manolo gouy wrote:

> The why is briefly explained in my other post "New hierarchy of
> Fl_Device subclasses".
> 
> The new hierarchy is :
> 
> Fl_Device              (all present and future devices)
>   Fl_Graphics_Device   (Quartz, GDI, or Xlib depending on the platform)
>     Fl_Display_Device  (above graphics system applied to display)
>     Fl_Printer         (above graphics system applied to printing)
>    (Fl_Clipboard_Device  -- would come here if it's adopted)
>   Fl_PS_Device         (PostScript backend)
>     Fl_PSfile_Device   (PostScript used in printing)
> 
> 
> Fl_Abstract_Printer    (a page-structured device)
>   Fl_Printer           (a page-structured printing device)
>   Fl_PSfile_Device     (a page-structured output sent to a local file)

What I would do is:

1. Put all the "printer" commands on the basic Fl_Device. For 
non-printers calling these can be no-ops or cause an assert, it depends 
on the function.

2. Instead of a subclass called "Fl_Printer", there would instead be an 
API called "give me the printer device". This will return a subclass of 
Fl_Device, programs should not have to care which one. Therefore if it 
is a subclass of the PS device on some platforms and the GDI device on 
others it does not matter.
_______________________________________________
fltk-dev mailing list
[email protected]
http://lists.easysw.com/mailman/listinfo/fltk-dev

Reply via email to