Grant,
> Seeing as how you are wanting to do something akin to a proxy for TCP
> traffic, I would suggest that you use a USB-to-Ethernet dongle for a 2nd
> Ethernet interface.
Thanks. I already got that idea too, but as I know almost nothing about the
possibilities ...
> As others have stated, a Layer 2 managed switch using 802.1Q VLAN tags is
> a perfectly viable option too.
>
> Which you use will probably depend on what you have available and / or
> want to purchase.
Well, I'm most always going for a solution that is easiest to manage and has
the largest usage area - even when I almost certainly will never move the
thing. IOW, when executed my choice will be the dongle.
>> No, no intermediate network.
>
> A routed network doesn't necessarily mean that there is something in
> between the networks.
As far as I can see B is in between the AB and BC networks ...
But yes, I was imagining a routing over yet another LAN segment.
> Do you want A and C to be able to establish TCP connections with each
> other? E.g. through a router of some sort?
Nope, never.
> Or do you want A to *ONLY* be able to establish a TCP connection with B,
> and likewise with B & C?
Yep, that. And to be honest, I could not care less how it happens, as
long as they can exchange data without actually being able to connect set up
random connections to each other. Hence my side-step in the direction of a
serial connection. Data can flow, but without specially written software
that data comes from nowhere and goes to nowhere else than from/into my
program.
> This may seem like a minutia difference. But it is a very important
> difference. If you're okay with A and C establishing a TCP connection
> with each other through an intermediate router (B), then much of this
> conversation is moot. Like a mic drop.
:-) I'm quite aware of that.
> Now you're talking more about something between NAT and an ALG / Proxy.
Not quite. As I said, the port information is dropped too. Always.
> You don't even need Samba / NFS / etc. on the intermediate host B.
Indeed.
> A will /think/ that it's talking to B when in fact it's /really/ talking
> to C. Likewise, C will /think/ that it's talking to B when in fact it's
> /really/ talking to A. It's just that A and C will have a different idea
> of what each other's address is.
Yep. But with a small twist: In my initial idea B was truly passive.
Receiving a connection from A would not provoke it to try to make a
connection to C. It would do nothing until C also connected to it. Only
than it would exchange datapackets.
Currently I'm cautiously sporting the idea that /maybe/ B could, when
receiving a connection from either side, open a connection to the other side
itself - on a fixed port ofcourse.
> No. Not as such. UUCP works by running a command on A that tells UUCP on
> A, B, and C, to transfer the file to C.
Ah, I misunderstood what it might do.
> Remember, even a file drop, can enable programs on A and C to communicate
> with each other.
Yes, I'm aware of that. The whole idea is that I /want/ them to
communicate - but only on my conditions.
>> Agreed. But I would not have any qualms to install SAMBA twice, once for
>> each interface, only coming together on the file-storage level.
>
> You only install Samba once. You only need to run one instance of it.
I understood that. It was just to make a point as to how far I would go to
keep everything apart.
>> maybe automatic clipboard exchange later.
>
> Now you're talking about something a LOT more involved.
Not really. I wrote something like that a number of years ago. Text only
though. That was all I needed it for. And I never actually finished the
"automatic" part though. I would have needed a data-hash check to stop data
storms, but eventually dropped that idea. Instead I simply used a "send"
button. Served two purposes: No data storms, and being able to copy-paste
private data without automatically having it land on other 'puters.
Regards,
Rudy Wieser
--- SoupGate-Win32 v1.05
* Origin: Agency HUB, Dunedin - New Zealand | FidoUsenet Gateway (3:770/3)
|