• Testing with BinkD...

    From Rj Clay@1:120/546 to All on Mon Jul 21 11:01:48 2014
    All,

    I'm still not sure if my IPv6 connectivity here at home is working well yet but I plan to start testing next week using BinkD v1.02 (which has been uploaded to Debian unstable and will likely transition to Debian testing ('Jessie') by next week.



    Jame

    --- BBBS/Li6 v4.10 Dada-1
    * Origin: BBBS Info at Rocasa (1:120/546)
  • From Michiel van der Vlist@2:280/5555 to Rj Clay on Tue Jul 22 10:26:16 2014
    Hello Rj,

    On Monday July 21 2014 11:01, you wrote to All:

    I'm still not sure if my IPv6 connectivity here at home is working
    well yet but I plan to start testing next week using BinkD v1.02
    (which has been uploaded to Debian unstable and will likely transition
    to Debian testing ('Jessie') by next week.

    When you are unsure about your IPv6 connectivity, why test it with an unstable binkd version? binkd version 1.1a-49 has been well tested by now with IPv6.


    Cheers, Michiel

    --- GoldED+/W32-MSVC 1.1.5-b20130111
    * Origin: 2001:470:1f15:1117::1 (2:280/5555)
  • From Markus Reschke@2:240/1661 to Michiel van der Vlist on Tue Jul 22 21:20:10 2014
    Hi Michiel!

    Jul 22 10:26 2014, Michiel van der Vlist wrote to Rj Clay:

    I'm still not sure if my IPv6 connectivity here at home is working
    well yet but I plan to start testing next week using BinkD v1.02
    (which has been uploaded to Debian unstable and will likely
    transition to Debian testing ('Jessie') by next week.

    MvdV> When you are unsure about your IPv6 connectivity, why test it with
    MvdV> an unstable binkd version? binkd version 1.1a-49 has been well
    MvdV> tested by now with IPv6.

    I'm running 1.02 for nearly three months now and haven't got any complaints yet. That version seems to be fine.

    Regards,
    Markus

    ---
    * Origin: *** theca tabellaria *** (2:240/1661)
  • From Michiel van der Vlist@2:280/5555 to Markus Reschke on Wed Jul 23 00:11:56 2014
    Hello Markus,

    On Tuesday July 22 2014 21:20, you wrote to me:

    MvdV>> When you are unsure about your IPv6 connectivity, why test it
    MvdV>> with an unstable binkd version? binkd version 1.1a-49 has been
    MvdV>> well tested by now with IPv6.

    I'm running 1.02 for nearly three months now and haven't got any complaints yet. That version seems to be fine.

    The Linux or the Win version?


    Cheers, Michiel

    --- GoldED+/W32-MSVC 1.1.5-b20130111
    * Origin: 2001:470:1f15:1117::1 (2:280/5555)
  • From Benny Pedersen@1:261/38.20 to Rj Clay on Wed Jul 23 01:18:46 2014
    Hello Rj!

    21 Jul 2014 11:01, Rj Clay wrote to All:

    I'm still not sure if my IPv6 connectivity here at home is working
    well yet but I plan to start testing next week using BinkD v1.02
    (which has been uploaded to Debian unstable and will likely transition
    to Debian testing ('Jessie') by next week.

    if its not working its fault in nodelist or /etc/gai.conf with ipv4/ipv6 preferences

    all else is dokumented in logs :=)

    for nodelist data use:

    ----- binkd.cfg begins -----
    root-domain binkp.net
    node 2:230/0 * password
    ----- binkd.cfg ends -----

    then binkp will try ipv4/ipv6 as /etc/gai.conf says if node is dual stacked

    default is btw ipv4 first, this is not a fault !

    ips should never be in nodelist, this would be a fault


    Regards Benny

    ... there can only be one way of life, and it works :)

    --- Msged/LNX 6.2.0 (Linux/3.12.21-gentoo-r1 (i686))
    * Origin: duggi.junc.org where qico is waiting (1:261/38.20)
  • From Benny Pedersen@1:261/38.20 to Michiel van der Vlist on Wed Jul 23 01:26:46 2014
    Hello Michiel!

    22 Jul 2014 10:26, Michiel van der Vlist wrote to Rj Clay:

    MvdV> When you are unsure about your IPv6 connectivity, why test it with
    MvdV> an unstable binkd version? binkd version 1.1a-49 has been well
    MvdV> tested by now with IPv6.

    debian :=)

    i dont get it either :/


    Regards Benny

    ... there can only be one way of life, and it works :)

    --- Msged/LNX 6.2.0 (Linux/3.12.21-gentoo-r1 (i686))
    * Origin: duggi.junc.org where qico is waiting (1:261/38.20)
  • From Alexey Vissarionov@2:5020/545 to Benny Pedersen on Wed Jul 23 08:20:20 2014
    Good ${greeting_time}, Benny!

