[Laszlo-dev] amazon-soap demo

[ono keiji]

Yes sure, the bug messages disappeared.
Thank you Henry.

Ono Keiji

Henry Minsky さんは書きました:
> In flash 9/10 (actionscript 3), the compiler requires much more strict
> type declarations, including
> number of arguments to methods, and the data types of those functions.
> 
> In addition, if a subclass overrides a method in a superclass, it's
> argument signature must match
> in the number and types of the arguments (except for optional arguments).
> 
> If you are writing in javascript, you must also explicitly declare a
> method as being an override by
> saying
> 
> override function foo (...)
> 
> (In LZX, we do this automatically for you when you define methods on classes)
> 
> IN the case of amazon-soap, there is a method named "stop", which
> conflicts with the built-in
> LzView stop() method, because it has the wrong arguments signature. If
> you look at the
> amazon/amazon.lzx (non-soap version of the demo) , that method has
> been renamed to "stopdrag",
> so you should make that change to the amazon-soap demo. You ought to
> check against this newer
> version of the amazon app for  other things have been
> modified to make it compile in swf9 as well.
> 
> 
> 
> On Wed, Dec 17, 2008 at 10:02 PM, ono keiji <keiji_ono at net8.co.jp> wrote:
>> Today, i have tested with rev.12167, but amazon-soap.lzx said compile error
>> with using swf9/10 as following,
>>
>> org.openlaszlo.sc.CompilerError: amazon.lzx: 581: Error: Incompatible
>> override, in line: public override function stop () {
>>
>> It dose not said with swf8.
>> What is the 'public override function' ?
>>
>> Ono Keiji
>>
>>
>> ono keiji さんは書きました:
>>> I am seeing amazon-soap.lzx now. As you know, it dose not work at this
>>> time.
>>> The one error said the cause of soap.lzx as following method.
>>>
>>>       <method name="makeProxyStubFunction" args="fn">
>>>       <![CDATA[
>>>           var stubinfo = this.proxyinfo[fn];
>>>           var stubfunc = function (){
>>>               var args = arguments.callee.args;
>>>               Debug.write( "soap.lzx: arg[1]", arguments[1] );
>>>               Debug.write( "soap.lzx: arg[0]" ,arguments[0] );
>>>               return LzSOAPService.invoke(
>>>                   arguments[1],
>>>                   arguments[0],
>>>                   args.superclass.requestheaders,
>>>                   stubinfo, false, null);
>>>           }
>>>
>>>           return stubfunc;
>>>       ]]>
>>>       </method>
>>>
>>> So my question are,
>>>    1.Dose 'LzSOAPService' in lzx file is the instance of
>>> LZSOAPService.java ?
>>>    2.If so, dose LzSOAPService.invoke() is also determined in
>>> LZSOAPService.java ?
>>>      It has different arguments with the java file.
>>>
>>> Thank you.
>>>
>>> Ono Keiji
>>>
>>> ono keiji さんは書きました:
>>>> Fumm, so next should i check the java sources under ./json/soap ?
>>>> I wounder why the second argument did not get executed in the function,
>>>> though the first one got in it.
>>>> I mean, in the *constant* values
>>>>
>>>>    value="${canvas.soaptest.makedoc( parent.name, s.text )}"
>>>>
>>>> 'parent.name' argument had been gotten executed in makedoc(), but
>>>> 's.text' had not been.
>>>> Could you tell me which the file is doing such converting ?
>>>>
>>>> Thank you.
>>>>
>>>> ono keiji
>>>>
>>>> Henry Minsky さんは書きました:
>>>>> Oh yes, that's right. Data which is bound in from from a datapath is
>>>>> now converted according to the
>>>>> attribute type. I don't think that applies to this code test case
>>>>> though.
>>>>>
>>>>>
>>>>>
>>>>> On Fri, Dec 12, 2008 at 11:20 AM, P T Withington <ptw at pobox.com> wrote:
>>>>>> On 2008-12-12, at 10:43EST, Henry Minsky wrote:
>>>>>>
>>>>>>> The schema data type of an attribute  ( "expression". "string",
>>>>>>> "color", etc ) only
>>>>>>> affects how the tag compiler compiles *constant* values in the source
>>>>>>> code, such as
>>>>>> And how data bindings are converted.
>>>>>>
>>
>