Re: [Vala] bugs
- From: Jürg Billeter <j bitron ch>
- To: pancake <pancake youterm com>
- Cc: vala paldo org
- Subject: Re: [Vala] bugs
- Date: Sat, 08 Sep 2007 10:50:34 +0200
On Sat, 2007-09-08 at 07:13 +0200, pancake wrote:
I have seen a glitch on the generated C code and a buggy example on the web.
This is the one provided by the signal example (which doesn't build)
http://live.gnome.org/Vala/Signals
static gint maman_bar_main (int args_length1, char** args)
{
MamanBar* bar;
gint _tmp0;
bar = g_new0 (MamanBar, 1);
maman_bar_run (bar);
return (_tmp0 = 0, (bar == NULL ? NULL : (bar = (maman_bar_free (bar), NULL))), _tmp0);
(bar == NULL ? NULL : (bar = (maman_bar_free (bar), NULL))); /* <------ CODE AFTER RETURN??? ------
*/
}
That's obviously a bug in valac, however it's safe to ignore for now;
it's more a cosmetic problem than anything else. We'll fix it as soon as
we have proper flow analysis in place.
NOTE: I have created an account on the wiki and fixed the signal example.
Thanks for fixing the example. Vala SVN now reports an error when trying
to define a signal in a type not deriving from GLib.Object.
I have tested the use of 'class Maman.Foo' vs 'namespace Maman; class Foo' and looks like
the C generated code is exactly the same. So I understand that both codes are equivalent.
Maybe that's new to me, because I'm a c# ignorant O:)
The implicit namespaces are not supported in C#, that's a Vala feature.
Jürg
[
Date Prev][
Date Next] [
Thread Prev][
Thread Next]
[
Thread Index]
[
Date Index]
[
Author Index]