    23 Jul 2014 01:18:46, you wrote to Rj Clay:

    for nodelist data use:
    root-domain binkp.net
    node 2:230/0 * password

    #include <FTA-1006>

    Very dangerous advise. All persistent links _should_ be defined explicitly, otherwise you may end up with no connectivity once something will happen to root-domain (as it once happened to fidonet.net).

    then binkp will try ipv4/ipv6 as /etc/gai.conf says if node is dual stacked

    default is btw ipv4 first, this is not a fault !

    It is. IPv6 _must_ be tried before IPv4. Changing this precedence is possible, but not advised outside of LANs.

    ips should never be in nodelist, this would be a fault

    No.
    However, using the hostnames is _recommended_.


    --
    Alexey V. Vissarionov aka Gremlin from Kremlin
    gremlin.ru!gremlin; +vii-cmiii-ccxxix-lxxix-xlii

    ... :wq!
    --- /bin/vi
    * Origin: http://openwall.com/Owl (2:5020/545)
  • From Alexey Vissarionov@2:5020/545 to Benny Pedersen on Wed Jul 23 08:33:00 2014
    Good ${greeting_time}, Benny!

    23 Jul 2014 01:26:46, you wrote to Michiel van der Vlist:

    MvdV>> When you are unsure about your IPv6 connectivity, why test it
    MvdV>> with an unstable binkd version? binkd version 1.1a-49 has been
    MvdV>> well tested by now with IPv6.
    debian :=)

    This brain-dead "OS" may have side-effects even with stable software.

    i dont get it either :/

    IPv6 connectivity, or?


    --
    Alexey V. Vissarionov aka Gremlin from Kremlin
    gremlin.ru!gremlin; +vii-cmiii-ccxxix-lxxix-xlii

    ... god@universe:~ # cvs up && make world
    --- /bin/vi
    * Origin: http://openwall.com/Owl (2:5020/545)
  • From Benny Pedersen@1:261/38.20 to Alexey Vissarionov on Wed Jul 23 14:54:34 2014
    Hello Alexey!

    23 Jul 2014 08:20, Alexey Vissarionov wrote to Benny Pedersen:

    for nodelist data use:
    root-domain binkp.net
    node 2:230/0 * password

    #include <FTA-1006>

    Very dangerous advise. All persistent links _should_ be defined explicitly, otherwise you may end up with no connectivity once
    something will happen to root-domain (as it once happened to fidonet.net).

    is there example logs when it did that ?, example logs where it happens to binkp.net aswell ?

    then binkp will try ipv4/ipv6 as /etc/gai.conf says if node is dual
    stacked

    default is btw ipv4 first, this is not a fault !

    It is. IPv6 _must_ be tried before IPv4. Changing this precedence is possible, but not advised outside of LANs.

    you deside in /etc/gai.conf

    default is ipv4 first

    ips should never be in nodelist, this would be a fault

    No.

    yep it will kill dual stacking

    However, using the hostnames is _recommended_.

    its the same what i meant, dammit :)

    but nodelist support both ipv4 and ipv6 hostnames, what an insight waste :/


    Regards Benny

    ... there can only be one way of life, and it works :)

    --- Msged/LNX 6.2.0 (Linux/3.12.21-gentoo-r1 (i686))
    * Origin: duggi.junc.org where qico is waiting (1:261/38.20)
  • From Benny Pedersen@1:261/38.20 to Alexey Vissarionov on Wed Jul 23 14:58:50 2014
    Hello Alexey!

    23 Jul 2014 08:33, Alexey Vissarionov wrote to Benny Pedersen:

    This brain-dead "OS" may have side-effects even with stable software.

    its like android, where defaults apps is in kernel :(

    i dont get it either :/
    IPv6 connectivity, or?

    current node here no, but try linode.junc.eu :)


    Regards Benny

    ... there can only be one way of life, and it works :)

    --- Msged/LNX 6.2.0 (Linux/3.12.21-gentoo-r1 (i686))
    * Origin: duggi.junc.org where qico is waiting (1:261/38.20)
  • From Markus Reschke@2:240/1661 to Michiel van der Vlist on Wed Jul 23 19:34:48 2014
    Hi Michiel!

