[ih] First file transfer on ARPANET

[John Day]

You are very likely right.  Also given that terminal handling in most 
OSs of the day was some of the strangest code you could come across. 
;-)  It was a place of many dragons.  Pseudo terminals were benign by 
comparison!!  ;-)

If you want to count test programs that just pushed data down the 
connection, file transfer was very likely first.


At 13:24 -0500 2012/12/12, Noel Chiappa wrote:
>     > From: Richard Bennett <richard at bennett.com>
>
>     > Given that the initial focus was remote login
>
>I wasn't there, and I don't offhand recall what I've seen about the timelines
>in historical materials I have read, but a word from personal memory about an
>analogous situation:
>
>When working on early TCP implementations, on machines which had no
>networking software on them at all (as was still common in '78), it was 'far'
>easier to bring up some sort of rudimentary file transfer than remote login.
>(In fact, TFTP was invented to allow file transfer before we even had TCP
>running, and IIRC EFTP was done for similar reasons at PARC - but don't quote
>me on that, that's just a dim recollection of something I read a long time
>ago.)
>
>Remote login takes a _lot_ more work on the server side (you have to create
>pseudo-teletypes, and hook them into the terminal handling code, and in
>general, depending on how involved the OS is, it can be a fair amount of work
>to handle remote users), whereas for simple file service (i.e. no
>login/authentication, just access to whatever's world readable/writeable),
>it's a very small amount of code.
>
>So my _guess_ is that while formal plans may have been to work on remote
>login, I'll bet whoever was actually writing code did some sort of file
>transfer first...
>
>
>     > From: John Day <jeanjour at comcast.net>
>
>     > I should point out that the ARPANET never did do a remote login
>     > protocol.
>     > ...
>     > The Telnet spec quite specifically says it is a terminal device driver
>     > protocol.
>
>You're splitting a rather fine hair (although I concede the accuracy of that
>hair): The TELNET protocol may _be_ a terminal device driver protocol, but it
>was mostly _used_ for remote login - and the server to which one connected at
>port 23 using it was a remote login service...
>
>	Noel