > developerWorks Profile
