• Feature request: soft IPv6 force

    From Michiel van der Vlist@2:280/5555 to Tony Langdon on Wed Feb 1 01:39:46 2017

    *** Answering a msg posted in area BINKD (Binkd mailer).

    Hello Tony,

    On Wednesday February 01 2017 09:34, you wrote to me:

    -64 Binkd ignores the OS preference and tries to first make an
    IPv6 connect. If that fails, it tries an IPv4 connect.
    [..]
    That could be useful in some scenarios. In my case, my IPv6
    connectivity (native) is better than IPv4 (tunneled), so it's better
    for me to try IPv6 first in most cases, then fall back to IPv4 if necessary.

    Native IPv6 and IPv4 via a tunnel is the opposite of what many of us have. Quite exceptional. For now. It may not be so exceptional in the future.

    But is IPv6 not the default on your system? Is is different from my situation (Windows, native dual stack) where IPv6 is the default connectiom method, except when the destination has a 6to4 IPv6 address? (2002::/16)


    heers, Michiel

    --- GoldED+/W32-MSVC 1.1.5-b20130111
    * Origin: he.net certified sage (2:280/5555)
  • From Janne Johansson@2:221/6 to Michiel van der Vlist on Wed Feb 1 16:44:06 2017
    On 2017-02-01 01:39, Michiel van der Vlist : Tony Langdon wrote:

    *** Answering a msg posted in area BINKD (Binkd mailer).

    Hello Tony,

    On Wednesday February 01 2017 09:34, you wrote to me:

    -64 Binkd ignores the OS preference and tries to first make an
    IPv6 connect. If that fails, it tries an IPv4 connect.
    [..]
    That could be useful in some scenarios. In my case, my IPv6 connectivity (native) is better than IPv4 (tunneled), so it's better
    for me to try IPv6 first in most cases, then fall back to IPv4 if necessary.

    Native IPv6 and IPv4 via a tunnel is the opposite of what many of us
    have. Quite exceptional. For now. It may not be so exceptional in the
    future.

    But is IPv6 not the default on your system? Is is different from my
    situation (Windows, native dual stack) where IPv6 is the default
    connectiom method, except when the destination has a 6to4 IPv6 address? (2002::/16)

    I wonder if "default connection method" might be "default resolving
    method" on some OSes, if given a hostname. There is a difference in
    which protocol OSes will try first if given both v4 and v6, so having a
    button to make binkd make a specific choice might be a good idea.

    ---
    * Origin: *** nntp://fidonews.mine.nu *** Finland *** (2:221/6.0)
  • From Markus Reschke@2:240/1661 to Janne Johansson on Wed Feb 1 16:27:04 2017
    Hello Janne!

    Feb 01 16:44 2017, Janne Johansson wrote to Michiel van der Vlist:

    I wonder if "default connection method" might be "default resolving method" on some OSes, if given a hostname. There is a difference in
    which protocol OSes will try first if given both v4 and v6, so having
    a button to make binkd make a specific choice might be a good idea.

    On linux you can configure how IPv4 and IPv6 are prioritized via /etc/gai.conf. To prefer 6to4 tunnels over IPv4 comment out the line starting with "label 2002::/16" and restart binkd. Yes, it's that simple ;)

    ciao,
    Markus

    ---
    * Origin: *** theca tabellaria *** (2:240/1661)
  • From Tommi Koivula@2:221/360.8110 to Markus Reschke on Wed Feb 1 18:47:48 2017

    01 Feb 17 16:27, you wrote to Janne Johansson:

    Feb 01 16:44 2017, Janne Johansson wrote to Michiel van der Vlist:

    I wonder if "default connection method" might be "default
    resolving method" on some OSes, if given a hostname. There is a
    difference in which protocol OSes will try first if given both v4
    and v6, so having a button to make binkd make a specific choice
    might be a good idea.