    Jul 23 00:11 2014, Michiel van der Vlist wrote to Markus Reschke:

    MvdV> The Linux or the Win version?

    Linux.

    Regards,
    Markus

    ---
    * Origin: *** theca tabellaria *** (2:240/1661)
  • From Michiel van der Vlist@2:280/5555 to Benny Pedersen on Wed Jul 23 23:52:10 2014
    Hello Benny,

    On Wednesday July 23 2014 14:54, you wrote to Alexey Vissarionov:

    ips should never be in nodelist, this would be a fault

    No.

    yep it will kill dual stacking

    No, it does not. We have discussed this before. Using a hostname is the prefered method when possible, but using two INA flags, one with the IPv6 and one with the IPv4 address is complient with FTS-5000 and FTSS-5001.


    Cheers, Michiel

    --- GoldED+/W32-MSVC 1.1.5-b20130111
    * Origin: 2001:470:1f15:1117::1 (2:280/5555)
  • From Michiel van der Vlist@2:280/5555 to Alexey Vissarionov on Thu Jul 24 00:21:01 2014
    Hello Alexey,

    On Wednesday July 23 2014 08:20, you wrote to Benny Pedersen:

    for nodelist data use:
    root-domain binkp.net
    node 2:230/0 * password

    Very dangerous advise. All persistent links _should_ be defined explicitly, otherwise you may end up with no connectivity once
    something will happen to root-domain (as it once happened to
    fidonet.net).

    True. OTOH, very few IP mailers follow the nodelist. So if you explicitly define the host name in the mailer config, it will not automatically be updated when it changes in the nodelist and connectivity with that node will be lost. It will not be restored until the sysop intervenes by manually editing the mailer's config file. We have seen examples of this when dyndns changed its policy and many sysops had to change their host names.

    binkp.net DOES follow the nodelist, so With the binkp.net method this problem is avoided.

    Each mathod has its pros band cons.

    then binkp will try ipv4/ipv6 as /etc/gai.conf says if node is
    dual stacked

    default is btw ipv4 first, this is not a fault !

    It is. IPv6 _must_ be tried before IPv4. Changing this precedence is possible, but not advised outside of LANs.

    Opinions differ. Ever heard op happy eyeballs?

    For FOIP it is no big deal if there is a 60 second delay because the nodelist advertises IPv6 connectivity but it is not there and it only falls back to IPv4 after a time-out. For other applications this can be a show stopper.


    Cheers, Michiel

    --- GoldED+/W32-MSVC 1.1.5-b20130111
    * Origin: 2001:470:1f15:1117::1 (2:280/5555)
  • From Konstantin Kuzov@2:5019/40 to Benny Pedersen on Wed Jul 23 12:57:28 2014
    Konnichi wa, *Benny-kun*! Aogu manako oyobi uketamawaru waga koe!
    Tomodachi _Benny Pedersen_ tsukuru airon _Rj Clay_
    Nichiji - /*23 ˆî« 14 01:18*/, Daizai - /*Testing with BinkD...*/:

    then binkp will try ipv4/ipv6 as /etc/gai.conf says if node is dual stacked

    default is btw ipv4 first, this is not a fault !

    You are wrong. Default gai.conf configuration prefers ipv6.

    If your distribution comes with gai.conf that prefers ipv4 over ipv6, then distribution maintainers have a brain desease.

    Ganbatte, *Benny*!

    [_N0SF3R@TU_]
    ... GoldED+/LNX 1.1.5 (Linux 3.11.5-gentoo CPU UNKNOWN)
    --- #[Kaori Sekken: Master.NoSFeRaTU[@]Gmail.com] [Kumi Nyaa]#
    * Origin: Ojisan, oriru mottekuru suna oyobi korosu sagaru kabe (2:5019/40)
  • From RJ Clay@1:120/544 to Benny Pedersen on Thu Jul 24 02:33:15 2014
    Benny,

    23 Jul 14 01:18, you wrote to me:

    21 Jul 2014 11:01, Rj Clay wrote to All:

    I'm still not sure if my IPv6 connectivity here at home is
    working well yet but I plan to start testing next week using
    BinkD v1.02 (which has been uploaded to Debian unstable and will
    likely transition to Debian testing ('Jessie') by next week.

    if its not working its fault in nodelist or /etc/gai.conf with
    ipv4/ipv6 preferences

    Won't have an idea until I can do testing...





    Jame


    --- GoldED+/LNX 1.1.5-b20110213
    * Origin: ftn.rocasa.net (1:120/544)
  • From RJ Clay@1:120/544 to Alexey Vissarionov on Thu Jul 24 02:46:21 2014
    Alexey.

