Re: Apple II Email (Network Connections)

[derek@slab.cisco.com (Derek Taubert)]

In <544e4o$acl@login.freenet.columbus.oh.us> Dave Althoff wrote:
> My apologies if this has already been covered, but with my extremely
> limited network knowledge, I'll throw in my $0.02...
> 
> It seems to me that there exist network servers designed to be connected
> via some sort of archaic router to dumb terminals.  This is where I know
> very little; perhaps it is as simple as using a multiport terminal server
> such as a mainframe server would use to talk to a bank of modems.  With
> that in mind, could one not use, for instance, a Linux server talking
> through such a network, cabled with RS-232-style wiring?  Then at the most
> basic level, each station could use a comm program to log in to the
> network server, collect mail, deposit outgoing mail messages, and log out.
>  Or not, as the case may be.

Ding!  Ding!  Ding!  This is the most viable solution to the problem that 
I've heard yet.

> The problem with this arrangement, of course, is the network router.  To
> the best of my knowledge, this type of serial-port networking is all but
> extinct (I say "all but" because such a network does exist in the building
> where I work...connected indirectly to the University Sun-box cluster and
> more directly to the now-obsolete administrative computer).  But the
> network hardware might be obtainable as commercial/institutional network
> users modernize their computer systems and install PC's and Ethernet.

Not true!  Many companies (like Cisco) make quite a bit of money selling 
multiport asynchronous serial terminal servers!  The Cisco 2511 for example 
has 16 serial ports on one end of the box, and ethernet on the other.  It 
would probably be cheaper and faster to let the terminal server translate 
TCP/IP than to try to develop something to make an Apple ][ do it.  Run a 
serial line for each Apple ][ and use your favorite SSC and terminal program 
to log in to a host on the ethernet network (such as a linux box).

> At the other end, this becomes a potentially useful system, since writing
> a mail client program for a ][e...to log into the network, pull down the
> user's mail, process the mail files, and upload outgoing messages...could
> be an almost trivial task, compared with the Herculean task of trying to
> write drivers to speak Appletalk via SSC, or fake an Ethernet connection,
> or whatever...on the ][e side, really the whole problem would be with
> developing the client software.  And the best part of it is that if
> current Internet mail standards are adopted, then the client program (with
> modifications...remove the log-on and capture feature) would be useful to
> users not connected to the sort of network your project describes.

I agree 100% with the above evaulation of the situation.  It would be so much 
easier to write a mail client to use simple ASCII transfers over a serial 
line and a corresponding server on the Unix end than to try to coax an Apple 
][ to speak TCP/IP.

--
 _             ___                    
| \ _ ._ _ |    | _.   |_  _ ._ _|_   taubert@geeks.org
|_/(/_| (/_|<   |(_||_||_)(/_|   |_   NeXTMail welcome