>
> On 11.01.2011, at 10:21, MacArthur, Ian (SELEX GALILEO, UK) wrote:
> > One for Manolo I guess. Though... There is the bigger question of
> > whether Fl_X should be exposed or not.
> >=20
> > What do folk think?=20
> >=20
> > I don't know - I'd guess not, for "safety" reasons, but there are some
> > advantages, too...
>
> This is bad. The xid should be available to the user.

Isn't the fl_xid(const Fl_Window*) function (documented
in the OS issues section) enough ?


> In the original code, including any header except "FL/x.H" would not
> load (or "try not to load") anything specific to the underlying OS.
> However including "FL/x.H" (and *NOT* "FL/mac.H" or "FL/win32.H",
> these are included automatically when using x.H!) would open up
> access to the system resources.

The new setup requires, under Mac OS for now, to define
FL_LIBRARY for FL/x.H to give full access to what's OS-specific.
Without FL_LIBRARY, there's access only to what's documented
(e.g., fl_gc, fl_xid()).
The benefit of that is to allow for a completely OS-independant
compilation of application programs, many of which include FL/x.H.
The drawback is the requirement to define FL_LIBRARY to get full
access to the OS-specific part of FLTK.
This has been done in reply to a user's request to compile
application programs with non Apple compilers on Mac OS.
The use of non Apple compilers on Mac OS is probably a very marginal
situation. But, the concept of OS-independent compilation of
FLTK applications seems interesting to me.
It's very possible to return to the previous setup, though,
if that's preferred.


> first_window, btw., is a static member of Fl_Window in FLTK2 (or
> rather fltk::Window). I find that a good idea.

Isn't Fl::first_window(void) just as good ?


_______________________________________________
fltk-dev mailing list
[email protected]
http://lists.easysw.com/mailman/listinfo/fltk-dev

Reply via email to