Le vendredi 05 décembre 2008 à 11:10 +0100, Olav Vitters a écrit : > > I’m not really convinced by the minidump approach, especially because of > > the confidentiality implications. Furthermore the current code still has > > non-free components and is not distributable. > > Non-free? By Debian guidelines or across distributions? Regardless of the distribution’s guidelines, a binary using this code is not distributable, because of GPL incompatibility. http://bugzilla.gnome.org/show_bug.cgi?id=480128 > You aren't shipping Bug-Buddy then? There is no need to go that far; we are simply shipping a version with google-breakpad support disabled. > > The approach I’ve been pondering for some time is instead to put the > > debugging information on a public server, shared by WebDAV, so that it > > is be possible to mount it with FUSE and let bug-buddy dump the trace > > directly. The problem with this approach is that gdb currently reads the > > I don't understand this bit. Do you mean that the core dump would be put > on a public server? Certainly not. The idea would be to allow bug-buddy to mount the debugging symbols on the user’s machine, to obtain the full backtrace here and to only send this (hopefully usable) trace on the network. Cheers, -- .''`. : :' : We are debian.org. Lower your prices, surrender your code. `. `' We will add your hardware and software distinctiveness to `- our own. Resistance is futile.
Description: Ceci est une partie de message =?ISO-8859-1?Q?num=E9riquement?= =?ISO-8859-1?Q?_sign=E9e?=