    23 Jul 14 08:20, you wrote to Benny Pedersen:

    then binkp will try ipv4/ipv6 as /etc/gai.conf says if node is
    dual stacked

    default is btw ipv4 first, this is not a fault !

    It is. IPv6 _must_ be tried before IPv4.

    Why?


    Changing this precedence is
    possible, but not advised outside of LANs.

    Which is where most people are...




    Jame




    --- GoldED+/LNX 1.1.5-b20110213
    * Origin: ftn.rocasa.net (1:120/544)
  • From RJ Clay@1:120/544 to Alexey Vissarionov on Thu Jul 24 02:48:07 2014
    Alexey,

    23 Jul 14 08:33, you wrote to Benny Pedersen:
    23 Jul 2014 01:26:46, you wrote to Michiel van der Vlist:

    MvdV>>> When you are unsure about your IPv6 connectivity, why test it
    MvdV>>> with an unstable binkd version? binkd version 1.1a-49 has been
    MvdV>>> well tested by now with IPv6.

    debian :=)

    This brain-dead "OS" may have side-effects even with stable software.

    "brain-dead"? Rather that than, for instance, centos...<g>


    Jame

    --- GoldED+/LNX 1.1.5-b20110213
    * Origin: ftn.rocasa.net (1:120/544)
  • From Rj Clay@1:120/546 to Michiel Van Der Vlist on Thu Jul 24 02:17:06 2014
    Michiel,

    On Monday July 21 2014 11:01, you wrote to All:
    I'm still not sure if my IPv6 connectivity here at home is working
    well yet but I plan to start testing next week using BinkD v1.02
    (which has been uploaded to Debian unstable and will likely transition
    to Debian testing ('Jessie') by next week.

    When you are unsure about your IPv6 connectivity, why test it with an unstable binkd version? binkd version 1.1a-49 has been well tested by
    now with IPv6.

    Who said anything about an unstable BinkD version? I referred to Debian Unstable (aka 'sid'), which in case you really don't know is where all new packages are uploaded to Debian.




    Jame



    --- BBBS/Li6 v4.10 Dada-1
    * Origin: BBBS Info at Rocasa (1:120/546)
  • From Michiel van der Vlist@2:280/5555 to RJ Clay on Thu Jul 24 12:20:03 2014
    Hello RJ,

    On Thursday July 24 2014 02:46, you wrote to Alexey Vissarionov:

    default is btw ipv4 first, this is not a fault !

    It is. IPv6 _must_ be tried before IPv4.

    Why?

    To promote IP6. It is not a hard must, just a recomendation.

    From your AKA's, 1:120/546 is the one that carries a host name (bbbb.rocasa.biz) that carries an AAAA record (2001:470:1f11:10b9::13). I tried to do a binkp connect but it failed.

    + 12:18 [1472] call to 1:120/546@fidonet
    12:18 [1472] trying bbbs.rocasa.biz [2001:470:1f11:10b9::13]...
    ? 12:18 [1472] connection to 1:120/546@fidonet failed: {W32 API error 10060}
    Connection timed out
    12:18 [1472] trying bbbs.rocasa.biz [24.11.201.119]...
    ? 12:19 [1472] connection to 1:120/546@fidonet failed: {W32 API error 10052}
    A socket operation was attempted to an unreachable network

    Looks like an IPv6 firewall problem. My binkd can not connect woth IPv4 either. I do not undertstand why.

    Changing this precedence is possible, but not advised outside of
    LANs.

    Which is where most people are...

    He means "but not advised when connecting to a host outside the LAN".


    Cheers, Michiel

    --- GoldED+/W32-MSVC 1.1.5-b20130111
    * Origin: 2001:470:1f15:1117::1 (2:280/5555)
  • From Benny Pedersen@1:261/38.20 to Michiel van der Vlist on Thu Jul 24 13:34:52 2014
    Hello Michiel!

    23 Jul 2014 23:52, Michiel van der Vlist wrote to Benny Pedersen:

    MvdV> No, it does not. We have discussed this before. Using a hostname is
    MvdV> the prefered method when possible,

    +1

    MvdV> but using two INA flags, one with
    MvdV> the IPv6 and one with the IPv4 address is complient with FTS-5000
    MvdV> and FTSS-5001.

    this is workarounds, its the same as have 2 hostnames that are single stacked, will work aswell yes, but badly dns/nodelist waste

    it works as designed :/


    Regards Benny

    ... there can only be one way of life, and it works :)

    --- Msged/LNX 6.2.0 (Linux/3.12.21-gentoo-r1 (i686))
    * Origin: duggi.junc.org where qico is waiting (1:261/38.20)
  • From Benny Pedersen@1:261/38.20 to Konstantin Kuzov on Thu Jul 24 13:39:52 2014
    Hello Konstantin!

