Hi,

>>
>>  That would make sense, but this should then be included in the
>> documentation, because there is nothing about this! This requires the server
>> to adhere to some conventions that are not at all described in the
>> documentation. What is id? What is returnValue?
>>
>
> If we documented it, more people would use Script Transport with all of its
> nasty security implications, so leaving it undocumented lets us tell each of
> you how nasty it really is to try to discourage you from using it. :-)
>

Yeah well... I prefer to let unknowing people shoot themselves in the foot
than prevent knowledgeable people from using the framework as they need !!!
:) Seriously, crossDomain is false by default: people won't use it if they
don't need it.


> You're right. It could use some documentation someplace. I'm not sure
> where. Probably in the Server Writer's Guide, but that's RPC-specific.
> Hmmm... Fabian likes doing wiki edits. Maybe he'll find some nice
> appropriate place for it.
>

It *definitely* needs to be documented somewhere. As it is the feature is
just unusable (unless you write the ML like I did). Why dont you write that
in http://qooxdoo.org/documentation/0.8/remote_io ? This seems the place for
it.


> Anyway, yes, the return value from your Script Transport server must be a
> string containing a call to a javascript (qooxdoo) function:
> qx.io.remote.transport.Script._requestFinished(). That function expects two
> parameters. The first is the ID that was provided to the server by the
> qooxdoo client in the GET request, parameter name "_ScriptTransport_id".
> Passing this id back to the qooxdoo client allows qooxdoo to associate the
> incoming response with the original request. The second parameter to the
> function is the JSON object that you want returned to the client.
>

I see, makes sense. Actually as I said I think it's the only (sane) way to
proceed as IE does not fire a load event when the script node is added (and
loaded) into the DOM. Even when I used the script transport "manually" I had
to do this kind of "hack" to get the control back to the client side code
(because of IE; with FF you can rely on standard events and do not have to
write explicitely a callback function).

A last note. With the script transport you can only use GET I think, not
POST. Can anyone confirm ?

Thanks
Jean-Noel


>
>
> Derrell
>
>
>
> ------------------------------------------------------------------------------
> Crystal Reports - New Free Runtime and 30 Day Trial
> Check out the new simplified licensign option that enables unlimited
> royalty-free distribution of the report engine for externally facing
> server and web deployment.
> http://p.sf.net/sfu/businessobjects
> _______________________________________________
> qooxdoo-devel mailing list
> [email protected]
> https://lists.sourceforge.net/lists/listinfo/qooxdoo-devel
>
>
------------------------------------------------------------------------------
Crystal Reports - New Free Runtime and 30 Day Trial
Check out the new simplified licensign option that enables unlimited
royalty-free distribution of the report engine for externally facing 
server and web deployment.
http://p.sf.net/sfu/businessobjects
_______________________________________________
qooxdoo-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/qooxdoo-devel

Reply via email to