Re: C99 complex types in Glib

[CaStarCo <castarco gmail com>]

Hello :)

2010/12/11 Emmanuele Bassi <ebassi gmail com>

hi;

glib development is discussed on the gtk-devel mailing list, but I can
directly answer to you right now.

On Sat, 2010-12-11 at 00:19 +0100, CaStarCo wrote:

> I write here because i have many doubts and because i want to
> contribute with my code to the Glib library ^^ .

cool, contributions are much welcome...

> The first doubt is : The Glib project is using the C99 spec or the C89
> spec?

for portability reasons, GLib follows the C89 spec. the C99 spec is not
fully implemented by all compilers on all platforms we're tracking, so
it cannot be used.

> If it's the second case... there is any way to contribute to Glib
> adding support to the complex floating point types that C99 introduced
> in the year 2000 ?

honestly, I doubt this is a compelling enough use case to introduce new
types; applications that require complex floating point types will also
require other ad hoc facilities that GLib does not provide.

I don't think so :p , one usage example: extending the Vala language, in any case, the complex.h header provides simple functions to work with complex numbers.

Another usage example: simplify the code of, for example, gcalctool.

In any case, if the reasons of not extending Glib are related with compatibility, that's a stronger reason :) .

 

nevertheless, if new types have been introduced and if they can be
checked at configure-time, then they may be supported by wrapping them,
in to provide a graceful degradation path.

I have a doubt about that. To introduce a type following the way you described... I must create another "mini library" inside the Glib?

 

code contributions for GLib go through Bugzilla:

 http://bugzilla.gnome.org/enter_bug.cgi?product=glib

ciao,
 Emmanuele.

Thanks for your atention! 

-- 
- Per la llibertat del coneixement -
- Per la llibertat de la ment...       -