Hello Janne!   
      
   25 Oct 2018 10:06, Janne Johansson wrote to Benny Pedersen:   
      
    JJ> On 2018-10-25 00:40, Benny Pedersen : Michiel van der Vlist wrote:   
    >> -áMvdV> and up).-á If the destination address is a 6to4 tunnel address   
    >> -áMvdV> (2002::/16) many OSs default to IPv4 if an IPv4 address is    
    >> -áMvdV> present.   
    >>   
    >> time to update to a later glibc where ipv6 is prefered over ipv4    
    >> failback   
      
    JJ> I think its mostly a setting in your resolv.conf or similar file to   
    JJ> change default priority of v4 or v6 and not so much a age-of-glibc    
    JJ> issue.   
      
      ----- gai.conf begins -----   
   # Configuration for getaddrinfo(3).   
   #   
   # So far only configuration for the destination address sorting is needed.   
   # RFC 3484 governs the sorting.  But the RFC also says that system   
   # administrators should be able to overwrite the defaults.  This can be   
   # achieved here.   
   #   
   # All lines have an initial identifier specifying the option followed by   
   # up to two values.  Information specified in this file replaces the   
   # default information.  Complete absence of data of one kind causes the   
   # appropriate default information to be used.  The supported commands include:   
   #   
   # reload     
   #    If set to yes, each getaddrinfo(3) call will check whether this file   
   #    changed and if necessary reload.  This option should not really be   
   #    used.  There are possible runtime problems.  The default is no.   
   #   
   # label         
   #    Add another rule to the RFC 3484 label table.  See section 2.1 in   
   #    RFC 3484.  The default is:   
   #   
   #label ::1/128       0   
   #label ::/0          1   
   #label 2002::/16     2   
   #label ::/96         3   
   #label ::ffff:0:0/96 4   
   #label fec0::/10     5   
   #label fc00::/7      6   
   #label 2001:0::/32   7   
   #   
   #    This default differs from the tables given in RFC 3484 by handling   
   #    (now obsolete) site-local IPv6 addresses and Unique Local Addresses.   
   #    The reason for this difference is that these addresses are never   
   #    NATed while IPv4 site-local addresses most probably are.  Given   
   #    the precedence of IPv6 over IPv4 (see below) on machines having only   
   #    site-local IPv4 and IPv6 addresses a lookup for a global address would   
   #    see the IPv6 be preferred.  The result is a long delay because the   
   #    site-local IPv6 addresses cannot be used while the IPv4 address is   
   #    (at least for the foreseeable future) NATed.  We also treat Teredo   
   #    tunnels special.   
   #   
   # precedence        
   #    Add another rule to the RFC 3484 precedence table.  See section 2.1   
   #    and 10.3 in RFC 3484.  The default is:   
   #   
   #precedence  ::1/128       50   
   #precedence  ::/0          40   
   #precedence  2002::/16     30   
   #precedence ::/96          20   
   #precedence ::ffff:0:0/96  10   
   #   
   #    For sites which prefer IPv4 connections change the last line to   
   #   
   #precedence ::ffff:0:0/96  100   
      
   #   
   # scopev4       
   #    Add another rule to the RFC 6724 scope table for IPv4 addresses.   
   #    By default the scope IDs described in section 3.2 in RFC 6724 are   
   #    used.  Changing these defaults should hardly ever be necessary.   
   #    The defaults are equivalent to:   
   #   
   #scopev4 ::ffff:169.254.0.0/112  2   
   #scopev4 ::ffff:127.0.0.0/104    2   
   #scopev4 ::ffff:0.0.0.0/96       14   
      ----- gai.conf ends -----   
      
   speeks for it self, if unchanged config, ipv6 is prefered   
      
      
    Regards Benny   
      
   ... there can only be one way of life, and it works :)   
      
   --- Msged/LNX 6.1.2 (Linux/4.14.65-gentoo (x86_64))   
    * Origin: I will always keep a PC running CPM 3.0 (2:230/0)