Comment # 10 on bug 931165 from Dr. Werner Fink

(In reply to Jan Engelhardt from comment #9)

> And that will be when? I suppose at SLE13GA time - the perfect time to sell 
> net.ifnames=0 in the release notes. Or if you really really want to stay, 
> there's always Base:System:Legacy/systemd. The reason why B:S:L exists is 
> precisely because you want to keep some prior version or extra patches.

The customer wish is what matters.

> So, on your Factory test system, which lines of enable_name_policy() get 
> executed?

I can swtich for and back (after removing
/etc/udev/rules.d/70-persistent-net.rules) and on openSUSE the part with

 -DNET_IFNAMES=0

is valid (the second part with ``predictable'' rules):

 static bool enable_name_policy(void) {
 [...]
 #if defined(NET_IFNAMES) && (NET_IFNAMES == 1)
 # warning Using persistent rules as a default
        if (r <= 0)
                return false;

        FOREACH_WORD_QUOTED(w, l, line, state)
                if (strneq(w, "net.ifnames=1", l))
                       return true;

        return false;
 #else
 # warning Using predictable rules as a default
        if (r <= 0)
                return true;

        FOREACH_WORD_QUOTED(w, l, line, state)
                if (strneq(w, "net.ifnames=0", l))
                        return false;

        return true;
 #endif
 }