Steve wrote:
>I think this is old code that predates the GTK-specific Director
>interface.
I think it isn't possible that this code predates the GTK-specific
Director interface, because I see no reason why it would be in
SciTEGTK.cxx then. However, if you can wait until tonight, I can
then send more information.

>(I've no idea if people still use this FILERX, but it can all be
>handled properly with the 'new' machinery.) There is then the pipe
>used to handle the single-instance machinery.  I agree that both
>of these exit points should execute exactly the same code!
I think FILERX is Windows-only. As for both exit points executing
the same code, that's why I proposed changes to have only one exit
point.

>But why not quit the main loop properly? It feels bad to exit a C++
>app like this, without an opportunity for proper call of destructors!
Well, probably, I don't know. I have no formal C++ training, actually.
These calls are in the code for the single instance, and I think
no class has been instantiated yet, so destructors should not be a
problem. I guess you could return a boolean value and check this
in main(). Or, wait... I may be wrong, is it called from Run()? Okay,
but what is the right way to exit the loop then?

Nicolas

_______________________________________________
Scite-interest mailing list
[email protected]
http://mailman.lyra.org/mailman/listinfo/scite-interest

Reply via email to