Re: NT and Cornell V1.0 TCP problem

Connecticut Couple (
Sat, 09 May 1998 09:02:35 PDT


I am currently experiencing the same problem with a Win95 system, again
with an ethernet card with a 10.x.x.x address and a dial up networking
connection. In my case, the local ethernet card needs to stay connected
since this machine also acts as a proxy for the internet connection. The
only work around I have found is to have someone else connect to me, and
then CU realizes what address it should be using. This only appears to
work on WP, cornell did not seem to change it's preferred address.
White Pine blames this on Microsoft, but personally I believe that it is
a normal situation to have more that a single network connection, and
that there SHOULD be a way (command line, preference selection, etc) to
select which adapter to use as a primary for the CU program. Hopefully
some of the development team is listening ;-)

I agree it is a binding order issue, I have seen some posts saying to
remove DUN & reinstall DUN, so that it is the LAST adapter that has been
installed, but this did not seem to work in my case. White Pine tech
support also suggested installing the latest DUN, which also had no

You did mention connecting connecting to a self reflect conference,
would you possibly be able to point me in the direction of one of

Glad there is a solution for this in the NT world, hopefully there will
be one in the Win95 worl as well ...

>From Sat May 9 02:12:51 1998
>Received: from (localhost [])
> by (8.8.8/8.8.8) with SMTP id FAA02551;
> Sat, 9 May 1998 05:05:38 -0400 (EDT)
>Received: from ( [])
> by (8.8.8/8.8.8) with ESMTP id BAA00414
> for <>; Sat, 9 May 1998 01:22:44
-0400 (EDT)
>Received: (from daemon@localhost)
> by (8.8.5/8.8.5) id BAA24586
> for; Sat, 9 May 1998 01:22:29
-0400 (EDT)
>Received: from ( [])
> by (8.8.5/8.8.5) with ESMTP id BAA24567
> for <>; Sat, 9 May 1998 01:22:27 -0400 (EDT)
>Received: from grpimo_ian ( [])
> by (8.8.8/8.8.8) via SMTP id XAA03604; Fri, 8 May
1998 23:22:20 -0600 (MDT)
> env-from (
>Message-Id: <>
>Date: Fri, 08 May 1998 23:16:53 -0600
>Precedence: bulk
>From: Ian <>
>To: <>
>Subject: Re: NT and Cornell V1.0 TCP problem
>In-Reply-To: <01bd7afc$fed52250$d55d8bcd@ikepc>
>Mime-Version: 1.0
>Content-Type: text/plain; charset="us-ascii"
>X-To: "ike relucio" <>
>X-Mailer: QUALCOMM Windows Eudora Pro Version 3.0.5 (32)
>X-PH: (Cornell Modified)
>X-Listprocessor-Version: 7.2(a) -- ListProcessor by CREN
>Hash: SHA1
> Your guess was right to assume that I am not an alpha tester for
>this new version. I just wasn't aware that non-testers couldn't ask
>non-documented questions on a CU list :-) Yes, I have used CU before
>(about 1.5 years now). These errors also occured on Cornell version
>0.92b2 as well as WP version 3.1.1.xxxxx. WP's are mighty buggier
>than Cornell's alpha releases if you ask me...
> I am well aware that I have two IP addresses when I dial up to the
>Internet. Yes this probably does confuse some software. I just
>thought that I might post my question just in case anyone else was
>experiencing this problem. I think I have found a fix which is now
>working well for me.
> It was most definately a "binding order problem". CU was just
>grabbing the first IP and adapter it saw in the binding order list.
> Since a NIC in Windows NT can have up to 25 IP addresses assigned
>per card, there has to be a binding order to these IPs. Five
>addresses per card may be configured using Control Panel; however,
>more may be added in the Registry. For details, see the IPAddress
>registry parameter in online Registry Help. NetBT (NetBIOS over TCP/IP
>per RFC 1001/1002) binds to only one IP address per interface card.
>When a NetBIOS name registration is sent, it contains the first IP
>address in the list of addresses assigned to the NIC. This was the IP
>that CU was attaching to. (MS KBASE Q165180, NT Server 4.0 Resource
>Kit: Networking Guide)
> If you have more than one adapter like a WAN Wrapper (AKA "Dial-up
>adapter" in Windows 95) and an ethernet LAN card, there *must* be a
>order to which these adapters are bound to their assigned IP
>addresses. All I did was change the order in which these *adapters*
>were bound. The reason it didn't work the first time was because I
>was re-ordering the protocols instead of the adapters. I neglected to
>select the "all adapters" pulldown menu located in the upper right of
>the control panel network binding window. So, I moved the WAN Wrapper
>to the top of the binding order list which seemed to solve the
>I also thought that it might have been the routing table as well. The
>routing table gave me a clue that the binding order might be the
>problem. When I ran "route print" the ethernet adapter's routing
>table showed up first with the WAN Wrapper following.
> Since I am running a non-routable IP as defined in RFC 791, turning
>on IP routing didn't work. The thing I think is strange (but cool) is
>how I am getting the reflector's conference list when CU uses
>as an IP. This only happens with Cornell's version 0.96a17.
>At 11:46 AM 5/9/98 +0800, you wrote:
>>i would guess that you are not one of the alpha testers of this new
>>have you also used cu-seeme before ? what version ? did you encounter
>>same problem ?
>>the problem is the fact that once you dial-in to your ISP, your
>>effectively now has two IP addresses --- the one for your LAN and the
>>assigned to you by your ISP. this confuses cu and probably a lot of
>>i don't believe there is anything like an "IP binding order" --- your
>>adapter card needs an ip and your dial-up networking adapter needs an
>ip ---
>>it's as simple as that. an adapter card can only have one ip address
>>associated with it so there isn't any 'binding order'
>>i don't have a lan card with a second ip address installed in my
>system but
>>i think the problem is in the way the ip routing table gets messed up
>>you dial-in to your isp. try comparing the output of the 'route
>>command before and after you dial-in to your isp. you will probably
>need to
>>run some batch file after dialing-in to correct the routing table.
>you will
>>probably need to set it in such a way that ip addresses that belong
>to your
>>lan network (10.x.x.x) gets routed to your lan network card and
>>else gets routed to the ip address assigned by your isp.
>>windows help entry for the 'route' command doesn't give you much info
>on how
>>to use the utility.
>>you might also need to turn on 'enable ip forwarding' in the tcp/ip
>>what happens if tcp/ip uses your lan ip address when it sends a
>packet out
>>to the internet ? well, since the lan ip address is not defined in
>any DNS
>>or router on the internet, there doesn't seem to be a way for the
>>packets to find their way back.
>>you might be better of asking yourself the question: "do i really,
>>need to be connected to tha lan at the same time that i want to
>connect to
>>the internet ?"
>>if your answer is no then make two hardware configurations and in one
>>configuration, enable the lan adapter and in the other one, disable
>it. so
>>when you want to connect to the internet, reboot and select the
>>configuration with the lan adapter card disabled, this way your lan
>>address will not be defined.
>Version: PGPfreeware 5.5.5 for non-commercial use <>

Get Your Private, Free Email at