    23 Jul 2014 12:57, Konstantin Kuzov wrote to Benny Pedersen:

    You are wrong. Default gai.conf configuration prefers ipv6.

    is your config only have # lines ?, if so its defaults for the vanilla glibc, if your glibc is patched, then its not default ipv6 first maybe, its opensource so we might both have diffrent problems with defaults :=)

    If your distribution comes with gai.conf that prefers ipv4 over ipv6, then distribution maintainers have a brain desease.

    i had to edit gai.conf to prefer ipv6 first

    Ganbatte, *Benny*!

    +1

    ... GoldED+/LNX 1.1.5 (Linux 3.11.5-gentoo CPU UNKNOWN)

    oh note unknown :)


    Regards Benny

    ... there can only be one way of life, and it works :)

    --- Msged/LNX 6.2.0 (Linux/3.12.21-gentoo-r1 (i686))
    * Origin: duggi.junc.org where qico is waiting (1:261/38.20)
  • From Benny Pedersen@1:261/38.20 to RJ Clay on Thu Jul 24 13:44:54 2014
    Hello RJ!

    24 Jul 2014 02:33, RJ Clay wrote to Benny Pedersen:

    if its not working its fault in nodelist or /etc/gai.conf with
    ipv4/ipv6 preferences

    Won't have an idea until I can do testing...

    yep fair, there is so much false info out there it could be writed big books about it :=)


    Regards Benny

    ... there can only be one way of life, and it works :)

    --- Msged/LNX 6.2.0 (Linux/3.12.21-gentoo-r1 (i686))
    * Origin: duggi.junc.org where qico is waiting (1:261/38.20)
  • From Benny Pedersen@1:261/38.20 to RJ Clay on Thu Jul 24 13:46:32 2014
    Hello RJ!

    24 Jul 2014 02:46, RJ Clay wrote to Alexey Vissarionov:

    then binkp will try ipv4/ipv6 as /etc/gai.conf says if node is
    dual stacked default is btw ipv4 first, this is not a fault !
    It is. IPv6 _must_ be tried before IPv4.
    Why?

    use the one that have lovest ping time, eg keep away from the one that have longest tcp route

    if you have ipv6 native not via tunnels then there could not be much differings in ping time, but if ipv6 is on tunnels it could be a big differings, same goes for if vpn ipv4 :=)

    Changing this precedence is
    possible, but not advised outside of LANs.

    incorrect, defaults is good, if its not make a bug report in glibc

    Which is where most people are...

    +1


    Regards Benny

    ... there can only be one way of life, and it works :)

    --- Msged/LNX 6.2.0 (Linux/3.12.21-gentoo-r1 (i686))
    * Origin: duggi.junc.org where qico is waiting (1:261/38.20)
  • From Benny Pedersen@1:261/38.20 to RJ Clay on Thu Jul 24 13:52:16 2014
    Hello RJ!

    24 Jul 2014 02:48, RJ Clay wrote to Alexey Vissarionov:

    This brain-dead "OS" may have side-effects even with stable software.
    "brain-dead"? Rather that than, for instance, centos...<g>

    +1000


    Regards Benny

    ... there can only be one way of life, and it works :)

    --- Msged/LNX 6.2.0 (Linux/3.12.21-gentoo-r1 (i686))
    * Origin: duggi.junc.org where qico is waiting (1:261/38.20)
  • From Michiel van der Vlist@2:280/5555 to Benny Pedersen on Fri Jul 25 01:30:24 2014
    Hello Benny,

    On Thursday July 24 2014 13:34, you wrote to me:

    MvdV>> No, it does not. We have discussed this before. Using a
    MvdV>> hostname is the prefered method when possible,

    +1

    So we agree that using a host name is the preferred method. Also note that "preferred method" implies it is not the only method. Sometimes a host name may not be available. In that case there is a "plan B": Using literal IP addresses.

    MvdV>> but using two INA flags, one with the IPv6 and one with the IPv4
    MvdV>> address is complient with FTS-5000 and FTSS-5001.

    this is workarounds,

    Call it what you want, the fact is that it is FTSC complient.

    its the same as have 2 hostnames that are single stacked,

    No, it is not. Hostnames can resolve to more than one IP address, an IP address can not.

    will work aswell yes, but badly dns/nodelist waste

    I disagree. Using host names is to make things easier for humans. Humans are better in remembering names than numbers. Especially big numbers. But machines do not have that limitation. Machines do not need names, they can remember the numbers just as well. From the machine's POV using host names is ineficient. Bypassing the dns system and dealing directly with IP numbers is much more efficient from the machine's POV. The information in the nodelist is primarily meant for machines. So...

    it works as designed :/

    Yes, it does.


