Re: Validity of creating a child window in a separate process to the parent (COM/win32)

Tech-Archive recommends: Fix windows errors by optimizing your registry

From: Louis Solomon [SteelBytes] (louis_at_steelbytes.spam-is-bad.com)
Date: 05/11/04 

Date: Tue, 11 May 2004 11:55:15 +1000

just don't try and run your own message loop.
leave that to the main app. 

-- 
Louis Solomon
www.steelbytes.com
"Missing Person" <missing_person@btinternet.com> wrote in message 
news:c7ooa1$ipo$1@sparta.btinternet.com...
>I have a COM based local-server (.exe) with say a method:
>
> HRESULT CreateWidget([in] long parent);
>
> Where the "parent" parameter is a window's HWND, cast to a long (I'm using 
> the default ole-marshaller and passing it as long keeps me from creating 
> custom proxy/stub .dll's).
>
> The local-server's response to the call is to create a child window from 
> the "parent" HWND. I.e. the child window is in a separate process to the 
> parent window.
>
> Essentially I would like the local-server's "rendering" to be performed 
> via a window, rather than have to marshal device independent bitmap 
> formats etc. and have the client application perform the render.
>
> The technique works, and I can rationalize (to myself) the reasons why 
> windows accepts this. But I'm worried that this technique is contrary to 
> some "standard" practice, and that at some point in the future a service 
> pack etc. will close this behavior off (for say security reasons). I 
> cannot find any on-line evidence from Microsoft that this is acceptable 
> practice or not.
>
> So I would like know; is creating a child window in a separate process to 
> its parent:
>
> - A supported and safe, "practice", (i.e. it works because the o/s 
> supports this)
>
> - Guaranteed to be a supported "practice" under at least the current MS 
> O/S's, e.g. 2000/XP (i.e. the behavior will continue to be supported)
>
> Regards.

Relevant Pages