Bug#443824: libtrackerclient0: should be linked with -Bsymbolic

2007-09-28 Thread Josselin Mouette
Le vendredi 28 septembre 2007 à 14:55 +0200, Michael Biebl a écrit : This is why it is a good idea for *all* Pango and GTK+ plugins to be linked with -Bsymbolic. This option guarantees that a function in libtrackerclient that calls foo will obtain the function foo defined in

Bug#443824: libtrackerclient0: should be linked with -Bsymbolic

2007-09-28 Thread Michael Biebl
Josselin Mouette schrieb: Le vendredi 28 septembre 2007 à 14:55 +0200, Michael Biebl a écrit : This is why it is a good idea for *all* Pango and GTK+ plugins to be linked with -Bsymbolic. This option guarantees that a function in libtrackerclient that calls foo will obtain the function foo

Bug#443824: libtrackerclient0: should be linked with -Bsymbolic

2007-09-28 Thread Michael Biebl
Josselin Mouette schrieb: Le vendredi 28 septembre 2007 à 11:43 +0200, Michael Biebl a écrit : To avoid weird crashes caused by namespace conflicts, libtrackerclient should be built with LDFLAGS += -Wl,-Bsymbolic. Hi Joss, the ld manpage was not particularly helpful regarding this topic.

Bug#443824: libtrackerclient0: should be linked with -Bsymbolic

2007-09-28 Thread Michael Biebl
Josselin Mouette schrieb: Package: libtrackerclient0 Version: 0.6.2-2 Severity: normal Hi, as libtrackerclient.so.0 is now dlopen()ed by GTK+ applications, the symbols it contains can now share the namespace of any of these applications, and that makes a lot. To avoid weird crashes

Bug#443824: libtrackerclient0: should be linked with -Bsymbolic

2007-09-24 Thread Josselin Mouette
Package: libtrackerclient0 Version: 0.6.2-2 Severity: normal Hi, as libtrackerclient.so.0 is now dlopen()ed by GTK+ applications, the symbols it contains can now share the namespace of any of these applications, and that makes a lot. To avoid weird crashes caused by namespace conflicts,