    Cheers, Michiel

    --- GoldED+/W32-MSVC 1.1.5-b20130111
    * Origin: 2001:470:1f15:1117::1 (2:280/5555)
  • From Konstantin Kuzov@2:5019/40 to Benny Pedersen on Thu Jul 24 20:27:34 2014
    Konnichi wa, *Benny-kun*! Aogu manako oyobi uketamawaru waga koe!
    Tomodachi _Benny Pedersen_ tsukuru airon _Konstantin Kuzov_
    Nichiji - /*24 ˆî« 14 13:39*/, Daizai - /*Testing with BinkD...*/:

    You are wrong. Default gai.conf configuration prefers ipv6.
    is your config only have # lines ?, if so its defaults for the vanilla glibc, if your glibc is patched, then its not default ipv6 first
    maybe, its opensource so we might both have diffrent problems with defaults :=)

    Yes, all lines were commented by default, until I tweak them to my preference. Currently I have sys-libs/glibc-2.19-r1 from official portage without vanilla flag, so it is actually patched version. ^_^

    If your distribution comes with gai.conf that prefers ipv4 over
    ipv6, then distribution maintainers have a brain desease.
    i had to edit gai.conf to prefer ipv6 first

    What distribution are you using?

    ... GoldED+/LNX 1.1.5 (Linux 3.11.5-gentoo CPU UNKNOWN)
    oh note unknown :)

    Thats because golded+ doesn't support cpu detection on x86_64 systems. Updated to my golded version which actually can. ^_^

    Ganbatte, *Benny*!

    [_N0SF3R@TU_]
    ... GoldED-NSF/LNX 1.1.5-b20140107 (Linux 3.11.5-gentoo iF6M23)
    --- #[Kaori Sekken: Master.NoSFeRaTU[@]Gmail.com] [Kumi Nyaa]#
    * Origin: Ojisan, oriru mottekuru suna oyobi korosu sagaru kabe (2:5019/40)
  • From Rj Clay@1:120/546 to Michiel Van Der Vlist on Mon Jul 28 11:53:36 2014
    Michiel,

    From your AKA's, 1:120/546 is the one that carries a host name (bbbb.rocasa.biz) that carries an AAAA record (2001:470:1f11:10b9::13).
    I tried to do a binkp connect but it failed.

    That should be bbbs.rocasa.biz. The FTN app there doesn't do IPv6 but I was planning to experiment with IPv6 IP proxies. Obviously I need to get the IPv6 connectivity working first...<g>


    My binkd can not connect woth IPv4 either I do not undertstand why.

    Wrong IP; that I need to update again...





    Jame


    --- BBBS/Li6 v4.10 Dada-1
    * Origin: BBBS Info at Rocasa (1:120/546)
  • From Rj Clay@1:120/546 to Benny Pedersen on Mon Jul 28 12:18:10 2014
    Benny,

    if you have ipv6 native

    I will have, but not here at home; and that system I've just started setting up, so it'll be a bit before I can deploy it.


    not via tunnels then there could not be much differing
    in ping time, but if ipv6 is on tunnels it could be a big differings,

    My ISP here at home (Comcast) does have native IPv6 in some states but not here, so I'll have to use IPv6 tunnels. I do have some IPv6 connectivity from on my workstation using a teredo tunnel but I don't know reliable it is... (I changed the gai defaults to IPv4 because it was taking so long to do package updates to the IPv6 connected repositories...)

    Before I had to move a couple of years ago, I had an HE tunnel being managed by my wrt54gl; I'll be seeing if I can get something working with that again.



    Jame

    --- BBBS/Li6 v4.10 Dada-1
    * Origin: BBBS Info at Rocasa (1:120/546)
  • From Markus Reschke@2:240/1661 to Rj Clay on Tue Jul 29 12:17:54 2014
    Hi Jame!

    Jul 28 12:18 2014, Rj Clay wrote to Benny Pedersen:

    here, so I'll have to use IPv6 tunnels. I do have some IPv6
    connectivity from on my workstation using a teredo tunnel but I don't know reliable it is... (I changed the gai defaults to IPv4 because it
    was taking so long to do package updates to the IPv6 connected repositories...)

