[erlang-questions] wx_object and OTP design pattern

[Frans Schneider]

What I have done so far is this:
I have an OTP application xxx with the xxx application behaviour, a 
supervisor and some gen_servers.
Based on the code from observer, I created a wx application with a 
hierarchy of wx_objects in line with functional parts of the 'view' I 
want to create. @Craig, I fully agree with you that the whole wx_object 
concept is indeed very nice, a specially since I would like to follow 
with the MVC design pattern.
 From xxx_app.erl, I start the wx application before starting xxx's 
supervisor.
The wx application has a application:stop(xxx) in it's terminate 
callback to stop xxx when the top frame window is closed.

To communicate with the 'view' from the rest of the xxx application 
(model and controller), I guess I should use some kind of pubsub 
protocol with named processes. Something in line with pureMVC 
(http://puremvc.org/) without the OOP stuff of course..

Any suggestions or comments are highly appreciated.

Frans

On 09/09/2014 07:28 PM, Frans Schneider wrote:
> Hi list,
>
> I want to incorporate wx in my application and was wondering how the 
> wx_object behaviour fits into the OTP application structure. Any 
> design patterns for this? All examples or applications using wxErlang 
> I found do not follow the OTP prinicples.
>
> Thanks,
>
> Frans