On Mon, 2010-08-02 at 09:06 -0700, Marcel Holtmann wrote: > > > > diff --git a/src/ipconfig.c b/src/ipconfig.c > > > index 6ffcd88..e6e63b2 100644 > > > --- a/src/ipconfig.c > > > +++ b/src/ipconfig.c > > > @@ -849,6 +849,7 @@ static struct connman_ipconfig *create_ipv6config(int > > > index) > > > > > > ipv6config->index = index; > > > ipv6config->type = CONNMAN_IPCONFIG_TYPE_IPV6; > > > + ipv6config->method = CONNMAN_IPCONFIG_METHOD_OFF; > > > > > > ipv6config->address = connman_ipaddress_alloc(AF_INET6); > > > if (ipv6config->address == NULL) { > > > > Because you don't yet support RA, this isn't going to break my normal > > IPv6 connectivity, right? > > it should not break anything. However the more testing we get on this > the better.
I kind of expected this to break with commit 933237c1 but it didn't. Is that expected? My home network appears in default.profile with IPv6.method=off IPv6.netmask_prefixlen=0 IPv4.method=dhcp IPv4.netmask_prefixlen=0 I remain concerned that at some point, this *is* going to break. Really, shouldn't we be defaulting to autoconf as the IPv6 method? -- dwmw2 _______________________________________________ connman mailing list connman@connman.net http://lists.connman.net/listinfo/connman