    On linux you can configure how IPv4 and IPv6 are prioritized via /etc/gai.conf. To prefer 6to4 tunnels over IPv4 comment out the line starting with "label 2002::/16" and restart binkd. Yes, it's that
    simple ;)

    In my ubuntu vps box all the lines in /etc/gai.conf are commented out. Like this:

    #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

    Anyway, this looks like the netsh way to do the same in windows.

    'Tommi

    ---
    * Origin: news://rbb.bbs.fi & news://fidonews.mine.nu (2:221/360.8110)
  • From Wilfred van Velzen@2:280/464 to Tommi Koivula on Wed Feb 1 17:56:46 2017
    Hi,

    On 2017-02-01 18:47:48, Tommi Koivula wrote to Markus Reschke:
    about: "Feature request: soft IPv6 force":

    In my ubuntu vps box all the lines in /etc/gai.conf are commented out. Like this:

    #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

    The same is true on openSUSE. But if I read the comments correctly, that's because those are the kernel defaults...

    Bye, Wilfred.


    --- FMail-W32 1.73.15.74-B20170201
    * Origin: Native IPv6 connectable node (2:280/464)
  • From Tony Langdon@3:633/410 to Michiel van der Vlist on Thu Feb 2 08:17:00 2017
    Michiel van der Vlist wrote to Tony Langdon <=-

    Native IPv6 and IPv4 via a tunnel is the opposite of what many of us have. Quite exceptional. For now. It may not be so exceptional in the future.

    I do have native IPv4, but it's only a single IP, which is heavily used. The BBSs get their own IPs via the tunnel (which routes a /29 of IPv4). Because I have a /56, I have plenty of IPv6 addresses to hand out. :)

    But is IPv6 not the default on your system? Is is different from my situation (Windows, native dual stack) where IPv6 is the default connectiom method, except when the destination has a 6to4 IPv6 address? (2002::/16)

    IPv6 is the default, but I would like to force the order for those corner cases, where the default is not what I want.


    ... And if one bad cluster should accidentally fail...
    --- MultiMail/Win32 v0.49
    * Origin: Freeway BBS - freeway.apana.org.au (3:633/410)
  • From Tommi Koivula@2:221/360 to Wilfred van Velzen on Fri Feb 3 08:20:40 2017

    In my ubuntu vps box all the lines in /etc/gai.conf
    are commented out. Like this:

    #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

    The same is true on openSUSE. But if I read the comments correctly, that's because those are the kernel defaults...

    So perhaps this might do what Markus suggested:

    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

    'Tommi

    --- Mozilla/5.0 (Windows NT 6.1; Win64; x64; rv:51.0) Gecko/20100101 SeaMonkey/
    * Origin: *** nntp://rbb.bbs.fi *** Lake Ylo *** Finland *** (2:221/360)
  • From Tommi Koivula@2:221/6 to Wilfred van Velzen on Fri Feb 3 18:30:40 2017
    Hi Wilfred.

    03 Feb 17 08:20:40, I wrote to you:

    In my ubuntu vps box all the lines in /etc/gai.conf
    are commented out. Like this:

    #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

    The same is true on openSUSE. But if I read the comments correctly,
    that's because those are the kernel defaults...

    So perhaps this might do what Markus suggested:

    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

    Confirmed:

    + 18:29 [4917] call to 2:5023/24@fidonet
    18:29 [4917] trying f24.n5023.z2.binkp.net [2002:3e94:91ed::2]...
    18:29 [4917] connected
    + 18:29 [4917] outgoing session with f24.n5023.z2.binkp.net:24554 [2002:3e94:91ed::2]
    - 18:29 [4917] OPT CRAM-MD5-a430691dcf91c9211a3f0f639d0acaee
    + 18:29 [4917] Remote requests MD mode
    - 18:29 [4917] SYS Hobbit Station
    - 18:29 [4917] ZYZ Dmitri Kamenski
    - 18:29 [4917] LOC Kaluga, Russia
    - 18:29 [4917] NDL 115200,TCP,BINKP
    - 18:29 [4917] TIME Fri, 3 Feb 2017 19:29:41 +0300
    - 18:29 [4917] VER binkd/1.1a-95/Win32 binkp/1.1
    + 18:29 [4917] addr: 2:5023/24@fidonet
    - 18:29 [4917] OPT EXTCMD GZ BZ2
    + 18:29 [4917] Remote supports EXTCMD mode
    + 18:29 [4917] Remote supports GZ mode
    + 18:29 [4917] Remote supports BZ2 mode
    + 18:29 [4917] done (to 2:5023/24@fidonet, OK, S/R: 0/0 (0/0 bytes))
    18:29 [4917] session closed, quitting...

