Re: RES: ReN: IPv6 tranisition issues

"Michael R. Cole" <[email protected]> Fri, 10 Jan 2003 04:12:49 -0500
Newsgroups gmane.ietf.ngtrans
Message-ID <002301c2b888$735e2d10$ca00a8c0@k6500a>
----- Original Message -----
From: "Christian Huitema" <[email protected]>
To: "Michael R. Cole" <[email protected]>; <[email protected]>
Sent: Wednesday, January 08, 2003 11:42 AM
Subject: RE: RES: ReN: (ngtrans) IPv6 tranisition issues


>
>
> > -----Original Message-----
> > From: Michael R. Cole [mailto:[email protected]]
> > Sent: Tuesday, January 07, 2003 11:00 PM
> > To: [email protected]
> > Subject: Re: RES: ReN: (ngtrans) IPv6 tranisition issues
> >
> >
> > ----- Original Message -----
> > From: "Fred L. Templin" <[email protected]>
> > To: "Keith Moore" <[email protected]>
> > Cc: "Marcelo Barbosa Lima" <[email protected]>; <[email protected]>;
> > <[email protected]>
> > Sent: Monday, January 06, 2003 2:03 PM
> > Subject: Re: RES: ReN: (ngtrans) IPv6 tranisition issues
> >
> >
> > > Keith,
> > >
> > > No offense intended, but I am beginning to suspect that you have
> > > somehow implemented an "automatic insult generator" in your mailer
> > > that is triggered anytime someone mentions the word "NAT". :^}
> > >
> > > Fred Templin
> > > [email protected]
> > >
> > > Keith Moore wrote:
> > > > On Mon, 6 Jan 2003 08:56:20 -0200
> > > > "Marcelo Barbosa Lima" <[email protected]> wrote:
> > > >
> > > >
> > > >>  I think that NAT is a good workaround and it can be done to
> support
> > all aplications
> > > >
> > > >
> > > > you are mistaken.  your statement is completely and utterly
> incorrect.
> > > >
> > > > Keith
> > >
> > >
> > >
> >
> > Keith Moore SHOULD have written:
> >
> > NAT breaks Unix applications but not Windows applications.
>
> Uh? For one, many networking application are platform independent, and
> are equally broken on Unix and Windows. Then, one of the reason
> Microsoft is pushing IPv6 is precisely because many high profile Windows
> applications are broken by NAT. Examples of broken applications are:
> Netmeeting; remote access to your home computer using "terminal server"
> or "remote assistance"; hosting a server on your home PC for DirectPlay
> games; the Voice over IP, video over IP, file transfer and application
> features of Messenger. These are just examples.
>
> Obviously, we at Microsoft are not sitting on our hands watching
> applications fail, and we have developed workarounds for some scenarios.
> In particular, we can traverse those NATs that support the UPNP
> "Internet Gateway Device" profile. But these workarounds are brittle;
> they don't work in all scenarios, they make application programming,
> testing and maintenance more complex; and they tend to generate calls to
> the support line. In particular, there is no solution in double NAT
> scenarios.
>
> -- Christian Huitema
>

The difference between you and Keith is that Keith was NOT willing to
mention which applications and NAT box protocols are problematic. It was as
if he were pulling his reasoning out of thin air and just not bothering to
cite some examples. You also gave me some insight as to why I have been able
to get Sygate Home Network to work very well and Wingate refused to work
with any of my applications. What I have noticed is that the better NAT
software uses both client and server-gateway modules. Is that what is needed
to implement Internet Gateway Device protocol?

Alas, I do not use Netmeeting and I do not have enough bandwidth to use
VOIP. I am also not too crazy allowing remote access to my computer. I am
also willing to bet that most residential NAT box will not support a server
but if you really want to run a web page you need to hire a server farm or
something like that anyways.

Firewalls and popup stoppers also break applications but in these cases the
idea is to deliberately break applications - one time General Motors Finance
thought it would be really cute if they highjacked my computer.

Thanks,

MIke Cole, [email protected]