Glynn Foster wrote:
> Hey,
>
> Brian Nitz wrote:
>
>> A few things happened to make crashes more common in recent GNOME builds:
>>
>> 1) The GNOME community enabled coreing on ASSERTs in default builds, so
>> some subtle bugs became less subtle.
>>
>
> I'm confused. Has something changed recently? or is this still the code in
> gnome-session/main.c? In which case it should only currently be triggered for
> all x.y.z where y is an unstable number.
>
Glynn, thanks for the correction. I may have been looking at a
vermillion build. IThat part of the gnome-session main.c code doesn't
seem to have changed between 2.20 and 2.22:
if (g_getenv ("GSM_VERBOSE_DEBUG"))
gsm_set_verbose (TRUE);
/* Help eradicate the critical warnings in unstable releases of GNOME */
versions = g_strsplit (VERSION, ".", 3);
if (versions && versions [0] && versions [1])
{
int major;
major = atoi (versions [1]);
if ((major % 2) != 0)
{
g_setenv ("G_DEBUG", "fatal_criticals", FALSE);
g_log_set_always_fatal (G_LOG_LEVEL_CRITICAL);
}
}
g_strfreev (versions);
So if the middle number of gnome-session --version is even, fail on
asserts shouldn't be set.
Do you happen to know if any individual desktop components look at their
own version number and attempt to override this setting? For example in
SNV_82:
gtkam --version gives 0.1.14
fc-cache --version gives fontconfig version 2.3.2.
pidgin --version gives Pidgin 2.1.1
...
_______________________________________________
opensolaris-discuss mailing list
opensolaris-discuss@opensolaris.org