    'Tommi

    ---
    * Origin: 2001:470:1f15:cb0:f1d0:2:221:6 (2:221/6)
  • From Michiel van der Vlist@2:280/5555 to Tommi Koivula on Sat Feb 4 16:54:53 2017
    Hello Tommi,

    On Friday February 03 2017 18:30, you wrote to Wilfred van Velzen:

    #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

    Confirmed:

    + 18:29 [4917] call to 2:5023/24@fidonet
    18:29 [4917] trying f24.n5023.z2.binkp.net [2002:3e94:91ed::2]...
    18:29 [4917] connected

    OK, so that way one can make an IPv6 connect to a 6to4 tunneled node. But that does not nullify my feature request for -64 and -46 options. What you did is changing the OS preference. So that will affect /all/ connections. Not just binkp. What I proposed was an override on the OS preference for a specific connection. I.e a binkp connect to a specific node.



    Cheers, Michiel

    --- GoldED+/W32-MSVC 1.1.5-b20130111
    * Origin: he.net certified sage (2:280/5555)
  • From Tommi Koivula@2:221/1.1 to Michiel van der Vlist on Sat Feb 4 19:27:22 2017
    Hi Michiel.

    04 Feb 17 16:54:52, you wrote to me:

    #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

    Confirmed:

    + 18:29 [4917] call to 2:5023/24@fidonet
    18:29 [4917] trying f24.n5023.z2.binkp.net [2002:3e94:91ed::2]...
    18:29 [4917] connected

    OK, so that way one can make an IPv6 connect to a 6to4 tunneled node. But that does not nullify my feature request for -64 and -46 options. What you did is changing the OS preference. So that will affect /all/ connections. Not just binkp. What I proposed was an override on the OS preference for a specific connection. I.e a binkp connect to a specific node.

    Of course. But this is ipv6 echo, not binkd echo. :)

    'Tommi

    ... he.net certified sage
    ---
    * Origin: IPv6 Point at [2001:470:1f15:cb0::4] (2:221/1.1)
  • From Michiel van der Vlist@2:280/5555 to Tommi Koivula on Sun Feb 5 00:44:35 2017
    Hello Tommi,

    On Saturday February 04 2017 19:27, you wrote to me:

    OK, so that way one can make an IPv6 connect to a 6to4 tunneled
    node. But that does not nullify my feature request for -64 and
    -46 options. What you did is changing the OS preference. So that
    will affect /all/ connections. Not just binkp. What I proposed
    was an override on the OS preference for a specific connection.
    I.e a binkp connect to a specific node.

    Of course. But this is ipv6 echo, not binkd echo. :)

    Indeed. Is there an equivalent setting for Windows?


    Cheers, Michiel

    --- GoldED+/W32-MSVC 1.1.5-b20130111
    * Origin: he.net certified sage (2:280/5555)
  • From Tommi Koivula@2:221/1.1 to Michiel van der Vlist on Sun Feb 5 09:29:40 2017
    Hi Michiel.

    05 Feb 17 00:44:34, you wrote to me:

    OK, so that way one can make an IPv6 connect to a 6to4 tunneled
    node. But that does not nullify my feature request for -64 and
    -46 options. What you did is changing the OS preference. So that
    will affect /all/ connections. Not just binkp. What I proposed
    was an override on the OS preference for a specific connection.
    I.e a binkp connect to a specific node.

