cc:Mail

[Grant Taylor]

On 10/7/20 2:10 PM, Tomas By wrote:
> Well, we are talking past each other.

That's entirely possible.  That's why I elaborated on what I meant, 
explicitly to give you an opportunity to confirm or refute.

> When I say client/server I mean the connection over serial port/modem 
> between the mobile client, not on the same LAN as the PO, and the PO.

Oh.  What I know as the remote client support.  I know even less about that.

But, yes, that could probably be a daemon of sorts without an actual P.O.

> Yes.
> 
> Well, I am not sure either. I need to have another look at that SMTP 
> gateway, I guess.

:-)

> But if it is like you say they would not have needed a separate 
> Internet bridge? They could just have used the SMTP gateway?

It might have been a different limitation from the time, imposed by the 
configuration in play.

I see a lot of documentation talking about doing SMTP between a PC 
running DOS or NT (the two platforms that the MS-Mail SMTP Gateway ran 
on) and a Unix (typically Solaris) box that would gateway between SMTP 
and UUCP.

I think the P.O. usually lived on one system which a lot of people 
accessed.  Then the SMTP gateway lived on another system, where it would 
gateway between the P.O. on one side ans SMTP on another side.

Given the purported instability of the various gateways, I think there 
was a fair bit of advantage in having it on it's own system so that it 
could be rebooted without impacting anything else.  The same would not 
be true if the gateway was run on the file server holding the P.O.

There's also using the SMTP gateway between multiple P.O.s, each on 
different departmental file servers, and the Internet.

All speculation about something I was not part of and have only read 
things about.  --  Save for the fact that I did get MS-Mail to SMTP 
Gateway working in a lab within the last couple of years.



-- 
Grant. . . .
unix || die