Matthais,
Actually, this is what Adobe does in DllMain (you'll laugh):
extern "C" BOOL APIENTRY DllMain(HANDLE hModule, DWORD ul_reason_for_call,
LPVOID);
BOOL APIENTRY DllMain(HANDLE hModule, DWORD ul_reason_for_call, LPVOID)
{
dllInstance = static_cast<HINSTANCE>(hModule);
return true;
}//DllMain
-Chris
>
> On 05.09.2012, at 11:28, "MacArthur, Ian (SELEX GALILEO, UK)" =
> <[email protected]> wrote:
>
> >=20
> >>> I wonder if we need to explicitly unregister the fltk window classes
> >> from the Windows context and then re-assert them, or maybe register =
> them
> >> with a different name each time, or something...?
> >>=20
> >>=20
> >> I am not sure if the entire endeavor will ever work. It's a long time =
> ago
> >> that I used the PS plugin API, and it was quite limiting. I believe =
> that
> >> there is a good reason for that: it is close to impossible to have
> >> independent UI systems running within one application.
> >>=20
> >> Even though FLTK is of course based on the same core libraries as
> >> Photoshop, there can be only on master. This master captures all the
> >> events sent from the OS and decides if they are sent on or not. FLTK
> >> requires a lot of events to function correctly, but if Photoshop =
> snaps
> >> those events away before they reach FLTK, FLTK will never function
> >> correctly.
> >=20
> >=20
> > Agree.
> >=20
> > But the kicker here is that it (at least appears) to work perfectly =
> the *first* time that the fltk window is shown. Then subsequent attempts =
> to show the fltk context fail...
> >=20
> > So I do wonder if there's something we can do to make every time just =
> like the first time, and hence at least give the appearance of working =
> Just Right (even if under the covers it is all a bit suspect!)
>
> Has it been tested that the dll is not unloaded? Maybe PS *does* =
> (partially, thanks to MSWindows) during calls to the plugin? A simple =
> debug breakpoint in the DLL initialization and deinitialization should =
> clarify that.
>
> On WIN32, this should help:
>
> BOOL APIENTRY DllMain(HANDLE hModule,=20
> DWORD ul_reason_for_call,=20
> LPVOID lpReserved)
> {
> switch( ul_reason_for_call ) {
> case DLL_PROCESS_ATTACH:
> ...
> case DLL_THREAD_ATTACH:
> ...
> case DLL_THREAD_DETACH:
> ...
> case DLL_PROCESS_DETACH:
> ...
> }
> return TRUE;
> }
>
> Maybe the original code needs to be replaced in the DLL to avoid =
> destruction of some statics?=
>
_______________________________________________
fltk mailing list
[email protected]
http://lists.easysw.com/mailman/listinfo/fltk
