[gst-devel] Re: [gst-cvs] thomasvs gst-plugins: gst-plugins/ext/mad/

[Bastien Nocera]

On Thu, 2002-12-12 at 17:08, David I. Lehn wrote:
> * Thomas Vander Stichele <thomasvs at users.sourceforge.net> [20021212 08:59]:
> > Index: gstmad.c
> ....
> >    mad = GST_MAD (gst_pad_get_parent (pad));
> > +  g_assert (GST_IS_MAD (mad));

Overkill, mad will be either a GstMad, or NULL.

> 
> I'm not sure what our current policy is but using g_return_if_fail()
> instead of g_assert() seems to be more the current fad these days.
> Maybe it doesn't matter since it shouldn't trigger either way. ;)

g_assert will make the whole application crash/exit. You just want a
warning, not the whole application to fall over.

In this case, I'm missing a bit of context. Is gst_pad_get_parent()
always supposed to work ? Or it might sometimes, not work, under normal
circumstances ?

-- 
/Bastien Nocera
http://hadess.net

#2  0x4205a2cc in printf ("Oh my %s\n", preferred_deity) from
/lib/i686/libc.so.6 printf ("Oh my %s\n", preferred_deity);
Segmentation fault