Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Posting-Version: version B 2.10.2 9/18/84 exptools; site ihwpt.UUCP
Path: utzoo!watmath!clyde!burl!ulysses!mhuxr!mhuxn!ihnp4!ihwpt!dafa
From: dafa@ihwpt.UUCP (David Fay)
Newsgroups: net.micro.mac
Subject: Re: Red Ryder questions.
Message-ID: <657@ihwpt.UUCP>
Date: Sat, 25-Jan-86 23:54:49 EST
Article-I.D.: ihwpt.657
Posted: Sat Jan 25 23:54:49 1986
Date-Received: Sun, 26-Jan-86 17:13:55 EST
References: <1170@utai.UUCP> <277@ism780c.UUCP> <651@ihwpt.UUCP> <298@ism780c.UUCP> <8671@amdcad.UUCP>
Organization: AT&T Bell Laboratories
Lines: 57

> In article <298@ism780c.UUCP> tim@ism780c.UUCP (Tim Smith) writes:
> >I agree that a sender initiated protocol can be nice.  But I think it
> >should be done by having the sender send something that means "I am
> >starting an xmodem ( or kermit, or whatever ) transfer", and then
> >using a standard xmodem or kermit or whatever.  Then if you have a
> >terminal program that knows about this, it can receive the file
> >automatically, and if you have one that doesn't, you can do a receive
> >command manually ( or menuly :-) ).
> 
> Correct me if I'm wrong, but isn't this what you get in MacTerminal
> when you select xmodem - other ( as opposed to Mac-to-Mac).

No. When you select xmodem - other in MacTerminal, you are selecting
a vanilla xmodem protocol, which is receiver initiated. You can tell,
because selecting it (as opposed to xmodem - macterminal) activates
the Receive File item under the File menu. In order for the transfer
to take place, the receiving machine must send out a signal that it
is ready to receive (which is what the program does when you select
Receive File). The transfer then takes place with a normal xmodem
protocol.

Selecting xmodem - macterminal, on the other hand, puts MacTerminal
in a mode in which it is constantly looking for an ESCAPE followed by
an "a". If it gets one it immediately goes into receive file mode
and the "thermometer" appears on the screen. This is what is called
a sender initiated transfer. The receiving machine doesn't have to
send anything out to initiate the transfer.

What Tim Smith is proposing, and I think he is right on the mark, is
that all transfers be sender initiated. A sender would precede the
transfer with a signal identifying the type of transfer. The terminal
program would constantly be looking for these signals. When it saw one,
it would go into the appropriate mode. The facility for
doing this is already built into MacTerminal and MacBinary
transfers. MacTerminal transfers are initiated with "ESC a", as already
mentioned, and MacBinary with "ESC b". It is a natural extension to
initiate Kermit transfers with "ESC k" or some such. In fact I am
told that Kermit, which I have never used, does sender initiated
transfers so I assume it is already set up with a some kind of
request-for-transfer signal.

It is a mystery to me why authors of terminal programs have ignored
the identifying signal of MacBinary. Dennis Brothers put it into the
MacBinary proposal specifically so that MacBinary could be sender
initiated. Yet no program to my knowledge takes advantage of it.

A friend and I have been working on this stuff recently because we
are setting up a Mac bulletin board running on a Unix machine. My friend
will be sending out a new version of macput shortly that moves it in the
direction that Tim Smith proposes.

-- 
--------------------
David Fay
AT&T Bell Laboratories
Naperville, IL 60566
ihnp4!ihexp!dafa