    You've done the right thing. Teredo was a temporary transition strategy and shouldn't be used anymore. AFAIK Microsoft is going to shut down their Teredo servers this year.

    Before I had to move a couple of years ago, I had an HE tunnel
    being managed
    by my wrt54gl; I'll be seeing if I can get something working with
    that again.

    The HE tunnels work quite reliable. I did use one too until I got native IPv6.

    Regards,
    Markus

    ---
    * Origin: *** theca tabellaria *** (2:240/1661)
  • From Michiel van der Vlist@2:280/5555 to Rj Clay on Tue Jul 29 14:31:32 2014
    Hello Rj,

    On Monday July 28 2014 11:53, you wrote to me:

    From your AKA's, 1:120/546 is the one that carries a host name
    (bbbb.rocasa.biz) that carries an AAAA record (2001:470:1f11:10b9::13).
    I tried to do a binkp connect but it failed.

    That should be bbbs.rocasa.biz.

    Yeah, typo.In the binkd config I got it right.

    The FTN app there doesn't do IPv6 but I was planning to experiment
    with IPv6 IP proxies. Obviously I need to get the IPv6 connectivity working first...<g>

    And until you have, you should not advertise IPv6 connectivity in the nodelist...

    My binkd can not connect woth IPv4 either I do not undertstand why.

    Wrong IP; that I need to update again...

    It works now:

    14:33 [3720] creating a poll for 1:120/546@fidonet (`d' flavour)
    14:33 [3720] clientmgr started
    + 14:33 [4088] call to 1:120/546@fidonet
    14:33 [4088] trying bbbs.rocasa.biz [2001:470:1f11:10b9::13]...
    ? 14:33 [4088] connection to 1:120/546@fidonet failed: {W32 API error 10060}
    Connection timed out
    14:33 [4088] trying bbbs.rocasa.biz [50.133.70.48]...
    14:33 [4088] connected
    + 14:33 [4088] outgoing session with 50.133.70.48
    - 14:33 [4088] SYS Rocasa BBBS
    - 14:33 [4088] ZYZ Robert Clay
    - 14:33 [4088] LOC Lansing, MI, USA
    - 14:33 [4088] VER BBBS/Li6 v4.10 Dada-1 binkp/1.0
    - 14:33 [4088] NDL XX,IBN
    + 14:33 [4088] addr: 1:120/546@fidonet
    + 14:33 [4088] addr: 8:7480/2@fmlynet (n/a or busy)
    + 14:33 [4088] addr: 110:205/2@linuxnet (n/a or busy)
    + 14:33 [4088] addr: 111:4150/402@stn (n/a or busy)
    + 14:33 [4088] addr: 618:200/17@micronet (n/a or busy)
    + 14:33 [4088] pwd protected session (plain text)
    + 14:33 [4088] done (to 1:120/546@fidonet, OK, S/R: 0/0 (0/0 bytes))
    14:33 [4088] session closed, quitting...
    14:33 [3720] the queue is empty, quitting...


    Cheers, Michiel

    --- GoldED+/W32-MSVC 1.1.5-b20130111
    * Origin: 2001:470:1f15:1117::1 (2:280/5555)
  • From Michiel van der Vlist@2:280/5555 to Rj Clay on Tue Jul 29 14:35:56 2014
    Hello Rj,

    On Monday July 28 2014 12:18, you wrote to Benny Pedersen:

    My ISP here at home (Comcast) does have native IPv6 in some states
    but not here, so I'll have to use IPv6 tunnels. I do have some IPv6 connectivity from on my workstation using a teredo tunnel but I don't
    know reliable it is...

    Forget about Teredo, it is VERY unreliable. I never got it to work properly.

    The best thing to do is disable Teredo.

    Before I had to move a couple of years ago, I had an HE tunnel being managed by my wrt54gl; I'll be seeing if I can get something working
    with that again.

    I have a he.net tunnel running that ends in my wrt54gl. Works like a charm. ;-)

    See my articles in Fidonews.


    Cheers, Michiel

    --- GoldED+/W32-MSVC 1.1.5-b20130111
    * Origin: 2001:470:1f15:1117::1 (2:280/5555)
  • From Michiel van der Vlist@2:280/5555 to Markus Reschke on Tue Jul 29 14:54:37 2014
    Hello Markus,

    On Tuesday July 29 2014 12:17, you wrote to Rj Clay:

    You've done the right thing. Teredo was a temporary transition
    strategy and shouldn't be used anymore.

    I never got it to work properly.

