Re: gi-mm

[Ritesh Khadgaray <khadgaray gmail com>]

Hi

On Sat, 2014-08-16 at 03:15 +0530, Ritesh Khadgaray wrote:
> Hi
>
> On Sun, 2014-08-10 at 14:01 +0200, Juan R. Garcia Blanco wrote:
> > Hi,
> >
> > I think that having a tool that translates gi xml files to C++ code
> > wouldn't allow us to use 'any gobject enabled library' from C++. C++ is
> > a compiled language, so even if you have installed in your system
> > typelibs of the libraries you want to use, you still need to generate C
> > ++ code, generate a C++ library and install that library. The story is,
> > however, much more different for interpreted languages.
> >
> > I do think that a gi-xml to mm-macros translator would be really useful
> > to ease wrapping of new libraries/types. Wrapping GObject libraries is
> > sometimes routine, and having some aid would be great.
> >
> > Best regards,
> > Juan.
> <snip/>
>
> I do see valabind, which seems to have the ability to generate C++
> binding from vapi . Has anyone worked with this before ?

* This lacks STL integration
* Does not uses namespaces
* The header currently uses inline for everything
* Lack of smart pointers
* Would need to implement wrap/unwrap function
* Closer to C, than C++

I believe, with a bit of work this would be much more usable than
current gmmproc method, when extending cxx interface for other
libraries. 


-- 
Cheers
-- ritz
Is it weird in here, or is it just me?
                -- Steven Wright