[Mono-list] Fwd: embedding mono runtime

Miguel de Icaza miguel@ximian.com
09 Aug 2002 08:43:52 -0400


Hello!

> I'm not sure, if this mail will make it to the list, I think, I only have my other address subscribed, which I can't use from here.
> 
> What I'm not completely sure about is, how I really should proceede.
> 
> To not block the application I have to invoke CIL methods, which will have to return quite seen, to not block the GUI.
> 
> Do managed objects still exist, after the runtime invocation exists (because the end of a method has been reached)? 

Yes, they will keep existing.  They are just "owned" by Mono initially.

> If I call the jit again, with the same domain, can I reuse the objects from my previous call ?

Yes, you can ;-)

> Do I have to add  calls to the mono event loop into my application's main event loop ?

Only if you need to invoke methods inside the Mono-world, otherwise you
should be fine.


> What I don't get at the moment is, what I have to do, to neither block the managed code nor the application itself

Well, I would say that you can do a number of things:

	* Simple commands can probably just be called synchronously,
	  as they will just execute quickly and return.

	* More complex commands that have a GUI, you would call into
	  Mono, and then you Qt# dialog, could launch the Qt# main
	  loop to get events processed.

	* Even more complex interactions would require you to have a
  	  thread, but this is just an extreme case.

Miguel