    Of course. But this is ipv6 echo, not binkd echo. :)

    Indeed. Is there an equivalent setting for Windows?

    You can start with

    netsh interface ipv6 show prefixpolicies

    and continue with

    netsh interface ipv6 set prefixpolicy

    and then you may do some googling. ;)

    'Tommi

    ... he.net certified sage
    ---
    * Origin: IPv6 Point at [2001:470:1f15:cb0::4] (2:221/1.1)
  • From Michiel van der Vlist@2:280/5555 to Tommi Koivula on Tue Feb 7 18:08:27 2017
    Hello Tommi,

    On Sunday February 05 2017 09:29, you wrote to me:

    You can start with

    netsh interface ipv6 show prefixpolicies
    ^^^
    D:\FIDO\FMAIL>netsh int ipv6 show prefixpolicy
    Der aktive Status wird abgefragt...

    Vorg„nger Label Pr„fix
    ---------- ----- --------------------------------
    5 5 2001::/32
    10 4 ::ffff:0:0/96
    20 3 ::/96
    30 2 2002::/16
    40 1 ::/0
    50 0 ::1/128

    and continue with

    netsh interface ipv6 set prefixpolicy

    The problem with netsh is that its documentation is lousy. :(

    My guess is that removing the entry for 2002::/16 will do the trick.

    and then you may do some googling. ;)

    I'd better do that before I make changes that I do not know how to undo...


    Cheers, Michiel

    --- GoldED+/W32-MSVC 1.1.5-b20130111
    * Origin: he.net certified sage (2:280/5555)
  • From Tommi Koivula@2:221/360 to Michiel van der Vlist on Wed Feb 8 12:18:51 2017
    Michiel van der Vlist wrote:

    netsh interface ipv6 show prefixpolicies
    MvdV> ^^^
    D:\FIDO\FMAIL>> netsh int ipv6 show prefixpolicy
    MvdV> Der aktive Status wird abgefragt...

    So you still have win XP.

    I have number seven:

    netsh interface ipv6 show prefixpolicies
    Querying active state...

    Precedence Label Prefix
    ---------- ----- --------------------------------
    50 0 ::1/128
    40 1 ::/0
    30 2 2002::/16
    20 3 ::/96
    10 4 ::ffff:0:0/96
    5 5 2001::/32

    I cannot understand why these netsh commands had to be changed between win versions.

    'Tommi

    --- Mozilla/5.0 (Windows NT 6.1; Win64; x64; rv:51.0) Gecko/20100101 SeaMonkey/
    * Origin: *** nntp://rbb.bbs.fi *** Lake Ylo *** Finland *** (2:221/360)
  • From Michiel van der Vlist@2:280/5555 to Tommi Koivula on Fri Feb 10 11:27:26 2017
    Hello Tommi,

    On Wednesday February 08 2017 12:18, you wrote to me:

    netsh interface ipv6 show prefixpolicies
    MvdV>> ^^^
    D:\FIDO\FMAIL>> netsh int ipv6 show prefixpolicy
    MvdV>> Der aktive Status wird abgefragt...

    So you still have win XP.

    Getting rid of XP is on my todo list. Its IPv6 implementation works, but it is incomplete. But I want to get rid of the 16 bit stuff before I make that plunge.

    I have number seven:

    I too have 7 on my other machine.


    netsh interface ipv6 show prefixpolicies
    Querying active state...

    Precedence Label Prefix
    ---------- ----- --------------------------------
    50 0 ::1/128
    40 1 ::/0
    30 2 2002::/16
    20 3 ::/96
    10 4 ::ffff:0:0/96
    5 5 2001::/32

    I cannot understand why these netsh commands had to be changed between
    win versions.

    Neither do I.


    Cheers, Michiel

    --- GoldED+/W32-MSVC 1.1.5-b20130111
    * Origin: he.net certified sage (2:280/5555)