    AFAIK Microsoft is going to shut down their Teredo servers this year.

    Maybe they already have. teredo.ipv6.microsoft.com resolves to an IP address, but does not ping from here.

    D:\FIDO\Golded>ping teredo.ipv6.microsoft.com

    Ping teredo.ipv6.microsoft.com.nsatc.net [157.56.106.189] mit 32 Bytes Daten:

    Zeitberschreitung der Anforderung.
    Zeitberschreitung der Anforderung.
    Zeitberschreitung der Anforderung.
    Zeitberschreitung der Anforderung.

    Ping-Statistik fr 157.56.106.189:
    Pakete: Gesendet = 4, Empfangen = 0, Verloren = 4 (100% Verlust),

    They tested teredo sunset last year: http://www.ietf.org/proceedings/87/slides/slides-87-v6ops-5.pdf

    I have not seen a definite date for the final shut down. They may as well do it right now if they didn't allready...

    The HE tunnels work quite reliable. I did use one too until I got
    native IPv6.

    My ISP is still dragging its feet. Until then I make do with a he.net tunnel and a SixXs tunnel. Both work reliable but not with the same speed as native IPv4.


    Cheers, Michiel

    --- GoldED+/W32-MSVC 1.1.5-b20130111
    * Origin: 2001:470:1f15:1117::1 (2:280/5555)
  • From Markus Reschke@2:240/1661 to Michiel van der Vlist on Wed Jul 30 18:36:34 2014
    Hi Michiel!

    Jul 29 14:54 2014, Michiel van der Vlist wrote to Markus Reschke:

    AFAIK Microsoft is going to shut down their Teredo servers this year.

    MvdV> Maybe they already have. teredo.ipv6.microsoft.com resolves to an
    MvdV> IP address, but does not ping from here.

    Could also be caused by a firewall or host settings. A missing echo reply doesn't imply that there's no host.

    MvdV> I have not seen a definite date for the final shut down. They may
    MvdV> as well do it right now if they didn't allready...

    Microsoft announced at IETF88 that they are going to shutdown the Teredo servers the first half of 2014. Maybe they already have.

    MvdV> My ISP is still dragging its feet. Until then I make do with a
    MvdV> he.net tunnel and a SixXs tunnel. Both work reliable but not with
    MvdV> the same speed as native IPv4.

    Be patient! ;) You'll get native IPv6 sooner or later. No worries!

    Regards,
    Markus

    ---
    * Origin: *** theca tabellaria *** (2:240/1661)
  • From Michiel van der Vlist@2:280/5555 to Markus Reschke on Thu Jul 31 02:24:56 2014
    Hello Markus,

    On Wednesday July 30 2014 18:36, you wrote to me:

    AFAIK Microsoft is going to shut down their Teredo servers this
    year.

    MvdV>> Maybe they already have. teredo.ipv6.microsoft.com resolves to
    MvdV>> an IP address, but does not ping from here.

    Could also be caused by a firewall or host settings. A missing echo
    reply doesn't imply that there's no host.

    Of course. Disabeling or shielding off ping is considered a safety measure by some, but that idea is not shared by the professionals and considered of no added value when one is running servers. So I would expect a server run by Microsoft to respond to a ping.

    MvdV>> I have not seen a definite date for the final shut down. They
    MvdV>> may as well do it right now if they didn't allready...

    Microsoft announced at IETF88 that they are going to shutdown the
    Teredo servers the first half of 2014. Maybe they already have.

    Well, we are past the first half of 2014, so maybe they have indeed switched them off. I would have expected an announcement if they actually had though...

    MvdV>> My ISP is still dragging its feet. Until then I make do with a
    MvdV>> he.net tunnel and a SixXs tunnel. Both work reliable but not
    MvdV>> with the same speed as native IPv4.

    Be patient! ;) You'll get native IPv6 sooner or later. No worries!

    I know I will get it sooner or later. One way or another. But my patience is really challenged. There is some hope though. My ISP runs its own speedtest: speedtest.ziggo.nl. Up to about a month ago, it was IPv4 only. But now there are two versions: speedtest6.ziggo.nl and speedtest4.ziggo.nl.

    If you call speedtest.ziggo.nl from an IPv4 only client, it is a it ever was. If you call it from a dual stack client, you get the IPv6 version which has a field to click on for the IPv4 version.

    Also their webiste www.ziggo.nl has an IPv6 address.

    So /something/ is happening...


    Cheers, Michiel

    --- GoldED+/W32-MSVC 1.1.5-b20130111
    * Origin: 2001:470:1f15:1117::1 (2:280/5555)