[ih] FTP design
day at std.com
Sat Aug 3 06:45:07 PDT 2002
At 07:14 -0400 8/3/02, David P. Reed wrote:
>I remember all the things you do John (but my connection to the FTP
>design, pre-internet as it was, was through Ken Pogran, who did the
>work for it on Multics). I'd just like to add my comment that
>"separate control and data connections" was hardly a mistake. It
>was very much on purpose, and for exactly the reason that John
>points out. Rather than invent a magical "out of band" feature in
>TCP, which would have multiplied the congestion control
>complexities, as well as introducing problems of mixing data and
>control coding schemes, just using two connections was simple.
Completely agree. Also, in those days FTPs could take awhile and OSs
had a tendency to go off and do other things for awhile. So there
was a desire to be able to send other commands while the transfer was
going on. If for no other reason than to determine that the other
end was still thinking about it.
>As far as the directionality of streams, one idea that was
>contemplated was the idea of having control host A create a transfer
>from host B to host C, while not having a control connection from B
Yes, the third party transfer was part of the spec.
>Finally, but Ken Pogran can clarify this more if you talk to him,
>FTP's multiple connection scheme was somewhat hard to implement in
>Multics security model, but rather than have Multics influence the
>design, the Multics guys figured out how to generalize the
>connection notification mechanism to get this to work right.
Yea, I vaguely remember that. Was that also related to the
convention that the data transfer sockets were a fixed offset from
the control (telnet) connection?
But I don't remember thinking there were any real "mistakes" with
FTP. I remember thinking after we finalized the last version in 73
that we were getting an idea of how to do richer FTP with file access
and make file systems more transparent over the net, but that was
really icing on the cake. Not really any big insights.
I have always contended that the biggest problem with the ARPANet was
that we got too much stuff right to begin with!! So that it worked
well enough that there wasn't really enough pain to warrant going
back to fix it. ;-) To some extent we are paying for that now.
More information about the Internet-history