Re: using GLIBMM_PROPERTIES_ENABLED & co

[Murray Cumming <murrayc murrayc com>]

On Mon, 2008-04-28 at 18:10 +0300, Ionutz Borcoman wrote:
> On Monday 28 April 2008 5:28:31 pm Murray Cumming wrote:
> > On Mon, 2008-04-28 at 13:16 +0300, Ionutz Borcoman wrote:
> > > Can I define myself these variables to change the behavior of gtkmm? How
> > > are the users (developers) of gtkmm cope with this if it is determined at
> > > compile time? If some Linux distro decides to use a different set of
> > > compile-time flags,
> >
> > No distro does this. And any non-embedded distro that did it on purpose
> > would be a rather stupid distro that we could feel free to ignore.
> 
> I'm still a bit confused. Am I supposed to add a bunch of #ifdef ... #else .. 
> #endif

Not unless you care about your build running on Maemo or something
similar.

>  or can I assume that no distro is going to change the ./configure 
> flags?

Yes.

>  Are all these flags only for embedded systems? 

Yes.

> The gtkmm tutorial is not helping with this, by the way. If these defines are 
> only for embedded systems like maemo, it would have been nice to have it 
> stated in the examples instead of leaving the impression that everybody 
> (including the desktop developers) have to use these #ifdef ... #endif.

Yes, though I was trying to avoid repeating the same comments over and
over in the examples, making that unusual case seem more significant
than it is.

But I do want the examples to build on all platforms, so I want to keep
the #ifdef in there. Maybe we can remove those before including those
source files in the online book. 

> It would have been much nicer if I could have done something like this:
> 
>   #define ENABLE_XXX
>   #include <gtkmm.h>
> 
> and simply get the XXX feature. Probably not realistic, I admit, but it would 
> have been nice :-/

No, sorry, that's not possible.

-- 
murrayc murrayc com
www.murrayc.com
www.openismus.com